<?xml version="1.0" encoding="UTF-8"?>
<!DOCTYPE article PUBLIC "-//NLM//DTD Journal Publishing with OASIS Tables v3.0 20080202//EN" "journalpub-oasis3.dtd">
<article xmlns:xlink="http://www.w3.org/1999/xlink" xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:oasis="http://docs.oasis-open.org/ns/oasis-exchange/table" xml:lang="en" dtd-version="3.0">
  <front>
    <journal-meta><journal-id journal-id-type="publisher">GMD</journal-id><journal-title-group>
    <journal-title>Geoscientific Model Development</journal-title>
    <abbrev-journal-title abbrev-type="publisher">GMD</abbrev-journal-title><abbrev-journal-title abbrev-type="nlm-ta">Geosci. Model Dev.</abbrev-journal-title>
  </journal-title-group><issn pub-type="epub">1991-9603</issn><publisher>
    <publisher-name>Copernicus Publications</publisher-name>
    <publisher-loc>Göttingen, Germany</publisher-loc>
  </publisher></journal-meta>
    <article-meta>
      <article-id pub-id-type="doi">10.5194/gmd-11-1257-2018</article-id><title-group><article-title>Intercomparison of Antarctic ice-shelf, ocean, and sea-ice interactions simulated by MetROMS-iceshelf and FESOM 1.4</article-title><alt-title>Intercomparison of Antarctic ice-shelf, ocean, and sea-ice interactions</alt-title>
      </title-group><?xmltex \runningtitle{Intercomparison of Antarctic ice-shelf, ocean, and sea-ice interactions}?><?xmltex \runningauthor{K. A. Naughten et al.}?>
      <contrib-group>
        <contrib contrib-type="author" corresp="yes" rid="aff1 aff2 aff3">
          <name><surname>Naughten</surname><given-names>Kaitlin A.</given-names></name>
          <email>k.naughten@unsw.edu.au</email>
        <ext-link>https://orcid.org/0000-0001-9475-9162</ext-link></contrib>
        <contrib contrib-type="author" corresp="no" rid="aff1 aff2">
          <name><surname>Meissner</surname><given-names>Katrin J.</given-names></name>
          
        </contrib>
        <contrib contrib-type="author" corresp="no" rid="aff4 aff3">
          <name><surname>Galton-Fenzi</surname><given-names>Benjamin K.</given-names></name>
          
        <ext-link>https://orcid.org/0000-0003-1404-4103</ext-link></contrib>
        <contrib contrib-type="author" corresp="no" rid="aff1 aff2">
          <name><surname>England</surname><given-names>Matthew H.</given-names></name>
          
        </contrib>
        <contrib contrib-type="author" corresp="no" rid="aff5">
          <name><surname>Timmermann</surname><given-names>Ralph</given-names></name>
          
        </contrib>
        <contrib contrib-type="author" corresp="no" rid="aff5">
          <name><surname>Hellmer</surname><given-names>Hartmut H.</given-names></name>
          
        </contrib>
        <contrib contrib-type="author" corresp="no" rid="aff6 aff5">
          <name><surname>Hattermann</surname><given-names>Tore</given-names></name>
          
        </contrib>
        <contrib contrib-type="author" corresp="no" rid="aff7">
          <name><surname>Debernard</surname><given-names>Jens B.</given-names></name>
          
        </contrib>
        <aff id="aff1"><label>1</label><institution>Climate Change Research Centre, Level 4 Mathews Building, UNSW Sydney, Sydney NSW 2052, Australia</institution>
        </aff>
        <aff id="aff2"><label>2</label><institution>ARC Centre of Excellence for Climate System Science, Australia</institution>
        </aff>
        <aff id="aff3"><label>3</label><institution>Antarctic Climate &amp; Ecosystems Cooperative Research Centre, Private Bag 80, Hobart TAS 7001, Australia</institution>
        </aff>
        <aff id="aff4"><label>4</label><institution>Australian Antarctic Division, 203 Channel Highway, Kingston TAS 7050, Australia</institution>
        </aff>
        <aff id="aff5"><label>5</label><institution>Alfred Wegener Institut, Postfach 12 01 61, 27515 Bremerhaven, Germany</institution>
        </aff>
        <aff id="aff6"><label>6</label><institution>Akvaplan-niva, P.O. Box 6606, Langnes, 9296 Tromsø, Norway</institution>
        </aff>
        <aff id="aff7"><label>7</label><institution>Norwegian Meteorological Institute, P.O. Box 43, Blindern, 0313 Oslo, Norway</institution>
        </aff>
      </contrib-group>
      <author-notes><corresp id="corr1">Kaitlin A. Naughten (k.naughten@unsw.edu.au)</corresp></author-notes><pub-date><day>10</day><month>April</month><year>2018</year></pub-date>
      
      <volume>11</volume>
      <issue>4</issue>
      <fpage>1257</fpage><lpage>1292</lpage>
      <history>
        <date date-type="received"><day>27</day><month>October</month><year>2017</year></date>
           <date date-type="rev-request"><day>1</day><month>November</month><year>2017</year></date>
           <date date-type="rev-recd"><day>20</day><month>February</month><year>2018</year></date>
           <date date-type="accepted"><day>9</day><month>March</month><year>2018</year></date>
      </history>
      <permissions>
        <copyright-statement>Copyright: © 2018 Kaitlin A. Naughten et al.</copyright-statement>
        <copyright-year>2018</copyright-year>
      <license license-type="open-access"><license-p>This work is licensed under the Creative Commons Attribution 4.0 International License. To view a copy of this licence, visit <ext-link ext-link-type="uri" xlink:href="https://creativecommons.org/licenses/by/4.0/">https://creativecommons.org/licenses/by/4.0/</ext-link></license-p></license></permissions><self-uri xlink:href="https://gmd.copernicus.org/articles/11/1257/2018/gmd-11-1257-2018.html">This article is available from https://gmd.copernicus.org/articles/11/1257/2018/gmd-11-1257-2018.html</self-uri><self-uri xlink:href="https://gmd.copernicus.org/articles/11/1257/2018/gmd-11-1257-2018.pdf">The full text article is available as a PDF file from https://gmd.copernicus.org/articles/11/1257/2018/gmd-11-1257-2018.pdf</self-uri>
      <abstract><title>Abstract</title>
    <p id="d1e186">An increasing number of Southern Ocean models now include Antarctic ice-shelf
cavities, and simulate thermodynamics at the ice-shelf/ocean interface. This
adds another level of complexity to Southern Ocean simulations, as ice
shelves interact directly with the ocean and indirectly with sea ice. Here,
we present the first model intercomparison and evaluation of present-day
ocean/sea-ice/ice-shelf interactions, as simulated by two models: a
circumpolar Antarctic configuration of MetROMS (ROMS: Regional Ocean
Modelling System coupled to CICE: Community Ice CodE) and the global model
FESOM (Finite Element Sea-ice Ocean Model), where the latter is run at two
different levels of horizontal resolution. From a circumpolar Antarctic
perspective, we compare and evaluate simulated ice-shelf basal melting and
sub-ice-shelf circulation, as well as sea-ice properties and Southern Ocean
water mass characteristics as they influence the sub-ice-shelf processes.
Despite their differing numerical methods, the two models produce broadly
similar results and share similar biases in many cases. Both models reproduce
many key features of observations but struggle to reproduce others, such as
the high melt rates observed in the small warm-cavity ice shelves of the
Amundsen and Bellingshausen seas. Several
differences in model design show a particular influence on the simulations.
For example, FESOM's greater topographic smoothing can alter the geometry of
some ice-shelf cavities enough to affect their melt rates; this improves at
higher resolution, since less smoothing is required. In the interior Southern
Ocean, the vertical coordinate system affects the degree of water mass
erosion due to spurious diapycnal mixing, with MetROMS' terrain-following
coordinate leading to more erosion than FESOM's <inline-formula><mml:math id="M1" display="inline"><mml:mi>z</mml:mi></mml:math></inline-formula> coordinate. Finally,
increased horizontal resolution in FESOM leads to higher basal melt rates for
small ice shelves, through a combination of stronger circulation and
small-scale intrusions of warm water from offshore.</p>
  </abstract>
    </article-meta>
  </front>
<body>
      

<sec id="Ch1.S1" sec-type="intro">
  <title>Introduction</title>
      <p id="d1e203">The Antarctic Ice Sheet (AIS) has significant potential to drive sea level
rise as climate change continues
<xref ref-type="bibr" rid="bib1.bibx13 bib1.bibx39 bib1.bibx135 bib1.bibx107" id="paren.1"/>. Palaeo records
indicate that the AIS was a major contributor to sea level change in past
climate events <xref ref-type="bibr" rid="bib1.bibx8 bib1.bibx108 bib1.bibx131 bib1.bibx21 bib1.bibx121" id="paren.2"/>,
and the mass balance of the modern-day AIS is already negative
<xref ref-type="bibr" rid="bib1.bibx133 bib1.bibx178 bib1.bibx146" id="paren.3"/>. The ocean is an important driver
of AIS retreat <xref ref-type="bibr" rid="bib1.bibx40 bib1.bibx78" id="paren.4"/>, as 40 % of the ice sheet
by area is grounded below sea level <xref ref-type="bibr" rid="bib1.bibx31" id="paren.5"/>. This geometry
provides the<?pagebreak page1258?> potential for the ocean to melt large regions of the AIS from
below. For example, the Amundsen sector of West Antarctica has bedrock
geometry favourable for a marine ice-sheet instability, and unstable retreat
may have already begun <xref ref-type="bibr" rid="bib1.bibx135" id="paren.6"/>.</p>
      <p id="d1e225">The ocean directly interacts with the AIS through ice shelves, which are the
floating extensions of the land-based ice sheet. The properties of ice-shelf
cavities, the pockets of ocean between ice shelves and the seafloor,
determine the basal melt rates of each ice shelf which ultimately affect the
mass balance of the AIS through dynamical processes <xref ref-type="bibr" rid="bib1.bibx19" id="paren.7"/>. The
seawater in ice-shelf cavities can be sourced from several different water
masses, which affect its temperature and salinity. Many of these source water
masses are influenced by sea-ice processes <xref ref-type="bibr" rid="bib1.bibx71 bib1.bibx120" id="paren.8"/>.</p>
      <p id="d1e234">A better understanding of ocean/sea-ice/ice-shelf interactions in Antarctica
is crucial, particularly given their importance for future sea level rise.
However, these interactions take place in observation-deficient regions. In
particular, there are very few direct measurements inside ice-shelf cavities,
and observations are also scarce in the sea-ice-covered regions of the
Southern Ocean <xref ref-type="bibr" rid="bib1.bibx136" id="paren.9"/>. Nonetheless, some measurements have been
made at great expense (e.g. <xref ref-type="bibr" rid="bib1.bibx119 bib1.bibx102 bib1.bibx169" id="altparen.10"/>).
While ice-shelf basal melt rates can be inferred using remote sensing methods
<xref ref-type="bibr" rid="bib1.bibx134 bib1.bibx15" id="paren.11"/>, large uncertainties remain regarding the
circulation patterns driving these melt rates, and no predictions for the
future can be made based on these data.</p>
      <p id="d1e246">Consequently, much of our understanding of ocean/sea-ice/ice-shelf
interactions is based on numerical modelling. In recent years, an increasing
number of ocean models have begun to resolve ice-shelf cavities and simulate
thermodynamic processes at the ice-shelf base (<xref ref-type="bibr" rid="bib1.bibx17" id="altparen.12"/>, and
references therein; <xref ref-type="bibr" rid="bib1.bibx101" id="altparen.13"/>). Given the variety of models
involved, and the relative lack of observations to constrain their tuning, it
is desirable to conduct model intercomparison projects (MIPs; see, e.g.
<xref ref-type="bibr" rid="bib1.bibx104" id="altparen.14"/>) by which several models run the same experiment and
their output is compared. The resulting insights into model similarities and
differences can ideally be attributed to model design choices, with the aim
of guiding future development.</p>
      <p id="d1e259">To date, the only MIPs considering ice-shelf cavities are the ongoing ISOMIP
experiments (Ice Shelf-Ocean Model Intercomparison Project)
<xref ref-type="bibr" rid="bib1.bibx67 bib1.bibx2" id="paren.15"/> which use idealised domains and simplified
forcing, and do not include coupled sea-ice models. The ISOMIP experiments
are undoubtedly valuable and are likely to provide particular insights
regarding the response of cavity circulation to warm versus cold forcing.
However, idealised experiments such as ISOMIP should be complemented by
intercomparisons over more realistic domains, with observationally derived
forcing and coupled sea-ice models. These model configurations are already
being used to better understand processes in observed cavities
<xref ref-type="bibr" rid="bib1.bibx164 bib1.bibx35" id="paren.16"/> and to provide future projections of
ice-shelf melt <xref ref-type="bibr" rid="bib1.bibx161 bib1.bibx57 bib1.bibx58" id="paren.17"/>, so analysis of
the similarities and differences between such models is timely. Another
important benefit of realistic domains is the opportunity to compare model
output to available observations, even if these observations are limited.
Therefore, an element of model evaluation, as well as model intercomparison,
can be included.</p>
      <p id="d1e271">In this paper, we present such an intercomparison of two ocean models, both
including ice-shelf thermodynamics and sea-ice components, from a circumpolar
Antarctic perspective. We focus on ice-shelf basal melt and sub-ice-shelf
circulation across eight regions of the Antarctic coastline but also
consider interior Southern Ocean and sea-ice processes as they affect
ice-shelf cavities. The model output is compared to relevant observations where
available. Finally, key findings and their implications, as well as
possibilities for future model development, are discussed.</p>
</sec>
<sec id="Ch1.S2">
  <title>Model descriptions</title>
      <p id="d1e280">Two coupled ocean/sea-ice/ice-shelf models are included in this
intercomparison: MetROMS-iceshelf (hereafter MetROMS) and Finite Element
Sea-ice Ocean Model 1.4 (hereafter FESOM). We run FESOM at two different
resolutions for a total of three experiments (see
Sect. <xref ref-type="sec" rid="Ch1.S3"/>). In this section, we describe the two models
and compare their scientific design.</p>

      <?xmltex \floatpos{t}?><fig id="Ch1.F1" specific-use="star"><label>Figure 1</label><caption><p id="d1e287">Horizontal resolution (km) of the MetROMS grid and both FESOM meshes
around Antarctica. Resolution is defined as the square root of the area of
each grid box (MetROMS) or triangular element (FESOM). Note that values above
20 km are not differentiated.</p></caption>
        <?xmltex \igopts{width=341.433071pt}?><graphic xlink:href="https://gmd.copernicus.org/articles/11/1257/2018/gmd-11-1257-2018-f01.png"/>

      </fig>

<sec id="Ch1.S2.SS1">
  <title>Overview</title>
      <p id="d1e301">MetROMS consists of the regional ocean model ROMS (Regional Ocean Modelling
System) <xref ref-type="bibr" rid="bib1.bibx145" id="paren.18"/> including ice-shelf thermodynamics
<xref ref-type="bibr" rid="bib1.bibx35" id="paren.19"/>, coupled to the sea-ice model CICE (Community Ice
CodE) <xref ref-type="bibr" rid="bib1.bibx66" id="paren.20"/> using the coupler MCT (Model Coupling Toolkit)
<xref ref-type="bibr" rid="bib1.bibx90 bib1.bibx68" id="paren.21"/>. The coupling was implemented by the Norwegian
Meteorological Institute <xref ref-type="bibr" rid="bib1.bibx12" id="paren.22"/> and is described in
<xref ref-type="bibr" rid="bib1.bibx115" id="text.23"/>. We use the development version 3.7 of the ROMS code,
version 5.1.2 of CICE, and version 2.9 of MCT.</p>
      <?pagebreak page1259?><p id="d1e323">FESOM  is a global ocean model
<xref ref-type="bibr" rid="bib1.bibx173" id="paren.24"/> with an internally coupled sea-ice model
<xref ref-type="bibr" rid="bib1.bibx11 bib1.bibx162" id="paren.25"/> and ice-shelf thermodynamics
<xref ref-type="bibr" rid="bib1.bibx164" id="paren.26"/>. It has an unstructured mesh in the horizontal,
consisting of triangular elements which allow for spatially varying
resolution. The numerical methods associated with the unstructured mesh are
detailed by <xref ref-type="bibr" rid="bib1.bibx171" id="text.27"/> and <xref ref-type="bibr" rid="bib1.bibx173" id="text.28"/>, while the implementation
of the ice-shelf component is discussed in <xref ref-type="bibr" rid="bib1.bibx164" id="text.29"/>.
<?xmltex \hack{\newpage}?></p>
</sec>
<sec id="Ch1.S2.SS2">
  <title>Domain and resolution</title>
      <p id="d1e352">Our configuration of MetROMS has a circumpolar Antarctic domain with a
northern boundary at <inline-formula><mml:math id="M2" display="inline"><mml:mn mathvariant="normal">30</mml:mn></mml:math></inline-formula><inline-formula><mml:math id="M3" display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula> S. Horizontal resolution is
quarter-degree scaled by cosine of latitude, and the South Pole is relocated
to achieve approximately equal resolution around the Antarctic coastline.
This leads to resolutions (defined as the square root of the area of each
grid box) of approximately 15–20 km in the Antarctic Circumpolar Current
(ACC), 8–10 km on the Antarctic continental shelf, and 5 km or finer at
the southernmost grounding lines of the Ross, Filchner–Ronne, and Amery ice
shelves (Fig. <xref ref-type="fig" rid="Ch1.F1"/>a).</p>
      <p id="d1e372">Our FESOM setup has a global domain with spatially varying horizontal
resolution. Here, we define resolution in FESOM as the square root of the
area of each triangular element; however, this metric may not be truly
comparable to MetROMS. When discussing resolution, the real question is the
smallest flow features that are captured by a mesh of a certain spacing. In
models with such different numerical methods as MetROMS (finite volume) and
FESOM (finite element), the smallest resolved feature may scale differently
with the mesh spacing. Numerical dissipation and stabilisation built into
different time-stepping routines can also influence this effective
resolution. Furthermore, MetROMS employs a staggered Arakawa C grid for the
ocean and B grid for the sea ice <xref ref-type="bibr" rid="bib1.bibx1" id="paren.30"/> by which different
variables are calculated at different locations within each grid box. In
FESOM, all variables are calculated at the same locations (nodes), analogous
to the Arakawa A grid. There is some evidence that this design tends to
resolve fewer features of fluid flow <xref ref-type="bibr" rid="bib1.bibx52" id="paren.31"/>, and indeed FESOM
appears to have a lower effective resolution than finite-difference C-grid
models with comparable nominal grid spacing.</p>
      <p id="d1e381">To account for these uncertainties, as well as to investigate the importance
of resolution on FESOM's performance, we have prepared two meshes: “low
resolution” (Fig. <xref ref-type="fig" rid="Ch1.F1"/>b) and “high resolution”
(Fig. <xref ref-type="fig" rid="Ch1.F1"/>c). The high-resolution mesh has approximately
double the number of 2-D nodes as the low-resolution mesh, but these extra
nodes are not evenly spaced throughout the domain. Outside the Southern
Ocean, the two meshes have virtually identical resolution (not shown),
ranging from 150 to 225 km in the abyssal Pacific, Atlantic, and Indian
oceans, and 50–75 km along coastlines. In the open Southern Ocean,
resolution ranges from 20 to 100 km for the low-resolution mesh and
15–50 km for the high-resolution mesh. Both meshes have finer resolution on
the Antarctic continental shelf (approximately 8–10 km for low resolution,
5–7 km for high resolution) and in ice-shelf cavities (5–10 km for low
resolution, 3–7 km for high resolution). The greatest difference between
the two meshes occurs in the Amundsen and Bellingshausen seas, with
approximate resolution of 11 km for the low-resolution mesh and 4 km for
the high-resolution mesh.</p>
      <p id="d1e388">In the vertical, MetROMS has 31 terrain-following levels using the
<inline-formula><mml:math id="M4" display="inline"><mml:mi>s</mml:mi></mml:math></inline-formula>-coordinate system, with increasing vertical resolution near the surface
and bottom, and coarsest resolution in the interior. FESOM employs a hybrid
<inline-formula><mml:math id="M5" display="inline"><mml:mi>z</mml:mi></mml:math></inline-formula>-<inline-formula><mml:math id="M6" display="inline"><mml:mi mathvariant="italic">σ</mml:mi></mml:math></inline-formula> vertical coordinate system, with the same discretisation for both
the low- and high-resolution meshes. The region south of the 2500 m isobath
surrounding Antarctica, which includes all ice-shelf cavities as well as the
continental shelf and slope, has <inline-formula><mml:math id="M7" display="inline"><mml:mi mathvariant="italic">σ</mml:mi></mml:math></inline-formula> coordinates with 22 levels. The
remainder of the domain has <inline-formula><mml:math id="M8" display="inline"><mml:mi>z</mml:mi></mml:math></inline-formula> coordinates, comprised of 38 levels weighted
towards the surface. Both models are free-surface, which leads to
time-varying vertical levels in MetROMS but only affects the uppermost layer
in FESOM.</p>
      <p id="d1e427">In both models, the use of terrain-following coordinates in the thin water
columns of ice-shelf cavities leads to enhanced vertical resolution, often
finer than 1 m in MetROMS, which limits the time step. Our configuration of
ROMS requires a baroclinic time step of 5 min for stability, with
30 barotropic time steps for each baroclinic. In CICE, the time step is
30 min for both dynamic and thermodynamic processes, and ocean/sea-ice
coupling is also performed every 30 min. FESOM is run with a time step of
10 min for the low-resolution mesh and 9 min for the high-resolution mesh. The
sea-ice model operates on the same time step as the ocean component.</p>
</sec>
<?pagebreak page1260?><sec id="Ch1.S2.SS3">
  <title>Smoothing of bathymetry and ice-shelf draft</title>
      <p id="d1e437">Steep bathymetry can be problematic for terrain-following coordinate ocean
models, as it has the potential to introduce pressure gradient errors
<xref ref-type="bibr" rid="bib1.bibx53" id="paren.32"/>. Both ROMS <xref ref-type="bibr" rid="bib1.bibx144" id="paren.33"/> and FESOM
<xref ref-type="bibr" rid="bib1.bibx171" id="paren.34"/> are designed to minimise this issue with the splines density
Jacobian method for the calculation of the pressure gradient force, which
reduces errors compared to the standard density Jacobian method.
Nevertheless, a particular challenge arises at ice-shelf fronts, which in
reality are cliff faces that can reach several hundred metres in depth but
which models must represent as sloping surfaces. This substantial change in
surface layer depth over as little as one grid cell creates steeply sloping
vertical layers with a large pressure gradient, and numerical errors in the
pressure gradient calculation could drive spurious circulation patterns
across the given ice-shelf front.</p>
      <p id="d1e449">In both models, some amount of smoothing of the bathymetry and ice-shelf
draft is necessary for numerical stability and to reduce pressure gradient
errors. On the other hand, excessive smoothing could alter the geometry of
the ice-shelf cavities to the point where circulation is affected. An
oversmoothed ice-shelf front would be too shallow and gently sloping,
providing a pathway for warm surface waters to easily enter the cavity, where
in reality a physical barrier exists. Near the grounding lines at the backs
of ice-shelf cavities, oversmoothing would remove the deepest ice which melts
most easily <xref ref-type="bibr" rid="bib1.bibx93" id="paren.35"/>. In this situation, the water column thickness
would be overestimated, allowing for greater transport of warm water to the
grounding line. In a coupled ice-sheet/ocean model, <xref ref-type="bibr" rid="bib1.bibx160" id="text.36"/>
demonstrated that increased water column thickness due to a thinning ice
shelf can more than compensate for the reduced melting expected from the
elevated in situ freezing point at the ice-shelf base. Therefore, a delicate
balance must be struck when smoothing model topographies, in order to achieve
the most accurate simulation.</p>
      <p id="d1e458">We prepared the MetROMS and FESOM domains using bathymetry, ice-shelf draft,
and land/sea masks from the RTopo-1.05 dataset <xref ref-type="bibr" rid="bib1.bibx163" id="paren.37"/>.
MetROMS follows a three-step smoothing procedure similar to that of
<xref ref-type="bibr" rid="bib1.bibx91" id="text.38"/>. First, the “deep ocean filter” consists of a single
pass of a Hanning filter (window size 3) over the bathymetry <inline-formula><mml:math id="M9" display="inline"><mml:mi>h</mml:mi></mml:math></inline-formula>, with
variable coefficients designed to remove isolated seamounts. Next, a
selective Hanning filter is repeatedly applied to both <inline-formula><mml:math id="M10" display="inline"><mml:mrow><mml:mi>log⁡</mml:mi><mml:mo>(</mml:mo><mml:mi>h</mml:mi><mml:mo>)</mml:mo></mml:mrow></mml:math></inline-formula> and
<inline-formula><mml:math id="M11" display="inline"><mml:mrow><mml:mi>log⁡</mml:mi><mml:mo>(</mml:mo><mml:msub><mml:mi>z</mml:mi><mml:mtext>ice</mml:mtext></mml:msub><mml:mo>)</mml:mo></mml:mrow></mml:math></inline-formula>, where <inline-formula><mml:math id="M12" display="inline"><mml:mrow><mml:msub><mml:mi>z</mml:mi><mml:mtext>ice</mml:mtext></mml:msub></mml:mrow></mml:math></inline-formula> is the ice-shelf draft, until the
slope parameter <inline-formula><mml:math id="M13" display="inline"><mml:mrow><mml:mi>r</mml:mi><mml:mo>=</mml:mo><mml:mo>|</mml:mo><mml:msub><mml:mi>h</mml:mi><mml:mi>i</mml:mi></mml:msub><mml:mo>-</mml:mo><mml:msub><mml:mi>h</mml:mi><mml:mrow><mml:mi>i</mml:mi><mml:mo>+</mml:mo><mml:mn mathvariant="normal">1</mml:mn></mml:mrow></mml:msub><mml:mo>|</mml:mo><mml:mo>/</mml:mo><mml:mo>(</mml:mo><mml:msub><mml:mi>h</mml:mi><mml:mi>i</mml:mi></mml:msub><mml:mo>+</mml:mo><mml:msub><mml:mi>h</mml:mi><mml:mrow><mml:mi>i</mml:mi><mml:mo>+</mml:mo><mml:mn mathvariant="normal">1</mml:mn></mml:mrow></mml:msub><mml:mo>)</mml:mo></mml:mrow></mml:math></inline-formula> satisfies the condition
<inline-formula><mml:math id="M14" display="inline"><mml:mrow><mml:mi>r</mml:mi><mml:mo>&lt;</mml:mo><mml:mn mathvariant="normal">0.25</mml:mn></mml:mrow></mml:math></inline-formula> everywhere (and similarly for <inline-formula><mml:math id="M15" display="inline"><mml:mrow><mml:msub><mml:mi>z</mml:mi><mml:mtext>ice</mml:mtext></mml:msub></mml:mrow></mml:math></inline-formula>). This selective
filter has coefficients scaled by the gradient of <inline-formula><mml:math id="M16" display="inline"><mml:mi>h</mml:mi></mml:math></inline-formula> or <inline-formula><mml:math id="M17" display="inline"><mml:mrow><mml:msub><mml:mi>z</mml:mi><mml:mtext>ice</mml:mtext></mml:msub></mml:mrow></mml:math></inline-formula>,
meaning that regions which are already smooth enough will not become
oversmoothed. Finally, both <inline-formula><mml:math id="M18" display="inline"><mml:mi>h</mml:mi></mml:math></inline-formula> and <inline-formula><mml:math id="M19" display="inline"><mml:mrow><mml:msub><mml:mi>z</mml:mi><mml:mtext>ice</mml:mtext></mml:msub></mml:mrow></mml:math></inline-formula> undergo two final passes
of a regular Hanning smoother to remove 2-D noise. Note that this separate
treatment of bathymetry and ice-shelf draft does not directly consider water
column thickness, for which some large gradients may remain.</p>
      <p id="d1e632">The smoothing procedure in FESOM is the same as described by
<xref ref-type="bibr" rid="bib1.bibx113" id="text.39"/>. The source bathymetry and ice-shelf draft are first
averaged over 4 min (<inline-formula><mml:math id="M20" display="inline"><mml:mrow><mml:mn mathvariant="normal">1</mml:mn><mml:mo>/</mml:mo><mml:mn mathvariant="normal">15</mml:mn></mml:mrow></mml:math></inline-formula><inline-formula><mml:math id="M21" display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula>) intervals. Then Gaussian filters are
applied to both fields, with spatially varying radii scaled by the desired
final resolution. For this reason, high-resolution regions of the domain
receive less smoothing than lower-resolution regions. The ice-shelf draft
undergoes one pass of the Gaussian filter, while the bathymetry undergoes
four passes with larger radii. Following interpolation to the unstructured
mesh, the ice-shelf draft field receives selective smoothing to satisfy the
critical steepness limitation of <xref ref-type="bibr" rid="bib1.bibx53" id="text.40"/> at all points. This
procedure limits the slope of the ice-shelf draft, and extremely high
resolution may be necessary to preserve steep slopes.</p>
      <p id="d1e662">Another region of concern is the grounding line, where water column thickness
approaches zero and vertical layers converge. Estimates of pressure gradient
error, such as that of <xref ref-type="bibr" rid="bib1.bibx53" id="text.41"/>, scale inversely with the vertical
layer thickness and can therefore diverge near the grounding line. To
alleviate this problem, a minimum water column thickness of 50 m is
enforced. In both models, the bathymetry is artificially deepened where
necessary to satisfy this condition during the smoothing process.</p>
</sec>
<sec id="Ch1.S2.SS4">
  <title>Ocean mixing</title>
      <?pagebreak page1261?><p id="d1e674">ROMS includes several options for tracer advection <xref ref-type="bibr" rid="bib1.bibx145" id="paren.42"/>,
and the choice of advection scheme is known to impact the simulation. The
centred and Akima fourth-order tracer advection schemes are dominated by
dispersive error, which can lead to undershoots of the freezing point and
spurious sea-ice formation in our MetROMS configuration
<xref ref-type="bibr" rid="bib1.bibx115" id="paren.43"/>. On the other hand, the upwind third-order tracer
advection scheme is dominated by dissipative error, which can result in high
levels of diapycnal mixing for some simulations
<xref ref-type="bibr" rid="bib1.bibx91 bib1.bibx98" id="paren.44"/>. Indeed, problematic diapycnal mixing
related to the upwind third-order scheme was observed in decadal-scale
simulations with our configuration of MetROMS (not shown). Therefore, the
25-year MetROMS simulation we present here uses the Akima fourth-order tracer
advection scheme, combined with explicitly parameterised Laplacian diffusion
applied along isoneutral surfaces, at a level strong enough to smooth out
most dispersive oscillations. This configuration shows minimal spurious
sea-ice formation, comparable to a simulation with flux-limited (i.e. locally
monotonic) upwind third-order advection <xref ref-type="bibr" rid="bib1.bibx115" id="paren.45"/>, and exhibits
less spurious diapycnal mixing than the upwind scheme. The diffusivity
coefficient is 150 m<inline-formula><mml:math id="M22" display="inline"><mml:msup><mml:mi/><mml:mn mathvariant="normal">2</mml:mn></mml:msup></mml:math></inline-formula> s<inline-formula><mml:math id="M23" display="inline"><mml:msup><mml:mi/><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">1</mml:mn></mml:mrow></mml:msup></mml:math></inline-formula>, which applies to the largest grid cell
(approximately 24 km resolution) and is scaled linearly for smaller cells.
Advection of momentum uses the upwind third-order scheme in the horizontal
and the centred fourth-order scheme in the vertical <xref ref-type="bibr" rid="bib1.bibx145" id="paren.46"/>,
and is combined with parameterised biharmonic viscosity along geopotential
surfaces, with a coefficient of 10<inline-formula><mml:math id="M24" display="inline"><mml:msup><mml:mi/><mml:mn mathvariant="normal">7</mml:mn></mml:msup></mml:math></inline-formula> m<inline-formula><mml:math id="M25" display="inline"><mml:msup><mml:mi/><mml:mn mathvariant="normal">2</mml:mn></mml:msup></mml:math></inline-formula> s<inline-formula><mml:math id="M26" display="inline"><mml:msup><mml:mi/><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">1</mml:mn></mml:mrow></mml:msup></mml:math></inline-formula> (scaled by grid size
as with diffusivity).</p>
      <p id="d1e744">FESOM computes advection of momentum using the characteristic Galerkin
method, and advection of tracers using the explicit second-order
flux-corrected-transport scheme <xref ref-type="bibr" rid="bib1.bibx173" id="paren.47"/>. The Laplacian approach is
used to explicitly parameterise both diffusivity and viscosity, with
coefficients 600 and 6000 m<inline-formula><mml:math id="M27" display="inline"><mml:msup><mml:mi/><mml:mn mathvariant="normal">2</mml:mn></mml:msup></mml:math></inline-formula> s<inline-formula><mml:math id="M28" display="inline"><mml:msup><mml:mi/><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">1</mml:mn></mml:mrow></mml:msup></mml:math></inline-formula>, respectively. These values apply
to a reference area of 5800 km<inline-formula><mml:math id="M29" display="inline"><mml:msup><mml:mi/><mml:mn mathvariant="normal">2</mml:mn></mml:msup></mml:math></inline-formula> and are scaled to the area of each
triangular element, scaling with the square root for diffusivity and linearly
for viscosity. At 10 km resolution (element area of 100 km<inline-formula><mml:math id="M30" display="inline"><mml:msup><mml:mi/><mml:mn mathvariant="normal">2</mml:mn></mml:msup></mml:math></inline-formula>), the
resulting diffusivity is 78.8 m<inline-formula><mml:math id="M31" display="inline"><mml:msup><mml:mi/><mml:mn mathvariant="normal">2</mml:mn></mml:msup></mml:math></inline-formula> s<inline-formula><mml:math id="M32" display="inline"><mml:msup><mml:mi/><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">1</mml:mn></mml:mrow></mml:msup></mml:math></inline-formula>, compared to
62.4 m<inline-formula><mml:math id="M33" display="inline"><mml:msup><mml:mi/><mml:mn mathvariant="normal">2</mml:mn></mml:msup></mml:math></inline-formula> s<inline-formula><mml:math id="M34" display="inline"><mml:msup><mml:mi/><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">1</mml:mn></mml:mrow></mml:msup></mml:math></inline-formula> in ROMS. The analogous viscosity terms cannot be
directly compared between ROMS and FESOM, since they do not use the same
parameterisation.</p>
      <p id="d1e832">A flow-dependent Smagorinsky viscosity term is also applied in FESOM
<xref ref-type="bibr" rid="bib1.bibx148 bib1.bibx149 bib1.bibx173" id="paren.48"/>. In <inline-formula><mml:math id="M35" display="inline"><mml:mi>z</mml:mi></mml:math></inline-formula>-coordinate regions,
tracer diffusion is rotated along isoneutrals, and the Gent–McWilliams eddy
parameterisation is used <xref ref-type="bibr" rid="bib1.bibx36 bib1.bibx37 bib1.bibx173" id="paren.49"/>. In
<inline-formula><mml:math id="M36" display="inline"><mml:mi mathvariant="italic">σ</mml:mi></mml:math></inline-formula>-coordinate regions, diffusivity and viscosity are both applied along
constant-<inline-formula><mml:math id="M37" display="inline"><mml:mi mathvariant="italic">σ</mml:mi></mml:math></inline-formula> surfaces.</p>
      <p id="d1e862">For weakly stratified regions such as the Southern Ocean, the choice of
vertical mixing parameterisation can have a significant impact on simulated
convection <xref ref-type="bibr" rid="bib1.bibx159" id="paren.50"/>. MetROMS employs the
Large–McWilliams–Doney interior closure scheme <xref ref-type="bibr" rid="bib1.bibx89" id="paren.51"/> which
includes the K-profile parameterisation (KPP) boundary layer
parameterisation. We implement the same KPP modification as in
<xref ref-type="bibr" rid="bib1.bibx16" id="text.52"/>, which imposes a minimum surface boundary layer depth
based on surface stress, in the case of stabilising conditions. This
modification is designed to address problems with excessive stratification
during periods of rapid sea-ice melt, and follows principles similar to the
FESOM vertical mixing parameterisation discussed below. A shallow bias in
mixed layer depths during the melt season is problematic for the accurate
simulation of Southern Ocean water masses, particularly in the Weddell Sea
<xref ref-type="bibr" rid="bib1.bibx159" id="paren.53"/>.</p>
      <p id="d1e878">The vertical mixing scheme in our configuration of FESOM
<xref ref-type="bibr" rid="bib1.bibx162" id="paren.54"/> consists of the Richardson number dependent
parameterisation of <xref ref-type="bibr" rid="bib1.bibx122" id="text.55"/>, modified to have maximum vertical
diffusivities and viscosities of 0.05 m<inline-formula><mml:math id="M38" display="inline"><mml:msup><mml:mi/><mml:mn mathvariant="normal">2</mml:mn></mml:msup></mml:math></inline-formula> s<inline-formula><mml:math id="M39" display="inline"><mml:msup><mml:mi/><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">1</mml:mn></mml:mrow></mml:msup></mml:math></inline-formula>. Over a depth defined
by the Monin–Obukhov length, calculated as a function of surface stress and
buoyancy forcing, an extra 0.01 m<inline-formula><mml:math id="M40" display="inline"><mml:msup><mml:mi/><mml:mn mathvariant="normal">2</mml:mn></mml:msup></mml:math></inline-formula> s<inline-formula><mml:math id="M41" display="inline"><mml:msup><mml:mi/><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">1</mml:mn></mml:mrow></mml:msup></mml:math></inline-formula> is applied to both vertical
diffusivity and viscosity. This combination was found by
<xref ref-type="bibr" rid="bib1.bibx159" id="text.56"/> to produce the most realistic representation of water
masses in the Weddell Sea, avoiding the excessive open-ocean convection which
is characteristic of traditional convective adjustment. We also tested the
KPP parameterisation (without the modification used by MetROMS) in short
simulations with our FESOM configuration (not shown). At least on the 5-year
timescale, hydrography in the offshore Weddell Sea was very similar between
KPP and the modified Pacanowski–Philander scheme. It is possible that longer
simulations would show more divergence, and this warrants further
investigation.</p>
</sec>
<sec id="Ch1.S2.SS5">
  <title>Ice-shelf thermodynamics</title>
      <p id="d1e939">With terrain-following coordinates, it is relatively straightforward to
include ice-shelf cavities in an Antarctic domain. In both ROMS and FESOM,
all of the terrain-following vertical layers subduct beneath the ice shelves.
The pressure of the ice-shelf draft must be considered in the calculation of
the pressure gradient. ROMS vertically integrates the density of water
displaced by ice, and assumes the density of this displaced water is a linear
function of depth, with coefficient <inline-formula><mml:math id="M42" display="inline"><mml:mrow><mml:mstyle displaystyle="false"><mml:mfrac style="text"><mml:mrow><mml:mo>∂</mml:mo><mml:mi mathvariant="italic">ρ</mml:mi></mml:mrow><mml:mrow><mml:mo>∂</mml:mo><mml:mi>z</mml:mi></mml:mrow></mml:mfrac></mml:mstyle><mml:mo>=</mml:mo><mml:mn mathvariant="normal">4.78</mml:mn><mml:mo>×</mml:mo><mml:msup><mml:mn mathvariant="normal">10</mml:mn><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">3</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula> kg m<inline-formula><mml:math id="M43" display="inline"><mml:msup><mml:mi/><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">4</mml:mn></mml:mrow></mml:msup></mml:math></inline-formula> and intercept given by the density in the first
model layer. FESOM computes the pressure gradient force from the vertically
integrated horizontal density gradient and assumes that the horizontal
pressure gradient is zero at the ice-shelf base. High-order interpolation for
density is done in the vertical to compute horizontal density gradients as
accurately as possible.</p>
      <p id="d1e986">ROMS and FESOM simulate ice-shelf thermodynamics: the heat and salt fluxes
associated with melting and refreezing at the ice-shelf base. However, any
net melting or freezing is not actually applied to the ice-shelf geometry. It
is assumed that glacial flow of the ice shelf, surface accumulation, and
basal melting are in dynamic equilibrium such that the geometry remains
constant. Removing this assumption necessitates coupling with an ice-sheet
model, which has recently been accomplished for FESOM <xref ref-type="bibr" rid="bib1.bibx160" id="paren.57"/>
and is under development for ROMS <xref ref-type="bibr" rid="bib1.bibx38" id="paren.58"/>. Ice-sheet/ocean
coupling is an emerging field of climate modelling, and the first generation
of models will be compared and evaluated as part of the MISOMIP experiments
(Marine Ice Sheet-Ocean Model Intercomparison Project)
<xref ref-type="bibr" rid="bib1.bibx2" id="paren.59"/>.</p>
      <?pagebreak page1262?><p id="d1e998">Both ROMS and FESOM <xref ref-type="bibr" rid="bib1.bibx33 bib1.bibx35 bib1.bibx164" id="paren.60"/>
implement the three-equation parameterisation of <xref ref-type="bibr" rid="bib1.bibx55" id="text.61"/> refined by
<xref ref-type="bibr" rid="bib1.bibx62" id="text.62"/>. The heat and salt exchange coefficients <inline-formula><mml:math id="M44" display="inline"><mml:mrow><mml:msub><mml:mi mathvariant="italic">γ</mml:mi><mml:mi>T</mml:mi></mml:msub></mml:mrow></mml:math></inline-formula> and
<inline-formula><mml:math id="M45" display="inline"><mml:mrow><mml:msub><mml:mi mathvariant="italic">γ</mml:mi><mml:mi>S</mml:mi></mml:msub></mml:mrow></mml:math></inline-formula> have the form
            <disp-formula id="Ch1.E1" content-type="numbered"><mml:math id="M46" display="block"><mml:mrow><mml:msub><mml:mi mathvariant="italic">γ</mml:mi><mml:mi>T</mml:mi></mml:msub><mml:mo>=</mml:mo><mml:mstyle displaystyle="true"><mml:mfrac style="display"><mml:mrow><mml:msup><mml:mi>u</mml:mi><mml:mo>*</mml:mo></mml:msup></mml:mrow><mml:mrow><mml:mi mathvariant="italic">κ</mml:mi><mml:mo>+</mml:mo><mml:mn mathvariant="normal">12.5</mml:mn><mml:mspace linebreak="nobreak" width="0.33em"/><mml:msup><mml:mi mathvariant="italic">Pr</mml:mi><mml:mfrac><mml:mn mathvariant="normal">2</mml:mn><mml:mn mathvariant="normal">3</mml:mn></mml:mfrac></mml:msup><mml:mo>-</mml:mo><mml:mn mathvariant="normal">6</mml:mn></mml:mrow></mml:mfrac></mml:mstyle><mml:mo>,</mml:mo><mml:mtext> and</mml:mtext><mml:mspace linebreak="nobreak" width="0.25em"/><mml:msub><mml:mi mathvariant="italic">γ</mml:mi><mml:mi>S</mml:mi></mml:msub><mml:mo>=</mml:mo><mml:mstyle displaystyle="true"><mml:mfrac style="display"><mml:mrow><mml:msup><mml:mi>u</mml:mi><mml:mo>*</mml:mo></mml:msup></mml:mrow><mml:mrow><mml:mi mathvariant="italic">κ</mml:mi><mml:mo>+</mml:mo><mml:mn mathvariant="normal">12.5</mml:mn><mml:mspace linebreak="nobreak" width="0.33em"/><mml:msup><mml:mi mathvariant="italic">Sc</mml:mi><mml:mfrac><mml:mn mathvariant="normal">2</mml:mn><mml:mn mathvariant="normal">3</mml:mn></mml:mfrac></mml:msup><mml:mo>-</mml:mo><mml:mn mathvariant="normal">6</mml:mn></mml:mrow></mml:mfrac></mml:mstyle><mml:mo>,</mml:mo></mml:mrow></mml:math></disp-formula>
          where <italic>Pr</italic> is the Prandtl number and <italic>Sc</italic> is the Schmidt number
(both dimensionless constants), and <inline-formula><mml:math id="M47" display="inline"><mml:mrow><mml:msup><mml:mi>u</mml:mi><mml:mo>*</mml:mo></mml:msup></mml:mrow></mml:math></inline-formula> is the friction velocity in
m s<inline-formula><mml:math id="M48" display="inline"><mml:msup><mml:mi/><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">1</mml:mn></mml:mrow></mml:msup></mml:math></inline-formula>, calculated as
            <disp-formula id="Ch1.E2" content-type="numbered"><mml:math id="M49" display="block"><mml:mrow><mml:msup><mml:mi>u</mml:mi><mml:mo>*</mml:mo></mml:msup><mml:mo>=</mml:mo><mml:mo movablelimits="false">max⁡</mml:mo><mml:mfenced close=")" open="("><mml:mrow><mml:msqrt><mml:mrow><mml:msub><mml:mi>C</mml:mi><mml:mtext>d</mml:mtext></mml:msub><mml:mfenced close=")" open="("><mml:mrow><mml:msup><mml:mi>u</mml:mi><mml:mn mathvariant="normal">2</mml:mn></mml:msup><mml:mo>+</mml:mo><mml:msup><mml:mi>v</mml:mi><mml:mn mathvariant="normal">2</mml:mn></mml:msup></mml:mrow></mml:mfenced></mml:mrow></mml:msqrt><mml:mo>,</mml:mo><mml:msubsup><mml:mi>u</mml:mi><mml:mtext>min</mml:mtext><mml:mo>*</mml:mo></mml:msubsup></mml:mrow></mml:mfenced><mml:mo>,</mml:mo></mml:mrow></mml:math></disp-formula>
          where <inline-formula><mml:math id="M50" display="inline"><mml:mrow><mml:msub><mml:mi>C</mml:mi><mml:mtext>d</mml:mtext></mml:msub></mml:mrow></mml:math></inline-formula> is the drag coefficient (<inline-formula><mml:math id="M51" display="inline"><mml:mrow><mml:mn mathvariant="normal">3</mml:mn><mml:mo>×</mml:mo><mml:msup><mml:mn mathvariant="normal">10</mml:mn><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">3</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula> in ROMS, <inline-formula><mml:math id="M52" display="inline"><mml:mrow><mml:mn mathvariant="normal">2.5</mml:mn><mml:mo>×</mml:mo><mml:msup><mml:mn mathvariant="normal">10</mml:mn><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">3</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula> in FESOM), <inline-formula><mml:math id="M53" display="inline"><mml:mi>u</mml:mi></mml:math></inline-formula> and <inline-formula><mml:math id="M54" display="inline"><mml:mi>v</mml:mi></mml:math></inline-formula> are the horizontal ocean velocity
components in the uppermost vertical layer, and <inline-formula><mml:math id="M55" display="inline"><mml:mrow><mml:msubsup><mml:mi>u</mml:mi><mml:mtext>min</mml:mtext><mml:mo>*</mml:mo></mml:msubsup></mml:mrow></mml:math></inline-formula> is a lower
bound for <inline-formula><mml:math id="M56" display="inline"><mml:mrow><mml:msup><mml:mi>u</mml:mi><mml:mo>*</mml:mo></mml:msup></mml:mrow></mml:math></inline-formula> which represents molecular diffusion (<inline-formula><mml:math id="M57" display="inline"><mml:mrow><mml:msup><mml:mn mathvariant="normal">10</mml:mn><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">3</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula> in ROMS, <inline-formula><mml:math id="M58" display="inline"><mml:mrow><mml:mn mathvariant="normal">2.5</mml:mn><mml:mo>×</mml:mo><mml:msup><mml:mn mathvariant="normal">10</mml:mn><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">4</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula> in FESOM). While the effect of the different drag coefficient
between the models is likely to be negligible, the larger minimum <inline-formula><mml:math id="M59" display="inline"><mml:mrow><mml:msup><mml:mi>u</mml:mi><mml:mo>*</mml:mo></mml:msup></mml:mrow></mml:math></inline-formula> in
ROMS will cause stronger melting in locations with very weak flow, such as at
the grounding line <xref ref-type="bibr" rid="bib1.bibx50" id="paren.63"/>.</p>
      <p id="d1e1327">The turbulence term <inline-formula><mml:math id="M60" display="inline"><mml:mi mathvariant="italic">κ</mml:mi></mml:math></inline-formula> in Eq. (<xref ref-type="disp-formula" rid="Ch1.E1"/>) has a different
formulation between the two models. FESOM follows a very similar approach to
<xref ref-type="bibr" rid="bib1.bibx73" id="text.64"/> by which
            <disp-formula id="Ch1.E3" content-type="numbered"><mml:math id="M61" display="block"><mml:mrow><mml:mi mathvariant="italic">κ</mml:mi><mml:mo>=</mml:mo><mml:mn mathvariant="normal">2.12</mml:mn><mml:mspace linebreak="nobreak" width="0.33em"/><mml:mi>log⁡</mml:mi><mml:mfenced open="(" close=")"><mml:mrow><mml:msup><mml:mi>u</mml:mi><mml:mo>*</mml:mo></mml:msup><mml:mstyle displaystyle="true"><mml:mfrac style="display"><mml:mi>D</mml:mi><mml:mi mathvariant="italic">ν</mml:mi></mml:mfrac></mml:mstyle></mml:mrow></mml:mfenced><mml:mo>-</mml:mo><mml:mn mathvariant="normal">3</mml:mn><mml:mo>,</mml:mo></mml:mrow></mml:math></disp-formula>
          where <inline-formula><mml:math id="M62" display="inline"><mml:mrow><mml:mi>D</mml:mi><mml:mo>=</mml:mo><mml:mn mathvariant="normal">10</mml:mn></mml:mrow></mml:math></inline-formula> m is a reference boundary layer depth, and <inline-formula><mml:math id="M63" display="inline"><mml:mrow><mml:mi mathvariant="italic">ν</mml:mi><mml:mo>=</mml:mo><mml:mn mathvariant="normal">1.95</mml:mn><mml:mo>×</mml:mo><mml:msup><mml:mn mathvariant="normal">10</mml:mn><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">6</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula> m<inline-formula><mml:math id="M64" display="inline"><mml:msup><mml:mi/><mml:mn mathvariant="normal">2</mml:mn></mml:msup></mml:math></inline-formula> s<inline-formula><mml:math id="M65" display="inline"><mml:msup><mml:mi/><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">1</mml:mn></mml:mrow></mml:msup></mml:math></inline-formula> is the kinematic viscosity. ROMS instead uses a
simplified version of <xref ref-type="bibr" rid="bib1.bibx103" id="text.65"/>'s approach by which
            <disp-formula id="Ch1.E4" content-type="numbered"><mml:math id="M66" display="block"><mml:mrow><?xmltex \hack{\hbox\bgroup\fontsize{9.5}{9.5}\selectfont$\displaystyle}?><mml:mi mathvariant="italic">κ</mml:mi><mml:mo>=</mml:mo><mml:mfenced open="{" close=""><mml:mtable columnspacing="1em" class="cases" rowspacing="0.2ex 0.2ex 5.690551pt" columnalign="left left" framespacing="0em"><mml:mtr><mml:mtd><mml:mrow><mml:mn mathvariant="normal">2.5</mml:mn><mml:mspace width="0.33em" linebreak="nobreak"/><mml:mi>log⁡</mml:mi><mml:mfenced open="(" close=")"><mml:mstyle displaystyle="false"><mml:mstyle displaystyle="false"><mml:mfrac style="text"><mml:mrow><mml:mfenced close=")" open="("><mml:mrow><mml:mn mathvariant="normal">5300</mml:mn><mml:mspace width="0.125em" linebreak="nobreak"/><mml:msup><mml:mtext>m</mml:mtext><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">2</mml:mn></mml:mrow></mml:msup><mml:mspace linebreak="nobreak" width="0.125em"/><mml:mtext>s</mml:mtext></mml:mrow></mml:mfenced><mml:mo>(</mml:mo><mml:msup><mml:mi>u</mml:mi><mml:mo>*</mml:mo></mml:msup><mml:msup><mml:mo>)</mml:mo><mml:mn mathvariant="normal">2</mml:mn></mml:msup></mml:mrow><mml:mrow><mml:mo>|</mml:mo><mml:mi>f</mml:mi><mml:mo>|</mml:mo></mml:mrow></mml:mfrac></mml:mstyle></mml:mstyle></mml:mfenced><mml:mo>+</mml:mo><mml:mn mathvariant="normal">7.12</mml:mn></mml:mrow></mml:mtd><mml:mtd><mml:mrow><mml:mspace linebreak="nobreak" width="0.25em"/><mml:mtext>if</mml:mtext><mml:mspace width="0.25em" linebreak="nobreak"/><mml:msup><mml:mi>u</mml:mi><mml:mo>*</mml:mo></mml:msup><mml:mo>&gt;</mml:mo><mml:msup><mml:mn mathvariant="normal">10</mml:mn><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">3</mml:mn></mml:mrow></mml:msup><mml:mspace linebreak="nobreak" width="0.125em"/><mml:msup><mml:mtext>m s</mml:mtext><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">1</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:mtd></mml:mtr><mml:mtr><mml:mtd><mml:mrow/></mml:mtd><mml:mtd><mml:mrow><mml:mspace width="0.25em" linebreak="nobreak"/><mml:mtext>and</mml:mtext></mml:mrow></mml:mtd></mml:mtr><mml:mtr><mml:mtd><mml:mrow/></mml:mtd><mml:mtd><mml:mrow><mml:mspace linebreak="nobreak" width="0.25em"/><mml:mo>|</mml:mo><mml:mi>f</mml:mi><mml:mo>|</mml:mo><mml:mo>&gt;</mml:mo><mml:msup><mml:mn mathvariant="normal">10</mml:mn><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">8</mml:mn></mml:mrow></mml:msup><mml:mspace width="0.125em" linebreak="nobreak"/><mml:msup><mml:mtext>s</mml:mtext><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">1</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:mtd></mml:mtr><mml:mtr><mml:mtd><mml:mn mathvariant="normal">0</mml:mn></mml:mtd><mml:mtd><mml:mrow><mml:mspace width="0.25em" linebreak="nobreak"/><mml:mtext>otherwise</mml:mtext></mml:mrow></mml:mtd></mml:mtr></mml:mtable></mml:mfenced><?xmltex \hack{$\egroup}?><mml:mo>,</mml:mo></mml:mrow></mml:math></disp-formula>
          where <inline-formula><mml:math id="M67" display="inline"><mml:mi>f</mml:mi></mml:math></inline-formula> is the Coriolis parameter in s<inline-formula><mml:math id="M68" display="inline"><mml:msup><mml:mi/><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">1</mml:mn></mml:mrow></mml:msup></mml:math></inline-formula>.</p>
      <p id="d1e1602">While refreezing is implicit in the three-equation formulation, none of our
configurations include an explicit frazil ice model such as that of
<xref ref-type="bibr" rid="bib1.bibx150" id="text.66"/> or <xref ref-type="bibr" rid="bib1.bibx35" id="text.67"/>.</p>
</sec>
<sec id="Ch1.S2.SS6">
  <title>Sea ice</title>
      <p id="d1e1617">MetROMS includes the sea-ice model CICE <xref ref-type="bibr" rid="bib1.bibx66" id="paren.68"/> which is a
multi-layer, multi-category model widely used in global coupled models as
well as regional and uncoupled setups. Our configuration of CICE has seven
ice layers plus one snow layer, and five ice thickness categories. It is
externally coupled to ROMS, i.e. runs on separate processors, with
communication driven by the coupler MCT <xref ref-type="bibr" rid="bib1.bibx90 bib1.bibx68" id="paren.69"/>. There
are six baroclinic ocean time steps (5 min) for each sea-ice time step
(30 min), and the coupler exchanges fields every sea-ice time step. Having
longer time steps for the sea ice than for the ocean is computationally
favourable, but it also introduces lags in ocean/sea-ice interactions,
because the coupled fields are time-averaged over the previous 30 min.</p>
      <p id="d1e1626">FESOM's sea-ice model is described by <xref ref-type="bibr" rid="bib1.bibx11" id="text.70"/>. It has a single
ice layer (plus one snow layer) and a single thickness category. It is
internally coupled with the ocean, running on the same processors and the
same time step. While the FESOM sea-ice model is generally less complex than
CICE, it nonetheless has been shown to reproduce key features of observed
Arctic and Antarctic sea ice <xref ref-type="bibr" rid="bib1.bibx162" id="paren.71"/>.</p>
      <p id="d1e1635">Our configuration of CICE uses the “mushy” thermodynamics scheme of
<xref ref-type="bibr" rid="bib1.bibx167" id="text.72"/>. It also includes the level-ice melt pond parameterisation
of <xref ref-type="bibr" rid="bib1.bibx65" id="text.73"/>, and the Delta–Eddington radiation scheme
<xref ref-type="bibr" rid="bib1.bibx6" id="paren.74"/>. In FESOM, sea-ice thermodynamics follows
<xref ref-type="bibr" rid="bib1.bibx125" id="text.75"/> with the zero-layer approach to heat conduction
<xref ref-type="bibr" rid="bib1.bibx143" id="paren.76"/>.</p>
      <p id="d1e1653">For sea-ice dynamics, CICE uses elastic–anisotropic–plastic rheology
<xref ref-type="bibr" rid="bib1.bibx166" id="paren.77"/> with the ridging-based ice strength formulation of
<xref ref-type="bibr" rid="bib1.bibx139" id="text.78"/>. FESOM has elastic–viscous–plastic rheology
<xref ref-type="bibr" rid="bib1.bibx5" id="paren.79"/> including a linear formulation of ice strength with
coefficient <inline-formula><mml:math id="M69" display="inline"><mml:mrow><mml:msup><mml:mi>P</mml:mi><mml:mo>*</mml:mo></mml:msup><mml:mo>=</mml:mo><mml:mn mathvariant="normal">15</mml:mn><mml:mspace linebreak="nobreak" width="0.125em"/><mml:mn mathvariant="normal">000</mml:mn></mml:mrow></mml:math></inline-formula> N m<inline-formula><mml:math id="M70" display="inline"><mml:msup><mml:mi/><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">2</mml:mn></mml:mrow></mml:msup></mml:math></inline-formula>. Sea-ice transport follows an
incremental remapping approach in CICE <xref ref-type="bibr" rid="bib1.bibx94" id="paren.80"/>, with the ridging
participation and redistribution functions of <xref ref-type="bibr" rid="bib1.bibx95" id="text.81"/>. FESOM
uses a backward Euler implicit advection scheme for sea-ice transport.</p>
</sec>
<sec id="Ch1.S2.SS7">
  <title>Surface exchange scheme</title>
      <p id="d1e1708">While MetROMS and FESOM are forced with the same atmospheric state (see
Sect. <xref ref-type="sec" rid="Ch1.S3.SS2"/>), the resulting surface fluxes differ based on
the bulk formulae implemented by the models. Our configuration of FESOM uses
constant exchange coefficients for heat and momentum fluxes, while MetROMS'
exchange coefficients vary in time and space. For ocean/atmosphere fluxes (in
ROMS), these coefficients are based on the COARE (Coupled Ocean-Atmosphere
Response Experiment) protocol <xref ref-type="bibr" rid="bib1.bibx25" id="paren.82"/>. For sea-ice/atmosphere
fluxes, CICE includes a stability-based atmospheric boundary interface
<xref ref-type="bibr" rid="bib1.bibx66" id="paren.83"/>. These differences in bulk formulae may affect the
simulations, particularly the momentum fluxes which have consequences for ACC
transport, Ekman pumping, and sea-ice formation and drift. A comparison of
ocean surface stress (not shown) reveals that these momentum fluxes are
typically stronger in MetROMS by up to 30 %.</p>
</sec>
</sec>
<sec id="Ch1.S3">
  <title>Experimental design</title>
      <p id="d1e1726">For this intercomparison, we simulated the 25-year period of 1992–2016 using
three model configurations: MetROMS, low-resolution FESOM, and
high-resolution FESOM (Fig. <xref ref-type="fig" rid="Ch1.F1"/>).</p>
<sec id="Ch1.S3.SS1">
  <title>Initial conditions</title>
      <?pagebreak page1263?><p id="d1e1736">All simulations are initialised using monthly averaged observational or
reanalysis products for January 1992. Initial ocean temperature and salinity
are taken from the ECCO2 reanalysis <xref ref-type="bibr" rid="bib1.bibx106 bib1.bibx177" id="paren.84"/>, and
extrapolated into ice-shelf cavities using a nearest-neighbour method in
Cartesian space. Initial ocean velocity and sea surface height are set to
zero.
<?xmltex \hack{\newpage}?>
Sea ice is initialised using the National Oceanic and Atmospheric Administration
(NOAA)/National Snow and Ice Data Center (NSIDC) Climate Data Record for Passive
Microwave Sea Ice Concentration <xref ref-type="bibr" rid="bib1.bibx105" id="paren.85"/>. Wherever the observed
Antarctic sea-ice concentration exceeds 0.15, the model is initialised with
concentration 1, ice thickness of 1 m, and snow thickness of 0.2 m. This is
the same method used by <xref ref-type="bibr" rid="bib1.bibx115" id="text.86"/> and is similar to that of
<xref ref-type="bibr" rid="bib1.bibx81" id="text.87"/>. FESOM, having a global domain, also requires initial
conditions for Arctic sea ice. We follow the same method as for the
Antarctic but set the initial ice thickness to 2 m, since Arctic sea ice
tends to be thicker <xref ref-type="bibr" rid="bib1.bibx86 bib1.bibx176" id="paren.88"/>. Initial sea-ice velocity is
set to zero.</p>
      <p id="d1e1757">Our experiments do not include a proper spinup to a quasi-equilibrium state.
For the purposes of this intercomparison around the Antarctic margin and
continental shelf, as well as in the ice-shelf cavities, we argue a full
spinup is not worth the computational expense. The processes we focus on –
onshore flow, dense shelf water formation, and ocean/ice-shelf interaction –
equilibrate much more quickly than the interior ocean. For example,
area-averaged basal melt rates in our experiments stabilise within
5–10 years for most ice shelves.</p>
</sec>
<sec id="Ch1.S3.SS2">
  <title>Atmospheric forcing</title>
      <p id="d1e1766">MetROMS and FESOM are both forced with the ERA-Interim atmospheric reanalysis
<xref ref-type="bibr" rid="bib1.bibx14" id="paren.89"/> using 6- and 12-hourly fields over the years 1992–2016. Due
to differing implementations of model thermodynamics, the two models are
forced with different combinations of atmospheric variables. Both models
utilise 6-hourly fields for near-surface air temperature, pressure, and
winds, which are linearly interpolated to each time step. Near-surface
humidity is derived from ERA-Interim's 6-hourly fields for dew-point
temperature; this conversion is performed in advance for MetROMS but at
runtime for FESOM. Both models read 12-hourly fields for precipitation
(split into rain and snow) and evaporation, which are not interpolated in
time but rather applied at a constant rate with a step change every 12 h,
as they represent total fluxes over the given 12 h period.</p>
      <p id="d1e1772">MetROMS diagnoses incoming shortwave radiation from ERA-Interim's 6-hourly
total cloud cover, which is interpolated to each time step. Incoming longwave
radiation is calculated internally. In FESOM, incoming shortwave and longwave
radiation are read directly from ERA-Interim, as 12-hourly fields which are
applied as step changes.</p>
      <p id="d1e1775">To account for the influence of iceberg calving on the Southern Ocean
freshwater budget, both models are forced with an additional surface
freshwater flux representing iceberg melt. For this field, we use the output
of <xref ref-type="bibr" rid="bib1.bibx99" id="text.90"/>, who modelled icebergs as interactive Lagrangian
particles in the ocean component of a general
circulation model (GCM) simulation. The initial sizes of
icebergs at calving fronts were determined from a statistical distribution
constrained by observations. <xref ref-type="bibr" rid="bib1.bibx99" id="text.91"/>'s monthly climatology of
iceberg melt is interpolated to each time step in our simulations and
repeated annually. River runoff from other continents is not considered.</p>
</sec>
<sec id="Ch1.S3.SS3">
  <title>Surface salinity restoring</title>
      <p id="d1e1790">A persistent feature of many Southern Ocean models
<xref ref-type="bibr" rid="bib1.bibx81 bib1.bibx60 bib1.bibx141 bib1.bibx168 bib1.bibx41" id="paren.92"/> is
spuriously deep convection in the Weddell Sea, leading to an unrealistic
open-ocean polynya as warm Circumpolar Deep Water is brought to the surface.
The possible causes of this widespread model bias include insufficient
surface freshwater flux <xref ref-type="bibr" rid="bib1.bibx81" id="paren.93"/> as well as insufficient summer
mixed layer depths <xref ref-type="bibr" rid="bib1.bibx159" id="paren.94"/>. In both circumstances, salinity in
the subsurface Winter Water layer increases until the weakly stratified water
column becomes unstable and overturns.</p>
      <p id="d1e1802">MetROMS is prone to deep convection in the Weddell Sea, and while tuning of
the sea-ice dynamics and ocean vertical mixing helped to delay the onset of
convection, the only permanent solution we found was surface salinity
restoring. Such restoring affects the salt budget and may contribute to drift
in the total salt content of the ocean, although it prevents drift in the
surface layer. This may impact the density structure of the Southern Ocean,
and particularly the ACC, as well as damping interannual variability.
However, these shortcomings were deemed preferable to spurious deep
convection for the purposes of our analysis. We restore MetROMS to the World
Ocean Atlas 2013 monthly climatology of surface salinity <xref ref-type="bibr" rid="bib1.bibx179" id="paren.95"/>,
linearly interpolated to each model time step and repeated annually.
Restoring has a timescale of 30 days and affects the uppermost layer, whose
thickness is time-varying but generally ranges from 1 to 3 m. We exclude the
Antarctic continental shelf from this restoring (defined as regions south of
<inline-formula><mml:math id="M71" display="inline"><mml:mn mathvariant="normal">60</mml:mn></mml:math></inline-formula><inline-formula><mml:math id="M72" display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula> S with bathymetry shallower than 1500 m, as well as all ice-shelf cavities), as significant freshening of Antarctic Bottom Water occurs
otherwise. Given the relatively scant observations on the continental shelf
making up the World Ocean Atlas products, restoring in this region may not be
appropriate.</p>
      <?pagebreak page1264?><p id="d1e1823">FESOM does not develop spurious deep convection in the Weddell Sea, even for
long simulations without restoring. Possible reasons for this differing
behaviour between MetROMS and FESOM are discussed in
Sect. <xref ref-type="sec" rid="Ch1.S4.SS2.SSS1"/>. Nonetheless, we apply the same surface salinity
restoring to FESOM as we do to MetROMS, so that the experiments are as
similar as possible. Restoring in FESOM is scaled with a constant depth of
10 m, which is the depth of the surface layer in <inline-formula><mml:math id="M73" display="inline"><mml:mi>z</mml:mi></mml:math></inline-formula>-coordinate regions,
neglecting free surface variations. We do not restore north of
<inline-formula><mml:math id="M74" display="inline"><mml:mn mathvariant="normal">30</mml:mn></mml:math></inline-formula><inline-formula><mml:math id="M75" display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula> S, as this region is outside the MetROMS domain.
<?xmltex \hack{\newpage}?></p>
</sec>
<sec id="Ch1.S3.SS4">
  <title>Northern boundary conditions</title>
      <p id="d1e1858">MetROMS, with its regional circumpolar domain, has lateral boundary
conditions at <inline-formula><mml:math id="M76" display="inline"><mml:mn mathvariant="normal">30</mml:mn></mml:math></inline-formula><inline-formula><mml:math id="M77" display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula> S. The ECCO2 reanalysis
<xref ref-type="bibr" rid="bib1.bibx106 bib1.bibx177" id="paren.96"/> provides temperature, salinity, and
meridional velocity (<inline-formula><mml:math id="M78" display="inline"><mml:mi>v</mml:mi></mml:math></inline-formula>) as monthly averages over the transient period
of 1992–2016. Sea surface height is taken from the AVISO annual mean
climatology <xref ref-type="bibr" rid="bib1.bibx3" id="paren.97"/> which is a single time record. Note that tides
are not considered, as discussed further in Sect. <xref ref-type="sec" rid="Ch1.S5"/>.</p>
      <p id="d1e1891">We follow the method described in <xref ref-type="bibr" rid="bib1.bibx115" id="text.98"/> to ensure stability
at the open boundary: zonal velocity <inline-formula><mml:math id="M79" display="inline"><mml:mi>u</mml:mi></mml:math></inline-formula> is clamped to zero, the bathymetry
is modified to be constant in latitude over the northernmost 15 rows of the
domain, and a sponge layer is applied over these rows (in which the
diffusivity coefficient linearly increases to 10 times its background value,
and the viscosity coefficient to 100 times). Northern boundary conditions are
applied using the Chapman scheme for sea surface height <xref ref-type="bibr" rid="bib1.bibx7" id="paren.99"/>,
the Flather scheme for barotropic <inline-formula><mml:math id="M80" display="inline"><mml:mi>v</mml:mi></mml:math></inline-formula> <xref ref-type="bibr" rid="bib1.bibx27" id="paren.100"/>, and the
radiation-nudging scheme for baroclinic <inline-formula><mml:math id="M81" display="inline"><mml:mi>v</mml:mi></mml:math></inline-formula>, temperature, and salinity
<xref ref-type="bibr" rid="bib1.bibx97" id="paren.101"/>.</p>
      <p id="d1e1928">The presence of lateral boundary conditions derived from observations may
give MetROMS an advantage for the accurate simulation of Southern Ocean water
masses, compared to FESOM which has a global domain. However, this
intercomparison focuses on the continental shelf and ice-shelf cavities.
These regions are relatively far field from <inline-formula><mml:math id="M82" display="inline"><mml:mn mathvariant="normal">30</mml:mn></mml:math></inline-formula><inline-formula><mml:math id="M83" display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula> S, compared to the
ACC and the interior Southern Ocean which are more tightly coupled to the
boundary conditions. For the relatively short (25-year) simulations shown
here, it is unlikely that continental shelf water masses will be
significantly influenced by nudging at <inline-formula><mml:math id="M84" display="inline"><mml:mn mathvariant="normal">30</mml:mn></mml:math></inline-formula><inline-formula><mml:math id="M85" display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula> S. Longer simulations
would likely show a larger response.</p>
</sec>
</sec>
<sec id="Ch1.S4">
  <title>Results</title>
<sec id="Ch1.S4.SS1">
  <title>Ocean</title>
<sec id="Ch1.S4.SS1.SSS1">
  <title>Drake Passage transport</title>
      <p id="d1e1978">The ACC has the strongest transport of any ocean current in the world and is
key to the thermal isolation of Antarctica. Transport of the ACC is
influenced by the Southern Hemisphere westerly winds as well as the density
structure of the Southern Ocean. By convention, zonal transport of the ACC is
evaluated through Drake Passage and is time-averaged to remove the seasonal
cycle. With respect to observations, Drake Passage transport was previously
thought to lie around 134 Sv <xref ref-type="bibr" rid="bib1.bibx10" id="paren.102"/>. However, recent
improvements in measuring systems have suggested a higher value. As part of
the cDrake project (Dynamics and Transport of the Antarctic Circumpolar
Current in Drake Passage), <xref ref-type="bibr" rid="bib1.bibx18" id="text.103"/> estimated a Drake Passage
transport of <inline-formula><mml:math id="M86" display="inline"><mml:mrow><mml:mn mathvariant="normal">173.3</mml:mn><mml:mo>±</mml:mo><mml:mn mathvariant="normal">10.7</mml:mn></mml:mrow></mml:math></inline-formula> Sv.
<?xmltex \hack{\newpage}?>
For our simulations, zonal transport through Drake Passage is calculated at
<inline-formula><mml:math id="M87" display="inline"><mml:mn mathvariant="normal">67</mml:mn></mml:math></inline-formula><inline-formula><mml:math id="M88" display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula> W over the period 2002–2016. The first 10 years of the
simulation (1992–2001) are excluded as spinup. This time-averaged Drake
Passage transport, including the standard deviation in annual averages, is
126.8 <inline-formula><mml:math id="M89" display="inline"><mml:mo>±</mml:mo></mml:math></inline-formula> 3.4 Sv in MetROMS, 158.6 <inline-formula><mml:math id="M90" display="inline"><mml:mo>±</mml:mo></mml:math></inline-formula> 2.8 Sv in low-resolution FESOM,
and 152.6 <inline-formula><mml:math id="M91" display="inline"><mml:mo>±</mml:mo></mml:math></inline-formula> 3.1 Sv in high-resolution FESOM. Compared to the
observations of <xref ref-type="bibr" rid="bib1.bibx18" id="text.104"/>, the values from all three of our
simulations are too low, especially in MetROMS. This occurs despite MetROMS'
stronger surface stress than in FESOM (Sect. <xref ref-type="sec" rid="Ch1.S2.SS7"/>).
Additionally, the MetROMS and low-resolution FESOM simulations exhibit
downward trends in Drake Passage transport over 2002–2016, which are
statistically significant at the 95 % level: <inline-formula><mml:math id="M92" display="inline"><mml:mo>-</mml:mo></mml:math></inline-formula>0.28 Sv yr<inline-formula><mml:math id="M93" display="inline"><mml:msup><mml:mi/><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">1</mml:mn></mml:mrow></mml:msup></mml:math></inline-formula> in
MetROMS and <inline-formula><mml:math id="M94" display="inline"><mml:mo>-</mml:mo></mml:math></inline-formula>0.17 Sv yr<inline-formula><mml:math id="M95" display="inline"><mml:msup><mml:mi/><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">1</mml:mn></mml:mrow></mml:msup></mml:math></inline-formula> in low-resolution FESOM. This weakening of
the ACC may be driven by degradation of Southern Ocean interior water masses
due to spurious diapycnal mixing, as discussed in Sect. <xref ref-type="sec" rid="Ch1.S4.SS1.SSS4"/>.
Furthermore, drift in the density structure may result from non-closure of
the surface freshwater budget, which is globally unconstrained by the
bulk-flux approach of our simulations. Since interior Southern Ocean
processes operate on much longer timescales than our experiments, and would
require long spinups to equilibrate, simulated ACC transport should be
interpreted with caution and is not the focus of this paper.</p>

      <?xmltex \floatpos{t}?><fig id="Ch1.F2" specific-use="star"><label>Figure 2</label><caption><p id="d1e2086">Mixed layer depth (m), calculated as the shallowest depth where
potential density is at least 0.03 kg m<inline-formula><mml:math id="M96" display="inline"><mml:msup><mml:mi/><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">3</mml:mn></mml:mrow></mml:msup></mml:math></inline-formula> greater than at the surface
or ice-shelf base. Results are shown for MetROMS, low-resolution FESOM, and
high-resolution FESOM averaged over the years 2002–2016 for summer (DJF) and
winter (JJA), as well as climatological observations by
<xref ref-type="bibr" rid="bib1.bibx126" id="text.105"/> recalculated with the same definition of mixed layer
depth. Note the different colour scale for summer and winter.</p></caption>
            <?xmltex \igopts{width=341.433071pt}?><graphic xlink:href="https://gmd.copernicus.org/articles/11/1257/2018/gmd-11-1257-2018-f02.png"/>

          </fig>

</sec>
<sec id="Ch1.S4.SS1.SSS2">
  <title>Mixed layer depth</title>
      <p id="d1e2116">The surface mixed layer represents the portion of the ocean which is directly
influenced by the atmosphere. The depth of the mixed layer is a key indicator
of the strength of convection, and heat loss to the atmosphere resulting from
convection will influence water mass properties. Regions of strong sea-ice
formation, such as coastal polynyas, are characterised by deep wintertime
mixed layers.</p>
      <p id="d1e2119">We calculate mixed layer depth using the density criterion of
<xref ref-type="bibr" rid="bib1.bibx141" id="text.106"/>: the shallowest depth at which the potential density is at
least 0.03 kg m<inline-formula><mml:math id="M97" display="inline"><mml:msup><mml:mi/><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">3</mml:mn></mml:mrow></mml:msup></mml:math></inline-formula> greater than at the surface (or at the ice-shelf
interface, in the case of ice-shelf cavities). Summer (DJF) and winter (JJA)
mixed layer depths in each simulation, averaged over the period 2002–2016,
are shown in Fig. <xref ref-type="fig" rid="Ch1.F2"/> for the entire Southern Ocean, and
Fig. <xref ref-type="fig" rid="Ch1.F3"/> zoomed into the Antarctic continental shelf.
Figure <xref ref-type="fig" rid="Ch1.F2"/> also includes climatological observations by
<xref ref-type="bibr" rid="bib1.bibx126" id="text.107"/> recalculated to use the same definition of mixed layer
depth as the models. We have not included these observations in
Fig. <xref ref-type="fig" rid="Ch1.F3"/>, as they are less reliable on the continental shelf
due to insufficient measurements.</p>
      <p id="d1e2149">In the ACC in summer (top row of Fig. <xref ref-type="fig" rid="Ch1.F2"/>), MetROMS shows a
ring of deeper mixed layers around 100 m surrounding the region stratified
by sea-ice meltwater. This spatial pattern agrees well with observations, but
the magnitude somewhat disagrees, as MetROMS' mixed layers are too deep in
the ACC and too shallow elsewhere. FESOM has a much<?pagebreak page1265?> more uniform summer mixed
layer depth which is 45 m (corresponding to the fourth layer in
<inline-formula><mml:math id="M98" display="inline"><mml:mi>z</mml:mi></mml:math></inline-formula>-coordinate regions) throughout most of the ACC, and generally shallower
in the <inline-formula><mml:math id="M99" display="inline"><mml:mi mathvariant="italic">σ</mml:mi></mml:math></inline-formula>-coordinate region of the continental shelf. Both models have
significantly deeper mixed layers in winter (bottom row of
Fig. <xref ref-type="fig" rid="Ch1.F2"/>; note the different colour scale) with the largest values
in the northern branch of the ACC where mode and intermediate waters subduct.
Observations indicate this feature should be strongest in the Pacific and
Australian sectors. MetROMS shows local maxima in both regions, but the
magnitude in the Pacific sector (approximately 250 m) is still quite low. FESOM
only captures this feature in the Pacific sector, but here it attains mixed
layer depths in excess of 500 m which exceeds observations. This
overestimation is less pronounced at high resolution. Elsewhere in the ACC,
FESOM's mixed layer depths (approximately 100 m) more or less agree with
observations, while in MetROMS they are too deep (approximately 200 m). The
tendency of MetROMS to have deeper mixed layers than FESOM may be influenced
by the differing surface stress between the two models
(Sect. <xref ref-type="sec" rid="Ch1.S2.SS7"/>).</p>

      <?xmltex \floatpos{t}?><fig id="Ch1.F3"><label>Figure 3</label><caption><p id="d1e2174">As Fig. <xref ref-type="fig" rid="Ch1.F2"/> for each model simulation, zoomed into the
Antarctic continental shelf.</p></caption>
            <?xmltex \igopts{width=241.848425pt}?><graphic xlink:href="https://gmd.copernicus.org/articles/11/1257/2018/gmd-11-1257-2018-f03.png"/>

          </fig>

      <?xmltex \floatpos{t}?><fig id="Ch1.F4" specific-use="star"><label>Figure 4</label><caption><p id="d1e2188">Temperature–salinity distribution south of <inline-formula><mml:math id="M100" display="inline"><mml:mn mathvariant="normal">65</mml:mn></mml:math></inline-formula><inline-formula><mml:math id="M101" display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula> S for
MetROMS, low-resolution FESOM, and high-resolution FESOM, averaged over the
years 2002–2016, and coloured based on depth (note the non-linear colour
scale). Each grid box (in MetROMS) or triangular prism (in FESOM) is sorted
into <inline-formula><mml:math id="M102" display="inline"><mml:mrow><mml:mn mathvariant="normal">1000</mml:mn><mml:mo>×</mml:mo><mml:mn mathvariant="normal">1000</mml:mn></mml:mrow></mml:math></inline-formula> temperature and salinity bins. The depth shown for
each bin is the volume-weighted average of the depths of the grid boxes or
triangular prisms within that bin. The dashed black line in each plot is the
surface freezing point, which has a slightly different formulation between
MetROMS and FESOM due to the different sea-ice thermodynamics schemes. The
dotted grey lines are potential density contours in kg m<inline-formula><mml:math id="M103" display="inline"><mml:msup><mml:mi/><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">3</mml:mn></mml:mrow></mml:msup></mml:math></inline-formula> <inline-formula><mml:math id="M104" display="inline"><mml:mo>-</mml:mo></mml:math></inline-formula> 1000.
Labels show different water masses: AABW indicates Antarctic Bottom Water,
WSBW indicates Weddell Sea bottom water, RSBW indicates Ross Sea bottom
water, CDW indicates Circumpolar Deep Water, MCDW indicates Modified
Circumpolar Deep Water, LSSW indicates low-salinity shelf water, HSSW
indicates high-salinity shelf water, AASW indicates Antarctic Surface Water,
and ISW indicates Ice-Shelf Water. Slanted labels below the freezing point
line show specific ice shelves' contributions to ISW.</p></caption>
            <?xmltex \igopts{width=426.791339pt}?><graphic xlink:href="https://gmd.copernicus.org/articles/11/1257/2018/gmd-11-1257-2018-f04.png"/>

          </fig>

      <?pagebreak page1266?><p id="d1e2243">Zooming into the continental shelf, the water column is largely stratified in
summer (top row of Fig. <xref ref-type="fig" rid="Ch1.F3"/>) but shows active regions of
dense water formation in winter (bottom row of Fig. <xref ref-type="fig" rid="Ch1.F3"/>; note
the different colour scale). Both MetROMS and FESOM form dense water in the
inner Ross and Weddell seas, with regions of mixed layer depth exceeding
500 m. Convection appears to be stronger in FESOM where these regions are
deeper and more widespread due to stronger sea-ice production
(Sect. <xref ref-type="sec" rid="Ch1.S4.SS2.SSS3"/>). In the Weddell Sea, dense water formation is
split into two regions on either side of the Filchner–Ronne Ice Shelf front,
with shallower mixed layers in the middle. In the Ross Sea, both models show
somewhat stronger convection on the western side of the Ross Ice Shelf front,
near McMurdo Sound, in agreement with observations <xref ref-type="bibr" rid="bib1.bibx70" id="paren.108"/>. A
small region of dense water formation in western Prydz Bay, adjacent to the
Amery Ice Shelf, is also present in both models. These regions are in
agreement with observed bottom water formation sites
<xref ref-type="bibr" rid="bib1.bibx29 bib1.bibx42 bib1.bibx59" id="paren.109"/>. FESOM also exhibits
deep mixed layers (<inline-formula><mml:math id="M105" display="inline"><mml:mo lspace="0mm">&gt;</mml:mo></mml:math></inline-formula> 500 m) in the Amundsen Sea, which were observed in
2012 but are not a consistent feature of this region <xref ref-type="bibr" rid="bib1.bibx20" id="paren.110"/>.
The presence of CDW on the Amundsen Sea continental shelf is sensitive to
mixed layer depth, as a completely destratified water column filled with cold
shelf water will prevent the development of a warmer bottom layer
<xref ref-type="bibr" rid="bib1.bibx127 bib1.bibx128" id="paren.111"/>. This mechanism has been proposed as a cause of
cold biases in Amundsen Sea ice-shelf cavities, and subsequent
underestimation of ice-shelf melt rates, in FESOM <xref ref-type="bibr" rid="bib1.bibx113" id="paren.112"/>. In
our simulations, these deep mixed layers have some dependence on resolution,
as they cover nearly the entire Amundsen Sea at low resolution but are more
restricted to the ice-shelf fronts at high resolution. Similarly,
low-resolution FESOM exhibits locally deepened mixed layers (approximately
250 m) near the southern entrance of George VI Ice Shelf in the
Bellingshausen Sea, while this feature is absent at high resolution.</p>
      <p id="d1e2275">Mixed layer depths in ice-shelf cavities show no significant seasonality
(note the different colour scales for summer and winter in
Fig. <xref ref-type="fig" rid="Ch1.F3"/>) and are generally shallow (<inline-formula><mml:math id="M106" display="inline"><mml:mo lspace="0mm">&lt;</mml:mo></mml:math></inline-formula> 50 m) except near
regions of persistent refreezing, which forms marine ice. This process
increases salinity at the ice-shelf base as freshwater is removed in the
form of frazil ice, providing a buoyancy forcing. Regions of marine ice
formation are detailed in Sect. <xref ref-type="sec" rid="Ch1.S4.SS3"/>, but their signature
can be seen here. The most affected region is the central Ronne Ice Shelf,
which has mixed layer depths of 300–400 m in MetROMS, 50–80 m in
low-resolution FESOM, and 70–120 m in high-resolution FESOM. Refreezing in
this region is indeed stronger and more widespread in MetROMS than in FESOM
(Sect. <xref ref-type="sec" rid="Ch1.S4.SS3.SSS1"/>). All three simulations exhibit mixed layer depths
exceeding 50 m in much of the Ross Ice Shelf, which has large areas of
refreezing (Sect. <xref ref-type="sec" rid="Ch1.S4.SS3.SSS5"/>). Only MetROMS shows increased mixed
layer depths (approximately 70 m) along the western edge of the Amery Ice Shelf,
which is a region of refreezing in MetROMS but not in FESOM
(Sect. <xref ref-type="sec" rid="Ch1.S4.SS3.SSS3"/>). Due to the lack of observations in ice-shelf
cavities, the true mixed layer depths in these regions are unknown.</p>
</sec>
<sec id="Ch1.S4.SS1.SSS3">
  <title>Water mass properties</title>
      <p id="d1e2302">Ice-shelf melt rates and sea-ice formation both influence, and are influenced
by, water mass properties on the continental shelf. Figure <xref ref-type="fig" rid="Ch1.F4"/>
plots the temperature–salinity (<inline-formula><mml:math id="M107" display="inline"><mml:mrow><mml:mi>T</mml:mi><mml:mo>/</mml:mo><mml:mi>S</mml:mi></mml:mrow></mml:math></inline-formula>) distribution south of <inline-formula><mml:math id="M108" display="inline"><mml:mn mathvariant="normal">65</mml:mn></mml:math></inline-formula><inline-formula><mml:math id="M109" display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula> S
in each simulation, averaged over 2002–2016, and colour-coded based on
depth. In this section, we identify the different water masses represented in
Fig. <xref ref-type="fig" rid="Ch1.F4"/>, and compare their properties between the two models.
Due to a scarcity of year-round measurements on the continental shelf, it is
not feasible to create a comparable figure using observations. However,
limited observations of some water masses exist and are compared to the
simulated water mass properties in the text below.</p>
      <p id="d1e2336">Just above the surface freezing temperature (dashed black lines in
Fig. <xref ref-type="fig" rid="Ch1.F4"/>, approximately <inline-formula><mml:math id="M110" display="inline"><mml:mo>-</mml:mo></mml:math></inline-formula>2 <inline-formula><mml:math id="M111" display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula>C) are two subsurface
water masses (100–500 m depth). Low-salinity shelf water (LSSW,
<inline-formula><mml:math id="M112" display="inline"><mml:mo>&lt;</mml:mo></mml:math></inline-formula> 34.5 psu) and high-salinity shelf water (HSSW, <inline-formula><mml:math id="M113" display="inline"><mml:mo>&gt;</mml:mo></mml:math></inline-formula> 34.5 psu) are both
the result of sea-ice formation, but HSSW is more affected by strong brine
rejection. LSSW shows similar properties in all three simulations,<?pagebreak page1267?> with
minimum salinities around 33.75 psu. HSSW is saltier in low-resolution FESOM
(up to 35.1 psu) than in high-resolution FESOM (up to 35 psu). This is the
main difference between the two FESOM simulations, which are otherwise very
similar in terms of water mass properties. MetROMS has fresher HSSW than
either FESOM simulation, with maximum salinities of approximately 34.8 psu.
The differing salinity of HSSW in each simulation corresponds to the relative
rates of sea-ice production, analysed in Sect. <xref ref-type="sec" rid="Ch1.S4.SS2.SSS3"/>.</p>
      <p id="d1e2374">At the higher end of the HSSW salinity range, and with temperatures up to
<inline-formula><mml:math id="M114" display="inline"><mml:mo>-</mml:mo></mml:math></inline-formula>1 <inline-formula><mml:math id="M115" display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula>C, is surface water (0–50 m) from the Ross Sea polynya. This
water mass is more prominent in the FESOM distributions than in MetROMS due
to its higher salinity. As with HSSW, FESOM's Ross Sea polynya is saltier at
low resolution.</p>
      <p id="d1e2393">The remainder of the surface water (50 m or shallower) is Antarctic Surface
Water (AASW) which has lower salinity, generally <inline-formula><mml:math id="M116" display="inline"><mml:mo>&lt;</mml:mo></mml:math></inline-formula> 34 psu, with
temperatures between the surface freezing point and 1 <inline-formula><mml:math id="M117" display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula>C. A spread
of points with particularly low salinity (<inline-formula><mml:math id="M118" display="inline"><mml:mo lspace="0mm">&lt;</mml:mo></mml:math></inline-formula> 33.7 psu) represents narrow
embayments on the western side of the Antarctic Peninsula from which
meltwater cannot easily escape.</p>
      <p id="d1e2420">Water masses below the surface freezing temperature are called Ice-Shelf
Water (ISW). The only way that a water mass can fall below this line
(neglecting numerical error in tracer advection) is from interaction with an
ice-shelf base. The freezing temperature of seawater decreases with depth due
to enhanced pressure, and at the deepest grounding lines it can approach
<inline-formula><mml:math id="M119" display="inline"><mml:mo>-</mml:mo></mml:math></inline-formula>3 <inline-formula><mml:math id="M120" display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula>C. Water which melts or refreezes at the ice-shelf base will
retain this freezing temperature until it is modified by mixing or by
melting/freezing at a different depth.</p>
      <p id="d1e2439">The temperature–salinity distributions of ISW follow distinct diagonals,
where the slope is the dilution ratio of melting or freezing ice in seawater
<xref ref-type="bibr" rid="bib1.bibx32" id="paren.113"/>. The three deepest ice-shelf cavities form the most
prominent diagonals in Fig. <xref ref-type="fig" rid="Ch1.F4"/>: in order of increasing
salinity, they are the Amery, the Filchner–Ronne, and the Ross. ISW beneath
the Ross Ice Shelf is saltiest in low-resolution FESOM and freshest in
MetROMS, consistent with the HSSW which feeds the cavity. In the Amery and
Filchner–Ronne cavities, high-resolution FESOM displays deeper water masses
than low-resolution FESOM, which is due to its better representation of deep
ice near the grounding line (Sect. <xref ref-type="sec" rid="Ch1.S4.SS3.SSS3"/>
and <xref ref-type="sec" rid="Ch1.S4.SS3.SSS1"/>).</p>
      <p id="d1e2451">The remaining water masses, in the deep Southern Ocean, have much longer
residence times and are therefore not fully spun up. Comparing their
simulated properties is useful to assess model drift (see also
Sect. <xref ref-type="sec" rid="Ch1.S4.SS1.SSS4"/>), but they should be evaluated with caution.</p>
      <p id="d1e2456">Antarctic Bottom Water (AABW) is the deepest water mass (1000 m or deeper)
with simulated salinity <inline-formula><mml:math id="M121" display="inline"><mml:mo>&gt;</mml:mo></mml:math></inline-formula> 34.5 psu and intermediate temperature (<inline-formula><mml:math id="M122" display="inline"><mml:mo lspace="0mm">-</mml:mo></mml:math></inline-formula>1 to
1.5 <inline-formula><mml:math id="M123" display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula>C). In both MetROMS and FESOM, the deepest AABW (below 2000 m)
forks into two distinct branches on either side of 34.7 psu. The
lower-salinity branch is Weddell Sea Bottom Water (WSBW) and the
higher-salinity branch is Ross Sea Bottom Water (RSBW). Limited observations
of these two water masses are available through the World Ocean Circulation
Experiment (WOCE) Atlas <xref ref-type="bibr" rid="bib1.bibx82 bib1.bibx155" id="paren.114"/>: track A23 through
the Weddell Sea (considering only the section south of <inline-formula><mml:math id="M124" display="inline"><mml:mn mathvariant="normal">65</mml:mn></mml:math></inline-formula><inline-formula><mml:math id="M125" display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula> S,
which has approximate longitude <inline-formula><mml:math id="M126" display="inline"><mml:mn mathvariant="normal">20</mml:mn></mml:math></inline-formula><inline-formula><mml:math id="M127" display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula> W, and below 2000 m) and
track S4P through the Ross Sea (considering only the section between
<inline-formula><mml:math id="M128" display="inline"><mml:mn mathvariant="normal">150</mml:mn></mml:math></inline-formula><inline-formula><mml:math id="M129" display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula> E and <inline-formula><mml:math id="M130" display="inline"><mml:mn mathvariant="normal">130</mml:mn></mml:math></inline-formula><inline-formula><mml:math id="M131" display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula> W, which has latitude <inline-formula><mml:math id="M132" display="inline"><mml:mn mathvariant="normal">67</mml:mn></mml:math></inline-formula><inline-formula><mml:math id="M133" display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula> S,
and below 2000 m). In these tracks, the salinity of WSBW ranges from 34.65 to
34.7 psu, and RSBW from 34.68 to 34.72 psu. The models' tendency for WSBW
to be fresher than RSBW is therefore supported by observations, and both
models are also in agreement with the observed salinity of WSBW. However,
they both overestimate the salinity of RSBW compared to these observations,
particularly FESOM which approaches 34.8 psu. Both water masses have more
uniform salinity in MetROMS than in FESOM, which is reflected by narrower red
lines in Fig. <xref ref-type="fig" rid="Ch1.F4"/>.</p>
      <p id="d1e2564">The same WOCE tracks measure temperatures from <inline-formula><mml:math id="M134" display="inline"><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">0.8</mml:mn></mml:mrow></mml:math></inline-formula> to <inline-formula><mml:math id="M135" display="inline"><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">0.2</mml:mn></mml:mrow></mml:math></inline-formula> <inline-formula><mml:math id="M136" display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula>C
for WSBW and <inline-formula><mml:math id="M137" display="inline"><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">0.4</mml:mn></mml:mrow></mml:math></inline-formula> to <inline-formula><mml:math id="M138" display="inline"><mml:mn mathvariant="normal">0.8</mml:mn></mml:math></inline-formula> <inline-formula><mml:math id="M139" display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula>C for RSBW. The observed tendency for
WSBW to be colder than RSBW is apparent in MetROMS (<inline-formula><mml:math id="M140" display="inline"><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">0.5</mml:mn></mml:mrow></mml:math></inline-formula> to
<inline-formula><mml:math id="M141" display="inline"><mml:mn mathvariant="normal">0.75</mml:mn></mml:math></inline-formula> <inline-formula><mml:math id="M142" display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula>C for WSBW, <inline-formula><mml:math id="M143" display="inline"><mml:mn mathvariant="normal">0.25</mml:mn></mml:math></inline-formula> to <inline-formula><mml:math id="M144" display="inline"><mml:mn mathvariant="normal">0.75</mml:mn></mml:math></inline-formula> <inline-formula><mml:math id="M145" display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula>C for RSBW) but the
two water masses have approximately the same temperature in FESOM (<inline-formula><mml:math id="M146" display="inline"><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">1</mml:mn></mml:mrow></mml:math></inline-formula> to
<inline-formula><mml:math id="M147" display="inline"><mml:mn mathvariant="normal">1</mml:mn></mml:math></inline-formula> <inline-formula><mml:math id="M148" display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula>C). The colder varieties of RSBW are absent in MetROMS, while
FESOM reaches temperatures which are significantly colder than WOCE
observations. In both models, simulated WSBW is too warm. However, these
observations do not sample the full spatial extent of the water masses, so
the true temperature and salinity may have a larger range.</p>
      <p id="d1e2699">Circumpolar Deep Water (CDW) is shallower than AABW (200–1000 m) and warmer
(<inline-formula><mml:math id="M149" display="inline"><mml:mrow><mml:mo>&gt;</mml:mo><mml:mn mathvariant="normal">0</mml:mn></mml:mrow></mml:math></inline-formula> <inline-formula><mml:math id="M150" display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula>C). In MetROMS, the temperature of CDW can exceed
<inline-formula><mml:math id="M151" display="inline"><mml:mn mathvariant="normal">3</mml:mn></mml:math></inline-formula> <inline-formula><mml:math id="M152" display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula>C, while it stays below approximately <inline-formula><mml:math id="M153" display="inline"><mml:mn mathvariant="normal">2.5</mml:mn></mml:math></inline-formula> <inline-formula><mml:math id="M154" display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula>C in
FESOM. The warmer CDW in MetROMS is consistent with increased southward
spreading of warmer CDW from the north around most of the continent, as
discussed in Sect. <xref ref-type="sec" rid="Ch1.S4.SS1.SSS4"/>. Observations of CDW in this region
suggest a temperature range of <inline-formula><mml:math id="M155" display="inline"><mml:mn mathvariant="normal">0.3</mml:mn></mml:math></inline-formula> to <inline-formula><mml:math id="M156" display="inline"><mml:mn mathvariant="normal">2.5</mml:mn></mml:math></inline-formula> <inline-formula><mml:math id="M157" display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula>C
<xref ref-type="bibr" rid="bib1.bibx142" id="paren.115"/>. Both models exhibit curling, finger-like structures on
the low-salinity (left) side of the CDW distribution. These features
represent meanders of the ACC over the boundary <inline-formula><mml:math id="M158" display="inline"><mml:mn mathvariant="normal">65</mml:mn></mml:math></inline-formula><inline-formula><mml:math id="M159" display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula> S, and these
meanders transport different properties southward in different geographical
locations. As CDW enters the subpolar gyres, it mixes with other water masses
to produce cooler Modified Circumpolar Deep Water (MCDW).</p>

      <?xmltex \floatpos{t}?><fig id="Ch1.F5" specific-use="star"><label>Figure 5</label><caption><p id="d1e2801">Temperature in <inline-formula><mml:math id="M160" display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula>C (left) and salinity in psu (right)
interpolated to <inline-formula><mml:math id="M161" display="inline"><mml:mn mathvariant="normal">0</mml:mn></mml:math></inline-formula><inline-formula><mml:math id="M162" display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula> E (Greenwich Meridian). Black contours show the
<inline-formula><mml:math id="M163" display="inline"><mml:mn mathvariant="normal">0.75</mml:mn></mml:math></inline-formula> <inline-formula><mml:math id="M164" display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula>C isotherm and the 34.5 psu isohaline. <bold>(a)</bold> Initial
conditions for January 1992 from the ECCO2 reanalysis
<xref ref-type="bibr" rid="bib1.bibx106 bib1.bibx177" id="paren.116"/>. <bold>(b–d)</bold> January 2016 monthly
average for MetROMS, low-resolution FESOM, and high-resolution FESOM,
respectively.</p></caption>
            <?xmltex \igopts{width=284.527559pt}?><graphic xlink:href="https://gmd.copernicus.org/articles/11/1257/2018/gmd-11-1257-2018-f05.png"/>

          </fig>

      <?xmltex \floatpos{t}?><fig id="Ch1.F6" specific-use="star"><label>Figure 6</label><caption><p id="d1e2862"><bold>(a)</bold> The 1992–2015 mean Antarctic sea-ice concentration for
February (top) and September (bottom), comparing NSIDC observations
(NOAA/NSIDC Climate Data Record of Passive Microwave Sea Ice Concentration)
<xref ref-type="bibr" rid="bib1.bibx105" id="paren.117"/>, MetROMS, low-resolution FESOM, and high-resolution FESOM.
<bold>(b)</bold> Time series of total Antarctic sea-ice extent in millions of
km<inline-formula><mml:math id="M165" display="inline"><mml:msup><mml:mi/><mml:mn mathvariant="normal">2</mml:mn></mml:msup></mml:math></inline-formula> for February (top) and September (bottom), comparing NSIDC
observations (NSIDC Sea Ice Index version 2) <xref ref-type="bibr" rid="bib1.bibx26" id="paren.118"/>, MetROMS,
low-resolution FESOM, and high-resolution FESOM.</p></caption>
            <?xmltex \igopts{width=426.791339pt}?><graphic xlink:href="https://gmd.copernicus.org/articles/11/1257/2018/gmd-11-1257-2018-f06.png"/>

          </fig>

</sec>
<sec id="Ch1.S4.SS1.SSS4">
  <title>Deep ocean drift</title>
      <p id="d1e2897">As our experiments do not include a full spinup, it is useful to examine
changes in the properties of deep water masses during the simulations and
compare the different ways the models are drifting. Some of these changes may
be forced, as our forcing period of 1992–2016 is not a steady-state<?pagebreak page1268?> climate.
Other changes may be due to model deficiencies, such as artificial diapycnal
mixing (by which water masses over-mix) or sea-ice biases affecting deep
water formation.</p>
      <p id="d1e2900">Figure <xref ref-type="fig" rid="Ch1.F5"/> shows meridional slices of temperature and salinity
along <inline-formula><mml:math id="M166" display="inline"><mml:mn mathvariant="normal">0</mml:mn></mml:math></inline-formula><inline-formula><mml:math id="M167" display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula> E (Greenwich Meridian), comparing the ECCO2 initial
conditions for January 1992 (a) with the January 2016 monthly average for
MetROMS (b), low-resolution FESOM (c), and high-resolution FESOM (d). Greater
smoothing of the FESOM bathymetry compared to MetROMS or ECCO2 is apparent,
as deep ocean seamounts in the coarse-resolution regions north of
<inline-formula><mml:math id="M168" display="inline"><mml:mn mathvariant="normal">55</mml:mn></mml:math></inline-formula><inline-formula><mml:math id="M169" display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula> S are less pronounced. This is somewhat alleviated with higher
resolution.</p>
      <p id="d1e2935">Antarctic Intermediate Water (AAIW), the subsurface water mass north of
approximately <inline-formula><mml:math id="M170" display="inline"><mml:mn mathvariant="normal">50</mml:mn></mml:math></inline-formula><inline-formula><mml:math id="M171" display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula> S characterised by relatively low salinity
(<inline-formula><mml:math id="M172" display="inline"><mml:mo lspace="0mm">&lt;</mml:mo></mml:math></inline-formula> 34.5 psu, shown as a black contour in Fig. <xref ref-type="fig" rid="Ch1.F5"/>), shows
some degree of erosion in all three simulations. Difficulty preserving AAIW
is a very common problem among ocean models and is generally attributed to
spurious diapycnal mixing <xref ref-type="bibr" rid="bib1.bibx22 bib1.bibx23" id="paren.119"/> with a potential
contribution from errors in surface forcing <xref ref-type="bibr" rid="bib1.bibx46" id="paren.120"/>. The erosion
is most severe in MetROMS, and is combined with freshening of the underlying
North Atlantic Deep Water (NADW). Since MetROMS has terrain-following
coordinates throughout the entire domain, whereas FESOM has <inline-formula><mml:math id="M173" display="inline"><mml:mi>z</mml:mi></mml:math></inline-formula> coordinates
everywhere except the Antarctic continental shelf, MetROMS would indeed be
expected to be more prone to diapycnal mixing in the deep ocean
<xref ref-type="bibr" rid="bib1.bibx45" id="paren.121"/>, particularly around steep regions of bathymetry such as
seamounts. The degree of AAIW erosion in MetROMS depends on the tracer
advection scheme <xref ref-type="bibr" rid="bib1.bibx98 bib1.bibx91" id="paren.122"/>, and our choice of the
Akima advection scheme over the upwind third-order scheme
(Sect. <xref ref-type="sec" rid="Ch1.S2.SS4"/>) was motivated by the less severe diapycnal
mixing in Akima. In FESOM, AAIW is slightly better preserved at low
resolution than at high resolution. This agrees with the results of
<?pagebreak page1269?><xref ref-type="bibr" rid="bib1.bibx98" id="text.123"/> showing that in non-eddy-resolving regimes, spurious
diapycnal mixing tends to increase as resolution is refined.</p>
      <p id="d1e2987">Another notable feature in Fig. <xref ref-type="fig" rid="Ch1.F5"/> is the larger volume of warm
CDW (<inline-formula><mml:math id="M174" display="inline"><mml:mo lspace="0mm">&gt;</mml:mo></mml:math></inline-formula> 0.75 <inline-formula><mml:math id="M175" display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula>C, shown as a black contour) south of
<inline-formula><mml:math id="M176" display="inline"><mml:mn mathvariant="normal">60</mml:mn></mml:math></inline-formula><inline-formula><mml:math id="M177" display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula> S in MetROMS. A slight warming of the underlying AABW is also
apparent, likely due to spurious entrainment of the CDW through diapycnal
mixing. The cause of this increased CDW upwelling in MetROMS is not obvious.
Warming and shoaling of CDW around most regions of Antarctica have
been observed over recent decades and attributed to changes in wind
stress <xref ref-type="bibr" rid="bib1.bibx142 bib1.bibx151 bib1.bibx152" id="paren.124"/>. With these observations
in mind, it is possible that this behaviour is due to MetROMS' surface
exchange scheme, which leads to stronger surface stress than in FESOM
(Sect. <xref ref-type="sec" rid="Ch1.S2.SS7"/>). However, CDW upwelling is also sensitive
to the tracer advection scheme in MetROMS and is more severe with the upwind
third-order advection scheme (not shown). Therefore, some component of
numerical error could be an additional contributing factor.</p>
</sec>
</sec>
<sec id="Ch1.S4.SS2">
  <title>Sea ice</title>
<sec id="Ch1.S4.SS2.SSS1">
  <title>Concentration and extent</title>
      <p id="d1e3041">Sea-ice concentration (the fraction of each grid cell covered by ice) and
extent (the area of grid cells with concentration exceeding 0.15) are the
most convenient variables for model evaluation due to the availability of
satellite observations. These variables are largely a reflection of
atmospheric conditions but are also influenced by ocean processes, such as
upwelling of warmer water from below, and the pathway of the ACC. Here, we
compare with the NOAA/NSIDC Climate Data Record of Passive Microwave Sea Ice
Concentration <xref ref-type="bibr" rid="bib1.bibx105" id="paren.125"/> and the NSIDC Sea Ice Index version 2 for
sea-ice extent <xref ref-type="bibr" rid="bib1.bibx26" id="paren.126"/>. We examine monthly averages for February
and September, which are the months of minimum and maximum Antarctic sea-ice
extent, respectively, over the period of 1992–2015 (observations for 2016 were
not yet available at the time of writing).</p>
      <p id="d1e3050">Figure <xref ref-type="fig" rid="Ch1.F6"/> compares time-averaged sea-ice concentration (a) as
well as time series of total sea-ice extent (b) for February and September,
between NSIDC observations, MetROMS, low-resolution FESOM, and
high-resolution FESOM. All three of our simulations underestimate the sea-ice
minimum, which is a common bias seen in other stand-alone ocean/sea-ice models
forced with ERA-Interim <xref ref-type="bibr" rid="bib1.bibx85" id="paren.127"/> as well as in fully coupled GCMs
<xref ref-type="bibr" rid="bib1.bibx168" id="paren.128"/>. The majority of simulated February sea ice is in the
Weddell Sea (Fig. <xref ref-type="fig" rid="Ch1.F6"/>a, top row), which agrees with observations,
although in both FESOM simulations it extends too far northeast into the
Weddell Gyre. Observed patches of coastal ice in the Amundsen and
Bellingshausen seas, as well as along the coast of East Antarctica, are
largely absent in MetROMS and almost completely absent in FESOM. The
time series in Fig. <xref ref-type="fig" rid="Ch1.F6"/>b (top panel) reveal that all three
simulations underestimate February total sea-ice extent by approximately a
factor of 2 compared to observations. However, they all display some of the
observed interannual variability, such as the high in 2008 and the low in
2011, likely because observed sea-ice cover is imprinted on the ERA-Interim
atmospheric fields used to force the models.</p>
      <p id="d1e3065">In FESOM, the sea-ice minimum is slightly greater at high resolution. This
difference is driven by summertime<?pagebreak page1270?> conditions in the southern Weddell Sea and
the east coast of the Antarctic Peninsula. In the low-resolution mesh,
smoother bathymetry near the peninsula allows a spurious southward excursion
of the southern boundary of the ACC in summer, which carries warmer water
into the region and melts more sea ice.</p>
      <p id="d1e3068">The sea-ice maximum in September is well captured by all three simulations,
which exhibit zonal asymmetry in line with observations
(Fig. <xref ref-type="fig" rid="Ch1.F6"/>a, bottom row). Sea-ice concentrations throughout most of
the ice pack are lower in MetROMS (approximately 0.94) than in both FESOM
simulations (approximately 0.995). Observations from NSIDC fall in the middle
(approximately 0.97), which is not significantly different from either model if
observational uncertainty is considered. Nonetheless, this difference between
the models influences the air–sea fluxes, which are modulated by the sea-ice
concentration. For example, the ocean in MetROMS will experience slightly
greater wind stress than in FESOM and therefore more turbulent mixing. In
particular, sea-ice concentration affects the air–sea heat fluxes, which may
shed some light on the spurious Weddell Sea deep convection seen in MetROMS
(without surface salinity restoring) but not in FESOM, as described in
Sect. <xref ref-type="sec" rid="Ch1.S3.SS3"/>. Winter sea-ice concentrations far below 1 in
MetROMS allow frazil ice to form in the middle of the ice pack, rather than
being restricted to coastal polynyas. This introduces a positive feedback by
which brine rejection increases the sea surface salinity, causing
destabilisation of the water column and upwelling of warm water, which melts
surrounding sea ice and exposes more open water to the cold atmosphere. By
contrast, FESOM's winter sea ice has concentrations near 1 almost everywhere,
which shields the ocean surface from atmospheric heat fluxes and the
resulting frazil ice formation and brine rejection. However, differences in
vertical mixing schemes between the two models could also affect their
sensitivity to spurious Weddell Sea deep convection <xref ref-type="bibr" rid="bib1.bibx159" id="paren.129"/>,
as discussed in Sect. <xref ref-type="sec" rid="Ch1.S2.SS4"/>.</p>
      <p id="d1e3081">While the general pattern of both models' September sea ice agrees with
observations, the northern edge of the ice pack is too far south in MetROMS
and too far north in FESOM, which is possibly related to differences in mixed
layer depth (Sect. <xref ref-type="sec" rid="Ch1.S4.SS1.SSS2"/>) or in the path of the ACC. These
discrepancies are reflected in the time series of September sea-ice extent
(Fig. <xref ref-type="fig" rid="Ch1.F6"/>b, bottom panel) where the NSIDC observations fall
between the MetROMS and FESOM simulations. Interannual variability is well
represented, with both models reproducing many of the highs and lows seen in
the observations. No significant difference in winter sea-ice cover is
apparent between the low-resolution and high-resolution FESOM simulations.</p>
</sec>
<sec id="Ch1.S4.SS2.SSS2">
  <title>Thickness</title>
      <p id="d1e3094">Sea-ice thickness is influenced by both thermodynamics (sea-ice formation and
melt) and dynamics (sea-ice transport). Observations of sea-ice thickness are
scarce and have large uncertainties <xref ref-type="bibr" rid="bib1.bibx64" id="paren.130"/>. A comprehensive
evaluation of MetROMS and FESOM with respect to sea-ice thickness is
therefore difficult, although a comparison of the two models can still be
made. Figure <xref ref-type="fig" rid="Ch1.F7"/> shows seasonal averages of sea-ice effective
thickness (concentration multiplied by height) in each simulation averaged over
1992–2016.</p>
      <p id="d1e3102">Sea ice is generally thicker in MetROMS than in either FESOM simulation,
particularly in the Weddell Sea, the Amundsen and Bellingshausen seas, and
along the coastline of East Antarctica. This difference may be due to complex
dynamic processes such as ridging and rafting, which are not considered by
single-layer sea-ice models such as the one used in FESOM. However, FESOM's
coastal sea ice is slightly thicker at high resolution, particularly in the
Amundsen and Bellingshausen seas.</p>
      <p id="d1e3105">In MetROMS, a particularly thick region of sea ice (approximately 3 m) exists on
the western edge of the Weddell Sea, along the Antarctic Peninsula. This
feature is also present in IceSAT observations <xref ref-type="bibr" rid="bib1.bibx83 bib1.bibx64" id="paren.131"/>,
and in situ measurements of second-year ice in the western Weddell Sea find
thicknesses of 2.4 to 2.9 m <xref ref-type="bibr" rid="bib1.bibx51" id="paren.132"/>. The region of thick ice is
less pronounced, but still visible, in the high-resolution FESOM simulation.
In low-resolution FESOM, the southward excursion of the southern boundary of
the ACC in summer (see Sect. <xref ref-type="sec" rid="Ch1.S4.SS2.SSS1"/>) prevents multi-year ice from
building up in this region, so the feature is mostly absent. All three
simulations show some sign of the Ronne polynya in winter (JJA) and spring
(SON), with thinner sea ice near the Ronne Depression. Thicker ice is present
directly in front of the Filchner Ice Shelf, especially in FESOM.</p>

      <?xmltex \floatpos{t}?><fig id="Ch1.F7" specific-use="star"><label>Figure 7</label><caption><p id="d1e3118">The 1992–2016 mean seasonal Antarctic sea-ice effective thickness
(concentration times height, measured in metres) for MetROMS, low-resolution
FESOM, and high-resolution FESOM.</p></caption>
            <?xmltex \igopts{width=341.433071pt}?><graphic xlink:href="https://gmd.copernicus.org/articles/11/1257/2018/gmd-11-1257-2018-f07.png"/>

          </fig>

      <?xmltex \floatpos{t}?><fig id="Ch1.F8" specific-use="star"><label>Figure 8</label><caption><p id="d1e3130">Sea-ice production (<inline-formula><mml:math id="M178" display="inline"><mml:mrow><mml:msup><mml:mn mathvariant="normal">10</mml:mn><mml:mn mathvariant="normal">9</mml:mn></mml:msup></mml:mrow></mml:math></inline-formula> m<inline-formula><mml:math id="M179" display="inline"><mml:msup><mml:mi/><mml:mn mathvariant="normal">3</mml:mn></mml:msup></mml:math></inline-formula> yr<inline-formula><mml:math id="M180" display="inline"><mml:msup><mml:mi/><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">1</mml:mn></mml:mrow></mml:msup></mml:math></inline-formula>) on the continental
shelf (defined as regions south of <inline-formula><mml:math id="M181" display="inline"><mml:mn mathvariant="normal">60</mml:mn></mml:math></inline-formula><inline-formula><mml:math id="M182" display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula> S with bathymetry shallower
than 1500 m), integrated over <inline-formula><mml:math id="M183" display="inline"><mml:mn mathvariant="normal">1</mml:mn></mml:math></inline-formula><inline-formula><mml:math id="M184" display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula> longitude bins. Results are shown
for MetROMS, low-resolution FESOM, high-resolution FESOM, and the
observation-based estimate of <xref ref-type="bibr" rid="bib1.bibx157" id="text.133"/> which uses ERA-Interim heat
fluxes for its calculation.</p></caption>
            <?xmltex \igopts{width=398.338583pt}?><graphic xlink:href="https://gmd.copernicus.org/articles/11/1257/2018/gmd-11-1257-2018-f08.png"/>

          </fig>

</sec>
<sec id="Ch1.S4.SS2.SSS3">
  <title>Sea-ice production</title>
      <p id="d1e3211">As discussed in Sect. <xref ref-type="sec" rid="Ch1.S4.SS1.SSS2"/>, the strength of sea-ice formation is
a key determinant of mixed layer depth, particularly in coastal polynyas on
the Antarctic continental shelf where most sea ice is formed.
Figure <xref ref-type="fig" rid="Ch1.F8"/> compares sea-ice production in each simulation to the
observation-based estimate of <xref ref-type="bibr" rid="bib1.bibx157" id="text.134"/>. Sea-ice production is
integrated over <inline-formula><mml:math id="M185" display="inline"><mml:mn mathvariant="normal">1</mml:mn></mml:math></inline-formula><inline-formula><mml:math id="M186" display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula> longitude bins on the continental shelf (defined
as in Sect. <xref ref-type="sec" rid="Ch1.S3.SS3"/>) and averaged over the observed period
of 1992–2013. Note that <xref ref-type="bibr" rid="bib1.bibx157" id="text.135"/>'s calculation is integrated daily,
but sea-ice production in the models is calculated based on 5-day averaged
fluxes. These fluxes account for both melting and freezing, so sea-ice
production is only accumulated over 5-day periods with net freezing. As a
result, diagnosed sea-ice production in the models may be underestimated in
regions which switch between melting and freezing on the 1- to 5-day timescale,
but this discrepancy is expected to be small.</p>
      <p id="d1e3242">Compared to <xref ref-type="bibr" rid="bib1.bibx157" id="text.136"/>, all three simulations overestimate sea-ice
production in the Ross and Weddell seas; this bias is somewhat larger in
FESOM and is slightly<?pagebreak page1271?> alleviated at high resolution. In Prydz Bay, all three
simulations display a peak in sea-ice formation; here, FESOM agrees with
observations, but MetROMS produces an overestimate. Further east in the
Australian sector, the models struggle to capture the observed peaks in
sea-ice formation seen in small coastal polynyas, such as the Dalton polynya
near <inline-formula><mml:math id="M187" display="inline"><mml:mn mathvariant="normal">120</mml:mn></mml:math></inline-formula><inline-formula><mml:math id="M188" display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula> E. The Amundsen polynya (approximately
<inline-formula><mml:math id="M189" display="inline"><mml:mn mathvariant="normal">110</mml:mn></mml:math></inline-formula><inline-formula><mml:math id="M190" display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula> W) is also not well captured by the models. However, further
east in the Amundsen Sea (approximately <inline-formula><mml:math id="M191" display="inline"><mml:mn mathvariant="normal">105</mml:mn></mml:math></inline-formula><inline-formula><mml:math id="M192" display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula> W), near the Pine
Island and Thwaites Ice Shelf fronts, FESOM overestimates sea-ice production.
The implications of these regional biases for water mass properties and
ice-shelf melt rates are discussed in Sect. <xref ref-type="sec" rid="Ch1.S4.SS3"/>.</p>
      <?pagebreak page1272?><p id="d1e3296">Note that <xref ref-type="bibr" rid="bib1.bibx157" id="text.137"/>'s calculation makes use of heat flux values from
ERA-Interim, in addition to satellite observations of sea ice. Therefore, if
biases in ERA-Interim are affecting our simulations, they may also be
affecting <xref ref-type="bibr" rid="bib1.bibx157" id="text.138"/>'s estimates to some extent.</p>

<?xmltex \floatpos{ph!}?><table-wrap id="Ch1.T1" specific-use="star"><label>Table 1</label><caption><p id="d1e3308">Ice-shelf basal mass loss (Gt yr<inline-formula><mml:math id="M193" display="inline"><mml:msup><mml:mi/><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">1</mml:mn></mml:mrow></mml:msup></mml:math></inline-formula>) for all ice shelves with
area exceeding 5000 km<inline-formula><mml:math id="M194" display="inline"><mml:msup><mml:mi/><mml:mn mathvariant="normal">2</mml:mn></mml:msup></mml:math></inline-formula> as measured by <xref ref-type="bibr" rid="bib1.bibx134" id="text.139"/>. In some cases,
multiple ice shelves have been combined (e.g. Brunt and Riiser–Larsen) because
the boundaries between them in the model domains are not distinct. The ice
shelves have been sorted into the eight regions analysed in
Sect. <xref ref-type="sec" rid="Ch1.S4.SS3"/>. Values are shown for the MetROMS,
low-resolution FESOM, and high-resolution FESOM simulations averaged over the
years 2002–2016 and are compared to the range of observational estimates
given by <xref ref-type="bibr" rid="bib1.bibx134" id="text.140"/> for the period 2003–2008. Mass loss values from
model simulations are marked with (<inline-formula><mml:math id="M195" display="inline"><mml:mo lspace="0mm">-</mml:mo></mml:math></inline-formula>) or (<inline-formula><mml:math id="M196" display="inline"><mml:mo lspace="0mm">+</mml:mo></mml:math></inline-formula>) if they fall below or above
(respectively) the range given by <xref ref-type="bibr" rid="bib1.bibx134" id="text.141"/>.</p></caption><oasis:table frame="topbot"><oasis:tgroup cols="5">
     <oasis:colspec colnum="1" colname="col1" align="left"/>
     <oasis:colspec colnum="2" colname="col2" align="right"/>
     <oasis:colspec colnum="3" colname="col3" align="right"/>
     <oasis:colspec colnum="4" colname="col4" align="right"/>
     <oasis:colspec colnum="5" colname="col5" align="right"/>
     <oasis:thead>
       <oasis:row rowsep="1">
         <oasis:entry colname="col1"/>
         <oasis:entry colname="col2">MetROMS</oasis:entry>
         <oasis:entry colname="col3">FESOM low-res</oasis:entry>
         <oasis:entry colname="col4">FESOM high-res</oasis:entry>
         <oasis:entry colname="col5">Rignot et al.</oasis:entry>
       </oasis:row>
     </oasis:thead>
     <oasis:tbody>
       <oasis:row rowsep="1">
         <oasis:entry namest="col1" nameend="col5">1. Filchner–Ronne </oasis:entry>
       </oasis:row>
       <oasis:row rowsep="1">
         <oasis:entry colname="col1"/>
         <oasis:entry colname="col2">46.0 (<inline-formula><mml:math id="M197" display="inline"><mml:mo lspace="0mm">-</mml:mo></mml:math></inline-formula>)</oasis:entry>
         <oasis:entry colname="col3">113.8</oasis:entry>
         <oasis:entry colname="col4">115.4</oasis:entry>
         <oasis:entry colname="col5"><inline-formula><mml:math id="M198" display="inline"><mml:mrow><mml:mn mathvariant="normal">155.4</mml:mn><mml:mo>±</mml:mo><mml:mn mathvariant="normal">45</mml:mn></mml:mrow></mml:math></inline-formula></oasis:entry>
       </oasis:row>
       <oasis:row rowsep="1">
         <oasis:entry namest="col1" nameend="col5">2. Eastern Weddell region </oasis:entry>
       </oasis:row>
       <oasis:row>
         <oasis:entry colname="col1">Brunt and Riiser–Larsen</oasis:entry>
         <oasis:entry colname="col2">29.2 (<inline-formula><mml:math id="M199" display="inline"><mml:mo lspace="0mm">+</mml:mo></mml:math></inline-formula>)</oasis:entry>
         <oasis:entry colname="col3">33.4 (<inline-formula><mml:math id="M200" display="inline"><mml:mo lspace="0mm">+</mml:mo></mml:math></inline-formula>)</oasis:entry>
         <oasis:entry colname="col4">34.6 (<inline-formula><mml:math id="M201" display="inline"><mml:mo lspace="0mm">+</mml:mo></mml:math></inline-formula>)</oasis:entry>
         <oasis:entry colname="col5"><inline-formula><mml:math id="M202" display="inline"><mml:mrow><mml:mn mathvariant="normal">9.7</mml:mn><mml:mo>±</mml:mo><mml:mn mathvariant="normal">16</mml:mn></mml:mrow></mml:math></inline-formula></oasis:entry>
       </oasis:row>
       <oasis:row>
         <oasis:entry colname="col1">Fimbul, Jelbart, and Ekström</oasis:entry>
         <oasis:entry colname="col2">30.3</oasis:entry>
         <oasis:entry colname="col3">41.8 (<inline-formula><mml:math id="M203" display="inline"><mml:mo lspace="0mm">+</mml:mo></mml:math></inline-formula>)</oasis:entry>
         <oasis:entry colname="col4">52.4 (<inline-formula><mml:math id="M204" display="inline"><mml:mo lspace="0mm">+</mml:mo></mml:math></inline-formula>)</oasis:entry>
         <oasis:entry colname="col5"><inline-formula><mml:math id="M205" display="inline"><mml:mrow><mml:mn mathvariant="normal">26.8</mml:mn><mml:mo>±</mml:mo><mml:mn mathvariant="normal">14</mml:mn></mml:mrow></mml:math></inline-formula></oasis:entry>
       </oasis:row>
       <oasis:row>
         <oasis:entry colname="col1">Nivl</oasis:entry>
         <oasis:entry colname="col2">3.4</oasis:entry>
         <oasis:entry colname="col3">5.4</oasis:entry>
         <oasis:entry colname="col4">5.9</oasis:entry>
         <oasis:entry colname="col5"><inline-formula><mml:math id="M206" display="inline"><mml:mrow><mml:mn mathvariant="normal">3.9</mml:mn><mml:mo>±</mml:mo><mml:mn mathvariant="normal">2</mml:mn></mml:mrow></mml:math></inline-formula></oasis:entry>
       </oasis:row>
       <oasis:row>
         <oasis:entry colname="col1">Lazarev</oasis:entry>
         <oasis:entry colname="col2">2.9 (<inline-formula><mml:math id="M207" display="inline"><mml:mo lspace="0mm">-</mml:mo></mml:math></inline-formula>)</oasis:entry>
         <oasis:entry colname="col3">4.9</oasis:entry>
         <oasis:entry colname="col4">4.9</oasis:entry>
         <oasis:entry colname="col5"><inline-formula><mml:math id="M208" display="inline"><mml:mrow><mml:mn mathvariant="normal">6.3</mml:mn><mml:mo>±</mml:mo><mml:mn mathvariant="normal">2</mml:mn></mml:mrow></mml:math></inline-formula></oasis:entry>
       </oasis:row>
       <oasis:row>
         <oasis:entry colname="col1">Baudouin and Borchgrevink</oasis:entry>
         <oasis:entry colname="col2">28.4</oasis:entry>
         <oasis:entry colname="col3">35.7</oasis:entry>
         <oasis:entry colname="col4">36.5</oasis:entry>
         <oasis:entry colname="col5"><inline-formula><mml:math id="M209" display="inline"><mml:mrow><mml:mn mathvariant="normal">21.6</mml:mn><mml:mo>±</mml:mo><mml:mn mathvariant="normal">18</mml:mn></mml:mrow></mml:math></inline-formula></oasis:entry>
       </oasis:row>
       <oasis:row rowsep="1">
         <oasis:entry colname="col1">Prince Harald</oasis:entry>
         <oasis:entry colname="col2">5.4 (<inline-formula><mml:math id="M210" display="inline"><mml:mo lspace="0mm">+</mml:mo></mml:math></inline-formula>)</oasis:entry>
         <oasis:entry colname="col3">2.1 (<inline-formula><mml:math id="M211" display="inline"><mml:mo lspace="0mm">+</mml:mo></mml:math></inline-formula>)</oasis:entry>
         <oasis:entry colname="col4">2.6 (<inline-formula><mml:math id="M212" display="inline"><mml:mo lspace="0mm">+</mml:mo></mml:math></inline-formula>)</oasis:entry>
         <oasis:entry colname="col5"><inline-formula><mml:math id="M213" display="inline"><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">2</mml:mn><mml:mo>±</mml:mo><mml:mn mathvariant="normal">3</mml:mn></mml:mrow></mml:math></inline-formula></oasis:entry>
       </oasis:row>
       <oasis:row rowsep="1">
         <oasis:entry namest="col1" nameend="col5">3. Amery </oasis:entry>
       </oasis:row>
       <oasis:row rowsep="1">
         <oasis:entry colname="col1"/>
         <oasis:entry colname="col2">91.0 (<inline-formula><mml:math id="M214" display="inline"><mml:mo lspace="0mm">+</mml:mo></mml:math></inline-formula>)</oasis:entry>
         <oasis:entry colname="col3">71.0 (<inline-formula><mml:math id="M215" display="inline"><mml:mo lspace="0mm">+</mml:mo></mml:math></inline-formula>)</oasis:entry>
         <oasis:entry colname="col4">71.4 (<inline-formula><mml:math id="M216" display="inline"><mml:mo lspace="0mm">+</mml:mo></mml:math></inline-formula>)</oasis:entry>
         <oasis:entry colname="col5"><inline-formula><mml:math id="M217" display="inline"><mml:mrow><mml:mn mathvariant="normal">35.5</mml:mn><mml:mo>±</mml:mo><mml:mn mathvariant="normal">23</mml:mn></mml:mrow></mml:math></inline-formula></oasis:entry>
       </oasis:row>
       <oasis:row rowsep="1">
         <oasis:entry namest="col1" nameend="col5">4. Australian sector </oasis:entry>
       </oasis:row>
       <oasis:row>
         <oasis:entry colname="col1">West</oasis:entry>
         <oasis:entry colname="col2">11.5 (<inline-formula><mml:math id="M218" display="inline"><mml:mo lspace="0mm">-</mml:mo></mml:math></inline-formula>)</oasis:entry>
         <oasis:entry colname="col3">10.2 (<inline-formula><mml:math id="M219" display="inline"><mml:mo lspace="0mm">-</mml:mo></mml:math></inline-formula>)</oasis:entry>
         <oasis:entry colname="col4">12.7 (<inline-formula><mml:math id="M220" display="inline"><mml:mo lspace="0mm">-</mml:mo></mml:math></inline-formula>)</oasis:entry>
         <oasis:entry colname="col5"><inline-formula><mml:math id="M221" display="inline"><mml:mrow><mml:mn mathvariant="normal">27.2</mml:mn><mml:mo>±</mml:mo><mml:mn mathvariant="normal">10</mml:mn></mml:mrow></mml:math></inline-formula></oasis:entry>
       </oasis:row>
       <oasis:row>
         <oasis:entry colname="col1">Shackleton</oasis:entry>
         <oasis:entry colname="col2">14.3 (<inline-formula><mml:math id="M222" display="inline"><mml:mo lspace="0mm">-</mml:mo></mml:math></inline-formula>)</oasis:entry>
         <oasis:entry colname="col3">17.4 (<inline-formula><mml:math id="M223" display="inline"><mml:mo lspace="0mm">-</mml:mo></mml:math></inline-formula>)</oasis:entry>
         <oasis:entry colname="col4">22.7 (<inline-formula><mml:math id="M224" display="inline"><mml:mo lspace="0mm">-</mml:mo></mml:math></inline-formula>)</oasis:entry>
         <oasis:entry colname="col5"><inline-formula><mml:math id="M225" display="inline"><mml:mrow><mml:mn mathvariant="normal">72.6</mml:mn><mml:mo>±</mml:mo><mml:mn mathvariant="normal">15</mml:mn></mml:mrow></mml:math></inline-formula></oasis:entry>
       </oasis:row>
       <oasis:row>
         <oasis:entry colname="col1">Totten and Moscow University</oasis:entry>
         <oasis:entry colname="col2">9.5 (<inline-formula><mml:math id="M226" display="inline"><mml:mo lspace="0mm">-</mml:mo></mml:math></inline-formula>)</oasis:entry>
         <oasis:entry colname="col3">4.4 (<inline-formula><mml:math id="M227" display="inline"><mml:mo lspace="0mm">-</mml:mo></mml:math></inline-formula>)</oasis:entry>
         <oasis:entry colname="col4">9.3 (<inline-formula><mml:math id="M228" display="inline"><mml:mo lspace="0mm">-</mml:mo></mml:math></inline-formula>)</oasis:entry>
         <oasis:entry colname="col5"><inline-formula><mml:math id="M229" display="inline"><mml:mrow><mml:mn mathvariant="normal">90.6</mml:mn><mml:mo>±</mml:mo><mml:mn mathvariant="normal">8</mml:mn></mml:mrow></mml:math></inline-formula></oasis:entry>
       </oasis:row>
       <oasis:row rowsep="1">
         <oasis:entry colname="col1">Mertz</oasis:entry>
         <oasis:entry colname="col2">4.1 (<inline-formula><mml:math id="M230" display="inline"><mml:mo lspace="0mm">-</mml:mo></mml:math></inline-formula>)</oasis:entry>
         <oasis:entry colname="col3">2.6 (<inline-formula><mml:math id="M231" display="inline"><mml:mo lspace="0mm">-</mml:mo></mml:math></inline-formula>)</oasis:entry>
         <oasis:entry colname="col4">4.6 (<inline-formula><mml:math id="M232" display="inline"><mml:mo lspace="0mm">-</mml:mo></mml:math></inline-formula>)</oasis:entry>
         <oasis:entry colname="col5"><inline-formula><mml:math id="M233" display="inline"><mml:mrow><mml:mn mathvariant="normal">7.9</mml:mn><mml:mo>±</mml:mo><mml:mn mathvariant="normal">3</mml:mn></mml:mrow></mml:math></inline-formula></oasis:entry>
       </oasis:row>
       <oasis:row rowsep="1">
         <oasis:entry namest="col1" nameend="col5">5. Ross Sea </oasis:entry>
       </oasis:row>
       <oasis:row>
         <oasis:entry colname="col1">Ross</oasis:entry>
         <oasis:entry colname="col2">53.8</oasis:entry>
         <oasis:entry colname="col3">95.1 (<inline-formula><mml:math id="M234" display="inline"><mml:mo lspace="0mm">+</mml:mo></mml:math></inline-formula>)</oasis:entry>
         <oasis:entry colname="col4">112.0 (<inline-formula><mml:math id="M235" display="inline"><mml:mo lspace="0mm">+</mml:mo></mml:math></inline-formula>)</oasis:entry>
         <oasis:entry colname="col5"><inline-formula><mml:math id="M236" display="inline"><mml:mrow><mml:mn mathvariant="normal">47.7</mml:mn><mml:mo>±</mml:mo><mml:mn mathvariant="normal">34</mml:mn></mml:mrow></mml:math></inline-formula></oasis:entry>
       </oasis:row>
       <oasis:row>
         <oasis:entry colname="col1">Sulzberger</oasis:entry>
         <oasis:entry colname="col2">14.0 (<inline-formula><mml:math id="M237" display="inline"><mml:mo lspace="0mm">-</mml:mo></mml:math></inline-formula>)</oasis:entry>
         <oasis:entry colname="col3">6.5 (<inline-formula><mml:math id="M238" display="inline"><mml:mo lspace="0mm">-</mml:mo></mml:math></inline-formula>)</oasis:entry>
         <oasis:entry colname="col4">9.2 (<inline-formula><mml:math id="M239" display="inline"><mml:mo lspace="0mm">-</mml:mo></mml:math></inline-formula>)</oasis:entry>
         <oasis:entry colname="col5"><inline-formula><mml:math id="M240" display="inline"><mml:mrow><mml:mn mathvariant="normal">18.2</mml:mn><mml:mo>±</mml:mo><mml:mn mathvariant="normal">3</mml:mn></mml:mrow></mml:math></inline-formula></oasis:entry>
       </oasis:row>
       <oasis:row rowsep="1">
         <oasis:entry colname="col1">Nickerson</oasis:entry>
         <oasis:entry colname="col2">5.3</oasis:entry>
         <oasis:entry colname="col3">2.0 (<inline-formula><mml:math id="M241" display="inline"><mml:mo lspace="0mm">-</mml:mo></mml:math></inline-formula>)</oasis:entry>
         <oasis:entry colname="col4">4.1</oasis:entry>
         <oasis:entry colname="col5"><inline-formula><mml:math id="M242" display="inline"><mml:mrow><mml:mn mathvariant="normal">4.2</mml:mn><mml:mo>±</mml:mo><mml:mn mathvariant="normal">2</mml:mn></mml:mrow></mml:math></inline-formula></oasis:entry>
       </oasis:row>
       <oasis:row rowsep="1">
         <oasis:entry namest="col1" nameend="col5">6. Amundsen Sea </oasis:entry>
       </oasis:row>
       <oasis:row>
         <oasis:entry colname="col1">Getz</oasis:entry>
         <oasis:entry colname="col2">88.1 (<inline-formula><mml:math id="M243" display="inline"><mml:mo lspace="0mm">-</mml:mo></mml:math></inline-formula>)</oasis:entry>
         <oasis:entry colname="col3">21.3 (<inline-formula><mml:math id="M244" display="inline"><mml:mo lspace="0mm">-</mml:mo></mml:math></inline-formula>)</oasis:entry>
         <oasis:entry colname="col4">30.6 (<inline-formula><mml:math id="M245" display="inline"><mml:mo lspace="0mm">-</mml:mo></mml:math></inline-formula>)</oasis:entry>
         <oasis:entry colname="col5"><inline-formula><mml:math id="M246" display="inline"><mml:mrow><mml:mn mathvariant="normal">144.9</mml:mn><mml:mo>±</mml:mo><mml:mn mathvariant="normal">14</mml:mn></mml:mrow></mml:math></inline-formula></oasis:entry>
       </oasis:row>
       <oasis:row>
         <oasis:entry colname="col1">Dotson</oasis:entry>
         <oasis:entry colname="col2">9.1 (<inline-formula><mml:math id="M247" display="inline"><mml:mo lspace="0mm">-</mml:mo></mml:math></inline-formula>)</oasis:entry>
         <oasis:entry colname="col3">1.6 (<inline-formula><mml:math id="M248" display="inline"><mml:mo lspace="0mm">-</mml:mo></mml:math></inline-formula>)</oasis:entry>
         <oasis:entry colname="col4">3.7 (<inline-formula><mml:math id="M249" display="inline"><mml:mo lspace="0mm">-</mml:mo></mml:math></inline-formula>)</oasis:entry>
         <oasis:entry colname="col5"><inline-formula><mml:math id="M250" display="inline"><mml:mrow><mml:mn mathvariant="normal">45.2</mml:mn><mml:mo>±</mml:mo><mml:mn mathvariant="normal">4</mml:mn></mml:mrow></mml:math></inline-formula></oasis:entry>
       </oasis:row>
       <oasis:row>
         <oasis:entry colname="col1">Thwaites</oasis:entry>
         <oasis:entry colname="col2">7.4 (<inline-formula><mml:math id="M251" display="inline"><mml:mo lspace="0mm">-</mml:mo></mml:math></inline-formula>)</oasis:entry>
         <oasis:entry colname="col3">2.5 (<inline-formula><mml:math id="M252" display="inline"><mml:mo lspace="0mm">-</mml:mo></mml:math></inline-formula>)</oasis:entry>
         <oasis:entry colname="col4">5.9 (<inline-formula><mml:math id="M253" display="inline"><mml:mo lspace="0mm">-</mml:mo></mml:math></inline-formula>)</oasis:entry>
         <oasis:entry colname="col5"><inline-formula><mml:math id="M254" display="inline"><mml:mrow><mml:mn mathvariant="normal">97.5</mml:mn><mml:mo>±</mml:mo><mml:mn mathvariant="normal">7</mml:mn></mml:mrow></mml:math></inline-formula></oasis:entry>
       </oasis:row>
       <oasis:row rowsep="1">
         <oasis:entry colname="col1">Pine Island</oasis:entry>
         <oasis:entry colname="col2">20.5 (<inline-formula><mml:math id="M255" display="inline"><mml:mo lspace="0mm">-</mml:mo></mml:math></inline-formula>)</oasis:entry>
         <oasis:entry colname="col3">1.9 (<inline-formula><mml:math id="M256" display="inline"><mml:mo lspace="0mm">-</mml:mo></mml:math></inline-formula>)</oasis:entry>
         <oasis:entry colname="col4">9.5 (<inline-formula><mml:math id="M257" display="inline"><mml:mo lspace="0mm">-</mml:mo></mml:math></inline-formula>)</oasis:entry>
         <oasis:entry colname="col5"><inline-formula><mml:math id="M258" display="inline"><mml:mrow><mml:mn mathvariant="normal">101.2</mml:mn><mml:mo>±</mml:mo><mml:mn mathvariant="normal">8</mml:mn></mml:mrow></mml:math></inline-formula></oasis:entry>
       </oasis:row>
       <oasis:row rowsep="1">
         <oasis:entry namest="col1" nameend="col5">7. Bellingshausen Sea </oasis:entry>
       </oasis:row>
       <oasis:row>
         <oasis:entry colname="col1">Abbot</oasis:entry>
         <oasis:entry colname="col2">25.0 (<inline-formula><mml:math id="M259" display="inline"><mml:mo lspace="0mm">-</mml:mo></mml:math></inline-formula>)</oasis:entry>
         <oasis:entry colname="col3">21.9 (<inline-formula><mml:math id="M260" display="inline"><mml:mo lspace="0mm">-</mml:mo></mml:math></inline-formula>)</oasis:entry>
         <oasis:entry colname="col4">36.3</oasis:entry>
         <oasis:entry colname="col5"><inline-formula><mml:math id="M261" display="inline"><mml:mrow><mml:mn mathvariant="normal">51.8</mml:mn><mml:mo>±</mml:mo><mml:mn mathvariant="normal">19</mml:mn></mml:mrow></mml:math></inline-formula></oasis:entry>
       </oasis:row>
       <oasis:row>
         <oasis:entry colname="col1">Stange</oasis:entry>
         <oasis:entry colname="col2">6.1 (<inline-formula><mml:math id="M262" display="inline"><mml:mo lspace="0mm">-</mml:mo></mml:math></inline-formula>)</oasis:entry>
         <oasis:entry colname="col3">5.1 (<inline-formula><mml:math id="M263" display="inline"><mml:mo lspace="0mm">-</mml:mo></mml:math></inline-formula>)</oasis:entry>
         <oasis:entry colname="col4">10.9 (<inline-formula><mml:math id="M264" display="inline"><mml:mo lspace="0mm">-</mml:mo></mml:math></inline-formula>)</oasis:entry>
         <oasis:entry colname="col5"><inline-formula><mml:math id="M265" display="inline"><mml:mrow><mml:mn mathvariant="normal">28.0</mml:mn><mml:mo>±</mml:mo><mml:mn mathvariant="normal">6</mml:mn></mml:mrow></mml:math></inline-formula></oasis:entry>
       </oasis:row>
       <oasis:row>
         <oasis:entry colname="col1">George VI</oasis:entry>
         <oasis:entry colname="col2">48.4 (<inline-formula><mml:math id="M266" display="inline"><mml:mo lspace="0mm">-</mml:mo></mml:math></inline-formula>)</oasis:entry>
         <oasis:entry colname="col3">14.0 (<inline-formula><mml:math id="M267" display="inline"><mml:mo lspace="0mm">-</mml:mo></mml:math></inline-formula>)</oasis:entry>
         <oasis:entry colname="col4">32.5 (<inline-formula><mml:math id="M268" display="inline"><mml:mo lspace="0mm">-</mml:mo></mml:math></inline-formula>)</oasis:entry>
         <oasis:entry colname="col5"><inline-formula><mml:math id="M269" display="inline"><mml:mrow><mml:mn mathvariant="normal">89.0</mml:mn><mml:mo>±</mml:mo><mml:mn mathvariant="normal">17</mml:mn></mml:mrow></mml:math></inline-formula></oasis:entry>
       </oasis:row>
       <oasis:row rowsep="1">
         <oasis:entry colname="col1">Wilkins</oasis:entry>
         <oasis:entry colname="col2">8.1</oasis:entry>
         <oasis:entry colname="col3">8.6</oasis:entry>
         <oasis:entry colname="col4">11.3</oasis:entry>
         <oasis:entry colname="col5"><inline-formula><mml:math id="M270" display="inline"><mml:mrow><mml:mn mathvariant="normal">18.4</mml:mn><mml:mo>±</mml:mo><mml:mn mathvariant="normal">17</mml:mn></mml:mrow></mml:math></inline-formula></oasis:entry>
       </oasis:row>
       <oasis:row rowsep="1">
         <oasis:entry namest="col1" nameend="col5">8. Larsen ice shelves </oasis:entry>
       </oasis:row>
       <oasis:row>
         <oasis:entry colname="col1">Larsen C</oasis:entry>
         <oasis:entry colname="col2">18.2</oasis:entry>
         <oasis:entry colname="col3">35.2</oasis:entry>
         <oasis:entry colname="col4">54.7</oasis:entry>
         <oasis:entry colname="col5"><inline-formula><mml:math id="M271" display="inline"><mml:mrow><mml:mn mathvariant="normal">20.7</mml:mn><mml:mo>±</mml:mo><mml:mn mathvariant="normal">67</mml:mn></mml:mrow></mml:math></inline-formula></oasis:entry>
       </oasis:row>
       <oasis:row rowsep="1">
         <oasis:entry colname="col1">Larsen D</oasis:entry>
         <oasis:entry colname="col2">2.9</oasis:entry>
         <oasis:entry colname="col3">2.1</oasis:entry>
         <oasis:entry colname="col4">3.3</oasis:entry>
         <oasis:entry colname="col5"><inline-formula><mml:math id="M272" display="inline"><mml:mrow><mml:mn mathvariant="normal">1.4</mml:mn><mml:mo>±</mml:mo><mml:mn mathvariant="normal">14</mml:mn></mml:mrow></mml:math></inline-formula></oasis:entry>
       </oasis:row>
       <oasis:row>
         <oasis:entry colname="col1">Total Antarctica</oasis:entry>
         <oasis:entry colname="col2">642 (<inline-formula><mml:math id="M273" display="inline"><mml:mo lspace="0mm">-</mml:mo></mml:math></inline-formula>)</oasis:entry>
         <oasis:entry colname="col3">586 (<inline-formula><mml:math id="M274" display="inline"><mml:mo lspace="0mm">-</mml:mo></mml:math></inline-formula>)</oasis:entry>
         <oasis:entry colname="col4">739 (<inline-formula><mml:math id="M275" display="inline"><mml:mo lspace="0mm">-</mml:mo></mml:math></inline-formula>)</oasis:entry>
         <oasis:entry colname="col5"><inline-formula><mml:math id="M276" display="inline"><mml:mrow><mml:mn mathvariant="normal">1325</mml:mn><mml:mo>±</mml:mo><mml:mn mathvariant="normal">235</mml:mn></mml:mrow></mml:math></inline-formula></oasis:entry>
       </oasis:row>
     </oasis:tbody>
   </oasis:tgroup></oasis:table></table-wrap>

</sec>
</sec>
<sec id="Ch1.S4.SS3">
  <title>Ice-shelf cavities</title>
      <p id="d1e4563">Basal melting of Antarctic ice shelves comprises a substantial source of
freshwater entering the Southern Ocean. <xref ref-type="bibr" rid="bib1.bibx134" id="text.142"/> estimate, based on
observations for the period 2003–2008, that total ice-shelf basal mass loss
occurs at a rate of <inline-formula><mml:math id="M277" display="inline"><mml:mrow><mml:mn mathvariant="normal">1325</mml:mn><mml:mo>±</mml:mo><mml:mn mathvariant="normal">235</mml:mn></mml:mrow></mml:math></inline-formula> Gt yr<inline-formula><mml:math id="M278" display="inline"><mml:msup><mml:mi/><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">1</mml:mn></mml:mrow></mml:msup></mml:math></inline-formula>. This estimate is prone to
errors in the calculation of basal melting at ice-shelf fronts (where
separating basal melting from calving is not straightforward) and relies on
atmospheric reanalyses which in turn have limited observations from which to
downscale. Another observation-based estimate, by <xref ref-type="bibr" rid="bib1.bibx15" id="text.143"/>, is
similar at <inline-formula><mml:math id="M279" display="inline"><mml:mrow><mml:mn mathvariant="normal">1454</mml:mn><mml:mo>±</mml:mo><mml:mn mathvariant="normal">174</mml:mn></mml:mrow></mml:math></inline-formula> Gt yr<inline-formula><mml:math id="M280" display="inline"><mml:msup><mml:mi/><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">1</mml:mn></mml:mrow></mml:msup></mml:math></inline-formula>.</p>

      <?xmltex \floatpos{t}?><fig id="Ch1.F9" specific-use="star"><label>Figure 9</label><caption><p id="d1e4623">Difference between simulated ice-shelf basal mass loss (2002–2016
average) and the central estimate given by <xref ref-type="bibr" rid="bib1.bibx134" id="text.144"/> for each ice
shelf in Table <xref ref-type="table" rid="Ch1.T1"/>, in MetROMS (blue), low-resolution FESOM
(purple), and high-resolution FESOM (green). The uncertainty ranges of
<xref ref-type="bibr" rid="bib1.bibx134" id="text.145"/> are also shown with black error bars. The eight regions
specified in Table <xref ref-type="table" rid="Ch1.T1"/> are labelled as follows: FR indicates
Filchner–Ronne, EWed indicates eastern Weddell region, Am indicates Amery, Aus indicates
Australian sector, RS indicates Ross Sea, AS indicates Amundsen Sea, BS indicates
Bellingshausen Sea, and Lr indicates Larsen ice shelves.</p></caption>
          <?xmltex \igopts{width=341.433071pt}?><graphic xlink:href="https://gmd.copernicus.org/articles/11/1257/2018/gmd-11-1257-2018-f09.png"/>

        </fig>

      <?xmltex \floatpos{t}?><fig id="Ch1.F10" specific-use="star"><label>Figure 10</label><caption><p id="d1e4644">Proportions of different water masses (defined in
Table <xref ref-type="table" rid="Ch1.T2"/>) as percentage volumes in ice-shelf cavities for each
simulation, based on temperature and salinity fields averaged over
2002–2016. Results are shown for the eight regions specified in
Table <xref ref-type="table" rid="Ch1.T1"/>, points 1–7, as well as the total for all Antarctic ice
shelves (Table <xref ref-type="table" rid="Ch1.T1"/>, point 8). All acronyms are the same as in Fig. <xref ref-type="fig" rid="Ch1.F4"/>.</p></caption>
          <?xmltex \igopts{width=398.338583pt}?><graphic xlink:href="https://gmd.copernicus.org/articles/11/1257/2018/gmd-11-1257-2018-f10.png"/>

        </fig>

      <p id="d1e4662">All three model simulations underestimate total ice-shelf basal mass loss
with respect to these observations, roughly by a factor of 2. The simulated
mass loss, averaged over 2002–2016, is 642 Gt yr<inline-formula><mml:math id="M281" display="inline"><mml:msup><mml:mi/><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">1</mml:mn></mml:mrow></mml:msup></mml:math></inline-formula> for MetROMS,
586 Gt yr<inline-formula><mml:math id="M282" display="inline"><mml:msup><mml:mi/><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">1</mml:mn></mml:mrow></mml:msup></mml:math></inline-formula> for low-resolution FESOM, and 739 Gt yr<inline-formula><mml:math id="M283" display="inline"><mml:msup><mml:mi/><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">1</mml:mn></mml:mrow></mml:msup></mml:math></inline-formula> for
high-resolution FESOM. A closer examination of individual ice shelves shows
that the bias in our simulations is a regional phenomenon.
Table <xref ref-type="table" rid="Ch1.T1"/> compares simulated basal mass loss to
<xref ref-type="bibr" rid="bib1.bibx134" id="text.146"/>'s estimates for 25 ice shelves, organised into eight
regions. The model biases are summarised in Fig. <xref ref-type="fig" rid="Ch1.F9"/>, which
plots the difference between the simulated values and <xref ref-type="bibr" rid="bib1.bibx134" id="text.147"/>'s
central estimates, as well as the uncertainty range, for each ice shelf. All
three simulations underestimate mass loss for ice shelves in the Amundsen
Sea, Bellingshausen Sea, and Australian sector. These three regions include
many warm-cavity ice shelves which, despite their small areas, exhibit
substantial basal mass loss in observations. Ice shelves in the remaining
five regions generally show either agreement between our model simulations
and <xref ref-type="bibr" rid="bib1.bibx134" id="text.148"/>'s observations, or an overestimation of mass loss by
the models (with the main exception being MetROMS' underestimation of the
Filchner–Ronne Ice Shelf). The following sections will analyse these eight
regions in more detail.</p>
      <p id="d1e4715">While biases in ice-shelf mass loss are largely region-specific, several
overarching factors are worth mentioning here. First, neither MetROMS nor
FESOM considers the effects of tides. Since the heat and salt transfer
coefficients in both models depend on ocean velocity adjacent to the
ice-shelf base, tidal currents would be expected to increase melt rates in all
ice-shelf cavities. Tides also cause enhanced vertical mixing, which further
influences melt rates <xref ref-type="bibr" rid="bib1.bibx50" id="paren.149"/>. Next, insufficient horizontal
resolution is likely to cause an underestimation of eddy transport of warm
CDW onto the continental shelf; this phenomenon is discussed more in
Sect. <xref ref-type="sec" rid="Ch1.S4.SS3.SSS6"/>. Finally, biases in the ERA-Interim atmospheric
forcing could affect water mass properties and therefore ice-shelf melt
rates; this is difficult to test due to a lack of observations around
Antarctica. Note also that the area of a given ice shelf in model simulations
does not necessarily agree with the area used in <xref ref-type="bibr" rid="bib1.bibx134" id="text.150"/>'s
calculations, particularly for small ice shelves which are not well resolved
by the models. Such disagreements may bias our comparison. However, a
comparison of area-averaged basal melt rates rather than area-integrated
basal mass loss (not shown) shows essentially the same biases. Furthermore, a
comparison with the mass loss estimates of <xref ref-type="bibr" rid="bib1.bibx15" id="text.151"/> yields a
similar pattern of biases.</p>

<?xmltex \floatpos{t}?><table-wrap id="Ch1.T2"><label>Table 2</label><caption><p id="d1e4732">Potential temperature (<inline-formula><mml:math id="M284" display="inline"><mml:mi>T</mml:mi></mml:math></inline-formula>) and salinity (<inline-formula><mml:math id="M285" display="inline"><mml:mi>S</mml:mi></mml:math></inline-formula>) ranges used to
categorise water masses in ice-shelf cavities in Fig. <xref ref-type="fig" rid="Ch1.F10"/>.
<inline-formula><mml:math id="M286" display="inline"><mml:mrow><mml:msub><mml:mi>T</mml:mi><mml:mtext>f</mml:mtext></mml:msub></mml:mrow></mml:math></inline-formula> is the surface freezing point as in Fig. <xref ref-type="fig" rid="Ch1.F4"/>. All
acronyms are the same as in Fig. <xref ref-type="fig" rid="Ch1.F4"/>.</p></caption><oasis:table frame="topbot"><oasis:tgroup cols="3">
     <oasis:colspec colnum="1" colname="col1" align="left"/>
     <oasis:colspec colnum="2" colname="col2" align="left"/>
     <oasis:colspec colnum="3" colname="col3" align="left"/>
     <oasis:thead>
       <oasis:row rowsep="1">
         <oasis:entry colname="col1"/>
         <oasis:entry colname="col2"><inline-formula><mml:math id="M287" display="inline"><mml:mi>T</mml:mi></mml:math></inline-formula> (<inline-formula><mml:math id="M288" display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula>C)</oasis:entry>
         <oasis:entry colname="col3"><inline-formula><mml:math id="M289" display="inline"><mml:mi>S</mml:mi></mml:math></inline-formula> (psu)</oasis:entry>
       </oasis:row>
     </oasis:thead>
     <oasis:tbody>
       <oasis:row>
         <oasis:entry colname="col1">ISW</oasis:entry>
         <oasis:entry colname="col2"><inline-formula><mml:math id="M290" display="inline"><mml:mrow><mml:mi>T</mml:mi><mml:mo>&lt;</mml:mo><mml:msub><mml:mi>T</mml:mi><mml:mtext>f</mml:mtext></mml:msub></mml:mrow></mml:math></inline-formula></oasis:entry>
         <oasis:entry colname="col3"/>
       </oasis:row>
       <oasis:row>
         <oasis:entry colname="col1">AASW</oasis:entry>
         <oasis:entry colname="col2"><inline-formula><mml:math id="M291" display="inline"><mml:mrow><mml:mi>T</mml:mi><mml:mo>≥</mml:mo><mml:msub><mml:mi>T</mml:mi><mml:mtext>f</mml:mtext></mml:msub></mml:mrow></mml:math></inline-formula></oasis:entry>
         <oasis:entry colname="col3"><inline-formula><mml:math id="M292" display="inline"><mml:mrow><mml:mi>S</mml:mi><mml:mo>&lt;</mml:mo><mml:mn mathvariant="normal">34</mml:mn></mml:mrow></mml:math></inline-formula></oasis:entry>
       </oasis:row>
       <oasis:row>
         <oasis:entry colname="col1">LSSW</oasis:entry>
         <oasis:entry colname="col2"><inline-formula><mml:math id="M293" display="inline"><mml:mrow><mml:msub><mml:mi>T</mml:mi><mml:mtext>f</mml:mtext></mml:msub><mml:mo>≤</mml:mo><mml:mi>T</mml:mi><mml:mo>≤</mml:mo><mml:mo>-</mml:mo><mml:mn mathvariant="normal">1.5</mml:mn></mml:mrow></mml:math></inline-formula></oasis:entry>
         <oasis:entry colname="col3"><inline-formula><mml:math id="M294" display="inline"><mml:mrow><mml:mn mathvariant="normal">34</mml:mn><mml:mo>≤</mml:mo><mml:mi>S</mml:mi><mml:mo>&lt;</mml:mo><mml:mn mathvariant="normal">34.5</mml:mn></mml:mrow></mml:math></inline-formula></oasis:entry>
       </oasis:row>
       <oasis:row>
         <oasis:entry colname="col1">HSSW</oasis:entry>
         <oasis:entry colname="col2"><inline-formula><mml:math id="M295" display="inline"><mml:mrow><mml:msub><mml:mi>T</mml:mi><mml:mtext>f</mml:mtext></mml:msub><mml:mo>≤</mml:mo><mml:mi>T</mml:mi><mml:mo>≤</mml:mo><mml:mo>-</mml:mo><mml:mn mathvariant="normal">1.5</mml:mn></mml:mrow></mml:math></inline-formula></oasis:entry>
         <oasis:entry colname="col3"><inline-formula><mml:math id="M296" display="inline"><mml:mrow><mml:mi>S</mml:mi><mml:mo>≥</mml:mo><mml:mn mathvariant="normal">34.5</mml:mn></mml:mrow></mml:math></inline-formula></oasis:entry>
       </oasis:row>
       <oasis:row>
         <oasis:entry colname="col1">MCDW</oasis:entry>
         <oasis:entry colname="col2"><inline-formula><mml:math id="M297" display="inline"><mml:mrow><mml:mi>T</mml:mi><mml:mo>&gt;</mml:mo><mml:mo>-</mml:mo><mml:mn mathvariant="normal">1.5</mml:mn></mml:mrow></mml:math></inline-formula></oasis:entry>
         <oasis:entry colname="col3"><inline-formula><mml:math id="M298" display="inline"><mml:mrow><mml:mi>S</mml:mi><mml:mo>≥</mml:mo><mml:mn mathvariant="normal">34</mml:mn></mml:mrow></mml:math></inline-formula></oasis:entry>
       </oasis:row>
     </oasis:tbody>
   </oasis:tgroup></oasis:table></table-wrap>

      <p id="d1e4995">The average annual minimum in total basal mass loss (calculated over 5-day
averages between 2002 and 2016) is 490 Gt yr<inline-formula><mml:math id="M299" display="inline"><mml:msup><mml:mi/><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">1</mml:mn></mml:mrow></mml:msup></mml:math></inline-formula> for MetROMS,
323 Gt yr<inline-formula><mml:math id="M300" display="inline"><mml:msup><mml:mi/><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">1</mml:mn></mml:mrow></mml:msup></mml:math></inline-formula> for low-resolution FESOM, and 379 Gt yr<inline-formula><mml:math id="M301" display="inline"><mml:msup><mml:mi/><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">1</mml:mn></mml:mrow></mml:msup></mml:math></inline-formula> for
high-resolution FESOM. The corresponding average annual maximum values are
1017, 1589, and 1988 Gt yr<inline-formula><mml:math id="M302" display="inline"><mml:msup><mml:mi/><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">1</mml:mn></mml:mrow></mml:msup></mml:math></inline-formula>, respectively. Note that the seasonal
cycle is larger in FESOM than in MetROMS, which is likely related to greater
summertime melting near ice-shelf fronts. Transport of warm AASW into ice-shelf
cavities is enhanced by FESOM's more significant smoothing of the ice-shelf front, as discussed in the following sections.</p>
      <p id="d1e5046">Interannual variability in ice-shelf melting is relatively small. In all
three simulations, the standard deviation in annually averaged mass loss from
individual ice shelves is typically 10–20 % of their 2002–2016 mean.
Furthermore, the mean and median of the annually averaged values are
typically very similar (within 10 % of each other) which indicates that
the long-term average is not skewed by a few years of unusually high or low
melt. The main exceptions are the Larsen C and D ice shelves in FESOM, which
experience large spikes in mass loss in some summers but not others. This
behaviour is tied to the sea-ice cover, as FESOM occasionally has ice-free
summers along the peninsula, allowing warmer AASW to develop. In MetROMS, the
Shackleton Ice Shelf shows the highest interannual variability in mass loss.
Here, a few cells on the western edge of the ice shelf are undercut by the
Antarctic coastal current, bringing periodic pulses of high melt.</p>
      <?pagebreak page1274?><p id="d1e5050">To aid our intercomparison, we have categorised the water in ice-shelf
cavities into five water masses based on discrete temperature and salinity
bounds, defined in Table <xref ref-type="table" rid="Ch1.T2"/>: ISW, MCDW, HSSW, LSSW, and AASW.
Figure <xref ref-type="fig" rid="Ch1.F10"/> plots the percent volume of each water mass in ice-shelf
cavities for the eight regions specified in Table <xref ref-type="table" rid="Ch1.T1"/> as
well as the total for all Antarctic ice-shelf cavities. These proportions are
based on temperature and salinity fields averaged over 2002–2016 for each
simulation, and neglect the seasonal cycle. AASW, which is mostly a
summertime phenomenon, may therefore be obscured. For Antarctica as a whole
(Fig. <xref ref-type="fig" rid="Ch1.F10"/>i), both FESOM simulations have more MCDW and HSSW, and
less ISW and LSSW, than in MetROMS. These differences are more pronounced in
low-resolution FESOM, while high-resolution FESOM is more similar to MetROMS.
The water mass proportions in each region (Fig. <xref ref-type="fig" rid="Ch1.F10"/>a–h), and
consequently the reasons for the overarching differences between the three
models, will be analysed in the following sections.</p>
<sec id="Ch1.S4.SS3.SSS1">
  <title>Filchner–Ronne Ice Shelf</title>
      <p id="d1e5068">The Filchner–Ronne Ice Shelf (FRIS) is the largest ice shelf in the Weddell
Sea region and the second largest (by area) in Antarctica. However, its melt
rates away from the grounding line are quite low, leading to relatively
modest basal mass<?pagebreak page1275?> loss for its size. For both FESOM simulations, basal mass
loss for FRIS falls within the range of observations given by
<xref ref-type="bibr" rid="bib1.bibx134" id="text.152"/> (Table <xref ref-type="table" rid="Ch1.T1"/>, point 1). MetROMS
significantly underestimates this rate, simulating about half the lower bound
given by <xref ref-type="bibr" rid="bib1.bibx134" id="text.153"/>. Figure <xref ref-type="fig" rid="Ch1.F11"/>a shows the spatial
distribution of this mass loss in the three simulations, with two-dimensional
ice-shelf melt/freeze fields averaged over 2002–2016.</p>

      <?xmltex \floatpos{t}?><fig id="Ch1.F11"><label>Figure 11</label><caption><p id="d1e5083">The Filchner–Ronne Ice Shelf cavity in MetROMS (left),
low-resolution FESOM (middle), and high-resolution FESOM (right). All fields
are averaged over the period 2002–2016. <bold>(a)</bold> Ice-shelf melt rate
(m yr<inline-formula><mml:math id="M303" display="inline"><mml:msup><mml:mi/><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">1</mml:mn></mml:mrow></mml:msup></mml:math></inline-formula>). <bold>(b)</bold> Vertically averaged ocean velocity
(m s<inline-formula><mml:math id="M304" display="inline"><mml:msup><mml:mi/><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">1</mml:mn></mml:mrow></mml:msup></mml:math></inline-formula>), where the colour scale shows magnitude and the arrows show
direction. <bold>(c)</bold> Bottom water temperature (<inline-formula><mml:math id="M305" display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula>C).
<bold>(d)</bold> Bottom water salinity (psu). <bold>(e)</bold> Ice-shelf draft (m) as
seen by each model. In panels <bold>(b)</bold>–<bold>(d)</bold>, the ice-shelf front is contoured in
black. Rn indicates Ronne Ice Shelf, Fi indicates Filchner Ice Shelf, RDp indicates Ronne
Depression, FDp indicates Filchner Depression, and BI indicates Berkner Island.</p></caption>
            <?xmltex \igopts{width=199.169291pt}?><graphic xlink:href="https://gmd.copernicus.org/articles/11/1257/2018/gmd-11-1257-2018-f11.png"/>

          </fig>

      <p id="d1e5147">Circulation patterns in the FRIS cavity have been inferred from a few
sub-ice-shelf observations <xref ref-type="bibr" rid="bib1.bibx117 bib1.bibx116" id="paren.154"/> and feature anticyclonic
flow around Berkner Island, a cyclonic gyre in the Filchner Ice Shelf cavity,
and HSSW inflow into the western Ronne Ice Shelf cavity via the Ronne
Depression. MetROMS displays the anticyclonic flow around Berkner Island
(Fig. <xref ref-type="fig" rid="Ch1.F11"/>b), which is stronger on the western and southern sides,
corresponding with locally increased melt rates. Circulation in the Filchner
Ice Shelf cavity is weak but mostly cyclonic. However, the observed
refreezing immediately east of Berkner Island <xref ref-type="bibr" rid="bib1.bibx79 bib1.bibx134" id="paren.155"/>
is absent.</p>
      <p id="d1e5158">In FESOM, southward flow of HSSW on the western flank of the Filchner
Depression drives a relatively strong anticyclonic gyre in the Filchner
cavity. Melting is therefore apparent on the western side of the Filchner Ice
Shelf front, with refreezing associated with outflow in the east. This
pattern is opposite to observations and may be caused by FESOM's strong HSSW
formation in the Filchner Depression, as evidenced by deep wintertime mixed
layers in Fig. <xref ref-type="fig" rid="Ch1.F3"/>. Melting at the Filchner Ice Shelf front
is stronger in the low-resolution FESOM simulation than the high-resolution simulation,
with vigorous melting immediately east of Berkner Island. This feature is
likely due to the slightly smoother ice-shelf front in the low-resolution
simulation (Fig. <xref ref-type="fig" rid="Ch1.F11"/>e), which allows for greater transport of warm
AASW into the cavity. At the southern coast of Berkner Island, FESOM's
inflowing current splits into two branches, one continuing westward along the
southern edge of the Ronne Ice Shelf cavity and the other turning northward
to continue the cyclonic flow around the island.
<?xmltex \hack{\newpage}?>
In the Ronne Ice Shelf cavity, observations by <xref ref-type="bibr" rid="bib1.bibx134" id="text.156"/> as well as
<xref ref-type="bibr" rid="bib1.bibx79" id="text.157"/> indicate significant areas of refreezing in the interior
combined with melting at the ice-shelf front. MetROMS captures both of these
features. FESOM exhibits refreezing in the interior (albeit weaker than in
MetROMS) but also at the ice-shelf front, with a band of melting between the
two regions. At the western edge of the cavity, a narrow band of refreezing
associated with outflow is present in both MetROMS and FESOM, which agrees
with observations by <xref ref-type="bibr" rid="bib1.bibx118" id="text.158"/>.</p>
      <p id="d1e5178">In MetROMS, the strongest melting occurs in the pockets of deep ice at the
back of the cavity. Compared to the remote sensing observations of
<xref ref-type="bibr" rid="bib1.bibx79" id="text.159"/>,<?pagebreak page1276?> MetROMS overestimates melt rates in these regions of the
Filchner Ice Shelf but underestimates them for the Ronne Ice Shelf. In FESOM,
melting is weaker and more widespread along these grounding line regions,
which more or less agrees with <xref ref-type="bibr" rid="bib1.bibx79" id="text.160"/>'s observations for the
Filchner grounding line, although melt rates in the interior Filchner Ice
Shelf are too high. These grounding line regions are largely bypassed by the
gyre transporting HSSW through the Filchner cavity and therefore remain
cooler and fresher. There may also be a small effect from FESOM's greater
smoothing of the ice-shelf draft, which causes the deepest ice to shoal by
approximately 100 m. All else being equal, shallower ice melts more slowly
due to its increased in situ freezing point. In this situation, the increase
is approximately 0.076 <inline-formula><mml:math id="M306" display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula>C, which is likely to be overwhelmed by
other factors. The pockets of deep ice are better preserved in the
high-resolution FESOM mesh than the low-resolution FESOM mesh, and show no
significant changes in melt rate. Instead, the slightly thinner water column
further inhibits HSSW transport to these grounding line regions, where the
temperature is slightly cooler. This tendency for changes in velocity to
offset changes in the in situ freezing point in the FRIS cavity was
demonstrated by <xref ref-type="bibr" rid="bib1.bibx160" id="text.161"/> for a fully coupled configuration with
an evolving ice-shelf draft.</p>
      <p id="d1e5199">Other FESOM simulations focusing on FRIS exhibit somewhat different melt rate
patterns. For example, <xref ref-type="bibr" rid="bib1.bibx160" id="text.162"/> simulate more vigorous melting
near the grounding line, as well as a larger area of refreezing in the
interior Ronne Ice Shelf which does not quite extend to the ice-shelf front.
However, these simulations used a different atmospheric forcing dataset,
which may influence sea-ice formation patterns and consequently sub-ice-shelf
circulation. Additionally, the mesh used by <?xmltex \hack{\mbox\bgroup}?><xref ref-type="bibr" rid="bib1.bibx160" id="text.163"/><?xmltex \hack{\egroup}?> has
higher resolution at the FRIS grounding line (approximately 1 km), which may allow
for better representation of ocean velocities beneath the deepest ice.</p>
      <p id="d1e5210">With respect to water mass properties, Fig. <xref ref-type="fig" rid="Ch1.F10"/>a reveals that the
FRIS cavity in MetROMS is almost entirely filled with ISW, with a small
contribution from HSSW (<inline-formula><mml:math id="M307" display="inline"><mml:mo lspace="0mm">&lt;</mml:mo></mml:math></inline-formula> 5 %). This dominance of ISW indicates that
it has a relatively long simulated residence time in the FRIS cavity. FESOM's
stronger dense water formation in the Filchner Depression means that HSSW has
a much larger presence (approximately 40 %), indicating more rapid flushing of
the cavity; this effect is slightly lessened at high resolution. As a result,
FESOM has warmer and saltier bottom water than MetROMS in most of the FRIS
cavity (Fig. <xref ref-type="fig" rid="Ch1.F11"/>c and d). However, observations south of Berkner
Island <xref ref-type="bibr" rid="bib1.bibx116" id="paren.164"/> reveal bottom water temperatures around
<inline-formula><mml:math id="M308" display="inline"><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">2.2</mml:mn></mml:mrow></mml:math></inline-formula> <inline-formula><mml:math id="M309" display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula>C, which implies that both MetROMS and FESOM are too warm.
The fact that MetROMS still underestimates total basal mass loss, despite a
warm bias, suggests that its relatively weak circulation and long residence
time are to blame for low melt rates.</p>
      <p id="d1e5247">The exceptionally large tides of the Weddell Sea <xref ref-type="bibr" rid="bib1.bibx28" id="paren.165"/>, with
tidal velocities up to 1 m s<inline-formula><mml:math id="M310" display="inline"><mml:msup><mml:mi/><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">1</mml:mn></mml:mrow></mml:msup></mml:math></inline-formula> <?xmltex \hack{\mbox\bgroup}?><xref ref-type="bibr" rid="bib1.bibx138" id="paren.166"/><?xmltex \hack{\egroup}?>, are
understood to have a strong impact on FRIS melt rates. Indeed, inclusion of
tides in one regional model of the southern Weddell Sea caused basal mass
loss from FRIS to approximately double <xref ref-type="bibr" rid="bib1.bibx96" id="paren.167"/>. Another model
<xref ref-type="bibr" rid="bib1.bibx111" id="paren.168"/> found no significant change in total mass loss but
rather an amplification of existing melt and freeze patterns. As mentioned
previously, neither MetROMS nor FESOM considers these effects.</p>

      <?xmltex \floatpos{t}?><fig id="Ch1.F12"><label>Figure 12</label><caption><p id="d1e5279"><bold>(a, b)</bold> As Fig. <xref ref-type="fig" rid="Ch1.F11"/>a and b for the eastern
Weddell ice-shelf cavities. <bold>(c)</bold> Temperature (<inline-formula><mml:math id="M311" display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula>C) and
<bold>(d)</bold> salinity (psu) interpolated to <inline-formula><mml:math id="M312" display="inline"><mml:mn mathvariant="normal">1</mml:mn></mml:math></inline-formula><inline-formula><mml:math id="M313" display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula> W, through the
Fimbul Ice Shelf. Br indicates Brunt Ice Shelf, RiL indicates Riiser–Larsen Ice
Shelf, Ek indicates Ekström Ice Shelf, Je indicates Jelbart Ice Shelf,
Fm indicates Fimbul Ice Shelf, Nv indicates Nivl Ice Shelf, Lz indicates Lazarev Ice
Shelf, Bo indicates Borchgrevink Ice Shelf, Bd indicates Baudouin Ice Shelf,
and PH indicates Prince Harald Ice Shelf.</p></caption>
            <?xmltex \igopts{width=241.848425pt}?><graphic xlink:href="https://gmd.copernicus.org/articles/11/1257/2018/gmd-11-1257-2018-f12.png"/>

          </fig>

</sec>
<sec id="Ch1.S4.SS3.SSS2">
  <title>Eastern Weddell region</title>
      <p id="d1e5329">We define the eastern Weddell region as the line of ice shelves east of FRIS,
stretching along the coastline of Queen Maud Land from the Brunt Ice Shelf
in the west to the Prince Harald Ice Shelf in the east. Compared with the
values given by <xref ref-type="bibr" rid="bib1.bibx134" id="text.169"/> (Table <xref ref-type="table" rid="Ch1.T1"/>, point 2), all
three simulations overestimate basal mass loss from the combined Brunt and
Riiser–Larsen ice shelves as well as the Prince Harald Ice Shelf. Both FESOM
simulations also overestimate mass loss from the combined Fimbul, Jelbart,
and Ekström ice shelves. Simulated mass loss for the other ice shelves in the
eastern Weddell region generally falls within the observational<?pagebreak page1277?> estimates.
However, some of the ice shelves are so small that they are barely resolved
by the MetROMS grid or the low-resolution FESOM mesh, particularly the Nivl,
Lazarev, and Prince Harald ice shelves.</p>
      <p id="d1e5337">A notable feature of the eastern Weddell region is an overhang of some
ice-shelf fronts past the continental shelf break, which in our simulations
allows the Antarctic coastal current to undercut the ice shelf
(Fig. <xref ref-type="fig" rid="Ch1.F12"/>b). This process is particularly strong in MetROMS,
where increased velocity corresponds with vigorous melting at the fronts of
the Brunt, Riiser–Larsen, and Fimbul ice shelves, with weaker melting or
refreezing further back in the cavities (Fig. <xref ref-type="fig" rid="Ch1.F12"/>a). This
pattern more or less agrees with observations by <xref ref-type="bibr" rid="bib1.bibx88" id="text.170"/> for the
Fimbul Ice Shelf. FESOM has a weaker coastal current than MetROMS in this
region, and melting is less concentrated at the ice-shelf fronts. In general,
melting is stronger in the high-resolution FESOM simulation than the
low-resolution FESOM simulation.</p>
      <p id="d1e5347">Observations of eastern Weddell ice-shelf cavities are scarce, but mooring
data do exist for the Fimbul Ice Shelf <xref ref-type="bibr" rid="bib1.bibx54" id="paren.171"/>. It reveals a
generally cold cavity with temperatures around <inline-formula><mml:math id="M314" display="inline"><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">1.9</mml:mn></mml:mrow></mml:math></inline-formula> <inline-formula><mml:math id="M315" display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula>C and
salinities around 34.3 psu, with occasional intrusions of warmer, saltier
MCDW at depth, and seasonal melting at the ice-shelf front due to
solar-warmed AASW. For comparison, Fig. <xref ref-type="fig" rid="Ch1.F12"/>c
and d plot meridional slices of simulated temperature and salinity through
the Fimbul Ice Shelf cavity at <inline-formula><mml:math id="M316" display="inline"><mml:mn mathvariant="normal">1</mml:mn></mml:math></inline-formula><inline-formula><mml:math id="M317" display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula> W. All three of our simulations
show a warmer and (in the case of FESOM) saltier cavity than seen in the
range of mooring readings (see Fig. 2 of <xref ref-type="bibr" rid="bib1.bibx54" id="altparen.172"/>). It
appears that MCDW is mixing too readily into the cavity, particularly in
FESOM where the continental shelf break is more gently sloping due to
smoothing of the bathymetry (Fig. <xref ref-type="fig" rid="Ch1.F12"/>c and d, first column
versus second and third columns). The high-resolution FESOM mesh has a
slightly steeper continental shelf break than the low-resolution mesh, and
the cavity is slightly cooler and fresher, indicating that oversmoothing of
the continental shelf break may be a contributing factor in the increased
transport of MCDW.</p>
      <p id="d1e5395">For the eastern Weddell region as a whole, FESOM has proportionally more MCDW
in ice-shelf cavities than MetROMS, as evidenced in Fig. <xref ref-type="fig" rid="Ch1.F10"/>b. As
for the Fimbul, this effect is slightly lessened at high resolution. The
eastern Weddell cavities in MetROMS are more dominated by LSSW (approximately
50 %) with a small amount of ISW (<inline-formula><mml:math id="M318" display="inline"><mml:mo lspace="0mm">&lt;</mml:mo></mml:math></inline-formula> 5 %). Traces of AASW are
present in both FESOM simulations but not in MetROMS, although it may still
exist on a seasonal basis.</p>
</sec>
<sec id="Ch1.S4.SS3.SSS3">
  <title>Amery Ice Shelf</title>
      <p id="d1e5413">All three model simulations overestimate basal mass loss from the Amery Ice
Shelf by about 50 % above the upper bound given by <xref ref-type="bibr" rid="bib1.bibx134" id="text.173"/>
for MetROMS and about 20 % for both FESOM simulations
(Table <xref ref-type="table" rid="Ch1.T1"/>, point 3). However, the sources of this bias are
quite different between the two models.</p>
      <p id="d1e5421">In MetROMS, the majority of melting occurs near the grounding line
(Fig. <xref ref-type="fig" rid="Ch1.F13"/>a), which has one of the deepest ice-shelf drafts in
Antarctica (<inline-formula><mml:math id="M319" display="inline"><mml:mrow><mml:mo>&gt;</mml:mo><mml:mn mathvariant="normal">2000</mml:mn></mml:mrow></mml:math></inline-formula> m) <xref ref-type="bibr" rid="bib1.bibx34" id="paren.174"/>. A cyclonic circulation
pattern is apparent (Fig. <xref ref-type="fig" rid="Ch1.F13"/>b), with refreezing along the
western side of the cavity. This spatial distribution of melting and freezing
agrees with observations <xref ref-type="bibr" rid="bib1.bibx174 bib1.bibx134" id="paren.175"/> and with previous
modelling <xref ref-type="bibr" rid="bib1.bibx35" id="paren.176"/>, but the melting at the grounding line
appears to be too strong (<inline-formula><mml:math id="M320" display="inline"><mml:mo lspace="0mm">&gt;</mml:mo></mml:math></inline-formula> 40 m yr<inline-formula><mml:math id="M321" display="inline"><mml:msup><mml:mi/><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">1</mml:mn></mml:mrow></mml:msup></mml:math></inline-formula>) and the refreezing too weak
and over an insufficient area. These two biases combine to cause the
overestimation of total basal mass loss simulated by MetROMS. It is possible
that both biases could be addressed with an explicit frazil ice
parameterisation with multiple size classes, as <xref ref-type="bibr" rid="bib1.bibx35" id="text.177"/>
showed that including such a parameterisation in ROMS both reduced melting at
the grounding line and increased refreezing on the western side of the
cavity. Additionally, the back of the Amery Ice Shelf cavity is very steep
and not well resolved by MetROMS, which could lead to pressure gradient
errors causing excessive melt.</p>
      <p id="d1e5470">FESOM exhibits much weaker melting than MetROMS at the back of the cavity,
even though bottom water temperatures are at least as warm
(Fig. <xref ref-type="fig" rid="Ch1.F13"/>c). We attribute this discrepancy at least partly to
shoaling of the ice-shelf draft from oversmoothing in our FESOM setups, which
raises the in situ freezing point and therefore reduces melting. The Amery
draft is so steep that in order for a FESOM mesh to preserve the <inline-formula><mml:math id="M322" display="inline"><mml:mo>&gt;</mml:mo></mml:math></inline-formula> 2000 m
deep ice near the grounding line, resolution of 1.5 km or finer is required
throughout the entire cavity. However, experimentation with such a mesh
revealed severe time step limitations: even a 1 min time step was prone to
numerical instabilities, compared to the 10 and 9 min time steps which are
stable for the two FESOM meshes we present here. Due to the computational
expense, we chose not to pursue simulations with this experimental mesh.
Nonetheless, some improvement can be seen between the low-resolution mesh and
the high-resolution mesh (Fig. <xref ref-type="fig" rid="Ch1.F13"/>e), in which the deep ice has
shoaled by approximately 1000 and 800 m, respectively, compared to the source
topography. The corresponding increases in the in situ freezing point are
0.76 and <inline-formula><mml:math id="M323" display="inline"><mml:mn mathvariant="normal">0.61</mml:mn></mml:math></inline-formula> <inline-formula><mml:math id="M324" display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula>C. In the high-resolution simulation, melt rates
near the grounding line exceed 10 m yr<inline-formula><mml:math id="M325" display="inline"><mml:msup><mml:mi/><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">1</mml:mn></mml:mrow></mml:msup></mml:math></inline-formula> on the eastern flank, which
is approximately double that of the low-resolution simulation. Such a strong
response is not due to the in situ freezing point alone, which is only
modestly different between the two simulations. Increased velocities near the
back of the cavity, possibly due to the steeper ice draft or better resolved
currents, also have an effect.</p>
      <p id="d1e5513">In both FESOM simulations, significant melting also occurs near the ice-shelf
front and throughout the outer third of the ice shelf, at a higher rate (up
to 5 m yr<inline-formula><mml:math id="M326" display="inline"><mml:msup><mml:mi/><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">1</mml:mn></mml:mrow></mml:msup></mml:math></inline-formula>) than in MetROMS. This melting is somewhat lessened at
high resolution, offsetting the increased melt rates at the grounding<?pagebreak page1278?> line.
It occurs primarily in summer, leading to a large seasonal cycle in total
basal mass loss for the entire ice shelf. Average annual minimums and
maximums (calculated over 5-day averages between 2002 and 2016) are 27 and
451 Gt yr<inline-formula><mml:math id="M327" display="inline"><mml:msup><mml:mi/><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">1</mml:mn></mml:mrow></mml:msup></mml:math></inline-formula> for low-resolution FESOM, and 30 and 309 Gt yr<inline-formula><mml:math id="M328" display="inline"><mml:msup><mml:mi/><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">1</mml:mn></mml:mrow></mml:msup></mml:math></inline-formula>
for high-resolution FESOM. In comparison, MetROMS has an average annual
minimum of 78 Gt yr<inline-formula><mml:math id="M329" display="inline"><mml:msup><mml:mi/><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">1</mml:mn></mml:mrow></mml:msup></mml:math></inline-formula> and maximum of 119 Gt yr<inline-formula><mml:math id="M330" display="inline"><mml:msup><mml:mi/><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">1</mml:mn></mml:mrow></mml:msup></mml:math></inline-formula>. The
seasonality and resolution dependence of melting in the outer third of the
cavity suggest that it is driven by warm AASW subducting beneath an
oversmoothed ice-shelf front. Circulation in FESOM is predominantly
anticyclonic, with no significant areas of refreezing. This reversed
circulation shows little sensitivity to the improved cavity geometry at high
resolution, indicating that it is more likely driven by hydrography, as for
the Filchner Ice Shelf (Sect. <xref ref-type="sec" rid="Ch1.S4.SS3.SSS1"/>).</p>
      <p id="d1e5580">Bottom water in FESOM is warmer and saltier than in MetROMS throughout the
cavity (Fig. <xref ref-type="fig" rid="Ch1.F13"/>c and d); these differences are lessened at
higher resolution. Figure <xref ref-type="fig" rid="Ch1.F10"/>c indicates that the Amery cavity is
dominated by LSSW in both MetROMS and high-resolution FESOM, with smaller
contributions from ISW (5–20 %) and MCDW (<inline-formula><mml:math id="M331" display="inline"><mml:mo lspace="0mm">&lt;</mml:mo></mml:math></inline-formula> 5 %). Low-resolution
FESOM has significantly more MCDW in the cavity (approximately 40 %), which may
be tied to oversmoothing of the continental shelf break as for the Fimbul Ice
Shelf (Sect. <xref ref-type="sec" rid="Ch1.S4.SS3.SSS2"/>).</p>

      <?xmltex \floatpos{t}?><fig id="Ch1.F13"><label>Figure 13</label><caption><p id="d1e5598">As Fig. <xref ref-type="fig" rid="Ch1.F11"/>, for the Amery Ice Shelf cavity.
PB indicates Prydz Bay.</p></caption>
            <?xmltex \igopts{width=241.848425pt}?><graphic xlink:href="https://gmd.copernicus.org/articles/11/1257/2018/gmd-11-1257-2018-f13.png"/>

          </fig>

</sec>
<sec id="Ch1.S4.SS3.SSS4">
  <title>Australian sector</title>
      <p id="d1e5615">Travelling east from the Amery, the remainder of the Australian sector of
Antarctica contains numerous small ice shelves along the coast of Wilkes
Land, including the West, Shackleton, Totten, and Mertz ice shelves.
<xref ref-type="bibr" rid="bib1.bibx134" id="text.178"/> estimate relatively high melt rates for these ice shelves,
which all three of our simulations fail to capture
(Table <xref ref-type="table" rid="Ch1.T1"/>, point 4).</p>
      <p id="d1e5623">There are several potential reasons for this consistent underestimation.
First, the observed production of HSSW which drives the majority of melting
for cold-cavity ice shelves <xref ref-type="bibr" rid="bib1.bibx71" id="paren.179"/> is enhanced in this region by
many small polynyas which are kept open by grounded icebergs
<xref ref-type="bibr" rid="bib1.bibx84 bib1.bibx48" id="paren.180"/> and coastline geometry <xref ref-type="bibr" rid="bib1.bibx156" id="paren.181"/>.
Neither MetROMS nor FESOM considers grounded icebergs in the configurations
used here, which may explain why sea-ice production is underestimated in this
region (Sect. <xref ref-type="sec" rid="Ch1.S4.SS2.SSS3"/>), implying a lack of HSSW. In fact,
Fig. <xref ref-type="fig" rid="Ch1.F10"/>d reveals no year-round presence of HSSW in Australian
sector ice-shelf cavities for any of the three simulations. The dominant
water mass from sea-ice formation is instead LSSW, which is fresher and less
dense.</p>
      <p id="d1e5639">Whether enhanced production of HSSW would lead to a decrease or increase in
basal melting depends on the local hydrography. For example, flooding a
cavity with relatively cold HSSW formed in coastal polynyas could prevent any
nearby MCDW from accessing the cavity, in which case stronger polynyas would
lead to decreased melt rates <xref ref-type="bibr" rid="bib1.bibx9 bib1.bibx80 bib1.bibx48" id="paren.182"/>.
Indeed, MCDW has recently been observed in front of the Totten Ice Shelf
<xref ref-type="bibr" rid="bib1.bibx44 bib1.bibx137 bib1.bibx147" id="paren.183"/> which could be a factor in its
considerable melt rate. However, MCDW can only access the cavity through a
trough in the continental shelf which was previously unknown and therefore
not included in RTopo-1.05 <xref ref-type="bibr" rid="bib1.bibx44" id="paren.184"/>. Even if this trough was
included in the bathymetry datasets used by MetROMS and FESOM, the models
would likely not resolve such a small-scale feature without increased
resolution. As seen in Fig. <xref ref-type="fig" rid="Ch1.F10"/>d, a significant amount of MCDW is
still present in Australian sector ice-shelf cavities for all three
simulations (approximately 20 % in MetROMS, 85 % in low-res FESOM, and
75 % in high-res FESOM). However, the fairly wide temperature range of
water masses we consider to be MCDW (Table <xref ref-type="table" rid="Ch1.T2"/>) means that the
degree of modification is important for ice-shelf melting.</p>
      <p id="d1e5655">Simulated ice-shelf melt rates are shown for the Australian sector in
Fig. <xref ref-type="fig" rid="Ch1.F14"/>a. Melt rates in MetROMS are generally more
concentrated at the ice-shelf front, and in FESOM more uniform throughout the
cavities. As suggested by FESOM's greater proportion of MCDW in
Fig. <xref ref-type="fig" rid="Ch1.F10"/>d,<?pagebreak page1279?> FESOM has slightly warmer and (in the case of
low-resolution FESOM) saltier bottom water in most ice-shelf cavities
(Fig. <xref ref-type="fig" rid="Ch1.F14"/>b and c). This is despite the fact that bottom
water offshore the continental shelf is warmer in MetROMS (approximately
<inline-formula><mml:math id="M332" display="inline"><mml:mn mathvariant="normal">0.5</mml:mn></mml:math></inline-formula> <inline-formula><mml:math id="M333" display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula>C) than in either FESOM simulation (approximately
<inline-formula><mml:math id="M334" display="inline"><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">0.2</mml:mn></mml:mrow></mml:math></inline-formula> <inline-formula><mml:math id="M335" display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula>C), and indicates that cross-shelf heat transport is
stronger in FESOM. Melt rates are enhanced in the high-resolution FESOM
simulation compared to the low-resolution FESOM simulation, and in most cases are also higher than
MetROMS. This pattern is likely due to stronger circulation as shown for the
Totten Ice Shelf in Fig. <xref ref-type="fig" rid="Ch1.F14"/>d. The MetROMS grid and
low-resolution FESOM mesh cannot adequately resolve circulation in such a
small cavity, which is represented by only a few dozen grid boxes or
triangular elements. Resolution is still less than ideal for the
high-resolution FESOM mesh, but this simulation manages to develop an
anticyclonic gyre beneath the ice shelf. Stronger transport through the
cavity at high resolution, as well as increased transfer coefficients due to
the faster velocity, causes basal mass loss for the Totten Ice Shelf to more
than double compared to the low-resolution FESOM simulation, even though
temperatures are slightly higher in the latter simulation. By comparison, the
ROMS configuration of <xref ref-type="bibr" rid="bib1.bibx48" id="text.185"/> for the Totten region was even
higher resolution (approximately 3 km), and instead simulated a cyclonic gyre.</p>

      <?xmltex \floatpos{t}?><fig id="Ch1.F14"><label>Figure 14</label><caption><p id="d1e5708"><bold>(a)</bold>–<bold>(c)</bold> As Fig. <xref ref-type="fig" rid="Ch1.F11"/>a, c, and d for the Australian
sector ice-shelf cavities. The dashed black lines in panels <bold>(b)</bold>
and <bold>(c)</bold> show the 1500 m isobath, which approximates the continental
shelf break. <bold>(d)</bold> As Fig. <xref ref-type="fig" rid="Ch1.F11"/>b, zoomed into the Totten Ice
Shelf cavity (region outlined in the rightmost panel of panel <bold>a</bold>).
We indicates West Ice Shelf, Sh indicates Shackleton Ice Shelf, Tt indicates Totten Ice
Shelf, MU indicates Moscow University Ice Shelf, and Mz indicates Mertz Ice Shelf.</p></caption>
            <?xmltex \igopts{width=241.848425pt}?><graphic xlink:href="https://gmd.copernicus.org/articles/11/1257/2018/gmd-11-1257-2018-f14.png"/>

          </fig>

</sec>
<sec id="Ch1.S4.SS3.SSS5">
  <title>Ross Sea</title>
      <p id="d1e5745">For the Ross Ice Shelf, the largest in Antarctica (by area), MetROMS falls
within <xref ref-type="bibr" rid="bib1.bibx134" id="text.186"/>'s estimate of basal mass loss while both FESOM
simulations produce an overestimate (Table <xref ref-type="table" rid="Ch1.T1"/>, point 5). In
this region, we also include the nearby Sulzberger Ice Shelf, for which all
three simulations underestimate basal mass loss, and the Nickerson Ice Shelf,
for which MetROMS and high-resolution FESOM agree with <xref ref-type="bibr" rid="bib1.bibx134" id="text.187"/>'s
observations, but low-resolution FESOM produces a slight underestimate.</p>
      <p id="d1e5756">All simulations show predominantly anticyclonic circulation beneath the Ross
Ice Shelf, with inflow in the west and outflow in the east
(Fig. <xref ref-type="fig" rid="Ch1.F15"/>b), in agreement with observations <xref ref-type="bibr" rid="bib1.bibx132" id="paren.188"/>. A
similar system of interconnected gyres is seen in all three simulations,
although circulation near the back of the cavity is stronger in MetROMS.</p>
      <p id="d1e5764">Patterns of refreezing are similar between simulations
(Fig. <xref ref-type="fig" rid="Ch1.F15"/>a), with large areas of refreezing near the back of the
cavity concentrated around Steershead Ice Rise and Crary Ice Rise. FESOM also
displays relatively strong refreezing near the edge of McMurdo Ice Shelf,
particularly in the high-resolution simulation, which has been observed
<xref ref-type="bibr" rid="bib1.bibx87" id="paren.189"/>.</p>
      <p id="d1e5772">The Ross Ice Shelf front is exceptionally steep and requires smoothing in all
three simulations for numerical stability. This smoothing allows relatively
warm AASW to slide under the ice-shelf front, which is visible as tongues of
warm water in meridional temperature slices through 180<inline-formula><mml:math id="M336" display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula> E
(Fig. <xref ref-type="fig" rid="Ch1.F15"/>c). The low-resolution FESOM simulation exhibits this
problem most severely, but it is reduced at high resolution as the ice-shelf
front requires less smoothing. The interior of the cavity is warmer in FESOM
than in MetROMS, and (particularly at low resolution) saltier
(Fig. <xref ref-type="fig" rid="Ch1.F15"/>d). This is due to an increased presence of HSSW, which
is warmer and saltier than the ISW which is more prominent in MetROMS
(Fig. <xref ref-type="fig" rid="Ch1.F10"/>e). The increased HSSW is in turn due to stronger sea-ice
formation in FESOM, as discussed in Sect. <xref ref-type="sec" rid="Ch1.S4.SS2.SSS3"/>. These
differences in hydrography explain why melting in the<?pagebreak page1280?> outer third of the
cavity remains stronger in FESOM than in MetROMS even when the oversmoothing
of the ice-shelf front has been addressed with higher resolution.</p>
      <p id="d1e5793">FESOM also displays stronger melting east of Roosevelt Island, where a small
amount of relatively warm MCDW enters the cavity (not shown). This water mass
has been observed <xref ref-type="bibr" rid="bib1.bibx132 bib1.bibx69" id="paren.190"/> and is likely not a model
artefact. Indeed, <xref ref-type="bibr" rid="bib1.bibx124" id="text.191"/> detect ice-shelf thinning in this region
over the period of 1994–2012.</p>

      <?xmltex \floatpos{t}?><fig id="Ch1.F15"><label>Figure 15</label><caption><p id="d1e5804"><bold>(a, b)</bold> As Fig. <xref ref-type="fig" rid="Ch1.F11"/>a and b for the Ross Sea
ice-shelf cavities. <bold>(c)</bold> Temperature (<inline-formula><mml:math id="M337" display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula>C) and
<bold>(d)</bold> salinity (psu) interpolated to 180<inline-formula><mml:math id="M338" display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula> E, through the Ross
Ice Shelf. Rs indicates Ross Ice Shelf, Sz indicates Sulzberger Ice Shelf,
Nk indicates Nickerson Ice Shelf, McM indicates McMurdo Ice Shelf, RI indicates Roosevelt Island,
CIR indicates Crary Ice Rise, and SH indicates Steershead Ice Rise.</p></caption>
            <?xmltex \igopts{width=241.848425pt}?><graphic xlink:href="https://gmd.copernicus.org/articles/11/1257/2018/gmd-11-1257-2018-f15.png"/>

          </fig>

</sec>
<sec id="Ch1.S4.SS3.SSS6">
  <title>Amundsen Sea</title>
      <p id="d1e5848">Ice shelves in the Amundsen Sea have been the subject of much attention in
recent years due to observed intrusions of unmodified CDW causing rapid
basal melting and grounding line retreat
<xref ref-type="bibr" rid="bib1.bibx56 bib1.bibx72 bib1.bibx77 bib1.bibx170" id="paren.192"/>. The Amundsen Sea has
the highest ice-shelf melt rates of any sector of Antarctica, corresponding
to large basal mass loss coming from a handful of relatively small ice
shelves. However, all three of our simulations severely underestimate these
mass loss values, as shown in Table <xref ref-type="table" rid="Ch1.T1"/>, point 6.</p>
      <p id="d1e5856">There are several likely reasons for this systematic bias, the first and most
well studied being resolution. Intrusion of CDW onto the continental shelf of
the Amundsen Sea depends on small-scale features in the bathymetry, which
cannot be resolved by model grids coarser than approximately 5 km
<xref ref-type="bibr" rid="bib1.bibx113" id="paren.193"/>. Of the three simulations in this intercomparison, only
high-resolution FESOM falls within this threshold. Eddy transport of heat is
also an important factor for cross-shelf CDW exchange. In order to fully
resolve this process, resolutions of 1 km or finer are required
<xref ref-type="bibr" rid="bib1.bibx153" id="paren.194"/>, which none of our simulations have. A partial
representation of eddy transport would be expected from eddy-permitting
simulations (approximately 2–4 km on the Antarctic continental shelf), which
high-resolution FESOM attains in the Amundsen and Bellingshausen seas. The
latitude dependence of the Rossby radius of deformation means that eddies are
much smaller, and therefore more computationally expensive to resolve, in the
polar regions compared to the tropics and the midlatitudes.</p>
      <p id="d1e5865">In FESOM, CDW transport into the Amundsen Sea has been shown to be sensitive
to the depth of transition between <inline-formula><mml:math id="M339" display="inline"><mml:mi mathvariant="italic">σ</mml:mi></mml:math></inline-formula> coordinates and <inline-formula><mml:math id="M340" display="inline"><mml:mi>z</mml:mi></mml:math></inline-formula> coordinates,
with a shallower transition favouring the transport of warmer CDW due to the
better alignment of <inline-formula><mml:math id="M341" display="inline"><mml:mi>z</mml:mi></mml:math></inline-formula> coordinates with isopycnals in this region
<xref ref-type="bibr" rid="bib1.bibx113" id="paren.195"/>. Our simulations have a relatively deep transition of
2500 m, which supports the onshore transport of cooler CDW. Additionally,
the very deep mixed layers in the Amundsen Sea which develop in both FESOM
simulations (Sect. <xref ref-type="sec" rid="Ch1.S4.SS1.SSS2"/> and <xref ref-type="sec" rid="Ch1.S4.SS2.SSS3"/>) produce a
slope front which further blocks CDW from the continental shelf. Any warm
water flowing along the bottom is eroded by the convection of cold LSSW.</p>
      <p id="d1e5897">All three simulations underestimate bottom water temperature throughout the
continental shelf (Fig. <xref ref-type="fig" rid="Ch1.F16"/>b), which has been observed at
approximately <inline-formula><mml:math id="M342" display="inline"><mml:mn mathvariant="normal">1</mml:mn></mml:math></inline-formula> <inline-formula><mml:math id="M343" display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula>C in the Pine Island Ice Shelf cavity
<xref ref-type="bibr" rid="bib1.bibx72 bib1.bibx77" id="paren.196"/> and <inline-formula><mml:math id="M344" display="inline"><mml:mn mathvariant="normal">0.5</mml:mn></mml:math></inline-formula> to <inline-formula><mml:math id="M345" display="inline"><mml:mn mathvariant="normal">1.2</mml:mn></mml:math></inline-formula> <inline-formula><mml:math id="M346" display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula>C throughout the
Amundsen Sea <xref ref-type="bibr" rid="bib1.bibx20" id="paren.197"/>. Temperatures are warmer in MetROMS
(approximately <inline-formula><mml:math id="M347" display="inline"><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">0.2</mml:mn></mml:mrow></mml:math></inline-formula> <inline-formula><mml:math id="M348" display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula>C) than in both FESOM simulations
(approximately <inline-formula><mml:math id="M349" display="inline"><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">1.7</mml:mn></mml:mrow></mml:math></inline-formula> <inline-formula><mml:math id="M350" display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula>C for low resolution and <inline-formula><mml:math id="M351" display="inline"><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">1</mml:mn></mml:mrow></mml:math></inline-formula> <inline-formula><mml:math id="M352" display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula>C for
high resolution), leading to higher melt rates (Fig. <xref ref-type="fig" rid="Ch1.F16"/>a).
All three simulations are dominated by MCDW in this region
(Fig. <xref ref-type="fig" rid="Ch1.F10"/>f); however, FESOM has a larger presence of LSSW than
MetROMS due to the deep mixed layers discussed previously. Consistent with
the reduced cross-shelf transport of CDW, the Antarctic Slope Front current
(not shown) is stronger in FESOM compared to MetROMS, in the Amundsen as well
as Bellingshausen Sea sectors.</p>
      <?pagebreak page1281?><p id="d1e6011">Increasing the resolution in FESOM causes a substantial increase in basal
mass loss for all Amundsen Sea ice shelves (Table <xref ref-type="table" rid="Ch1.T1"/>,
point 6). These changes are most pronounced for the Pine Island Ice Shelf,
where mass loss increases by approximately a factor of 5. A major contributor
to this increased melting is better resolution of troughs in the bathymetry
which provide a pathway for warmer water to access the continental shelf. In
the easternmost trough, near the Thwaites and Pine Island ice shelves, bottom
water temperature increases by approximately 1 <inline-formula><mml:math id="M353" display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula>C as a result of
increased resolution. Salinity also decreases throughout the Amundsen Sea
due to entrainment from additional meltwater (Fig. <xref ref-type="fig" rid="Ch1.F16"/>c).</p>
      <p id="d1e6027">Melting is further enhanced in the high-resolution FESOM simulation due to
stronger circulation, as shown for the Pine Island Ice Shelf in
Fig. <xref ref-type="fig" rid="Ch1.F16"/>d. As for the Totten Ice Shelf
(Sect. <xref ref-type="sec" rid="Ch1.S4.SS3.SSS4"/>), circulation in this small cavity is not
well resolved by MetROMS or low-resolution FESOM, but high-resolution FESOM
develops an anticyclonic gyre which increases melt rates due to larger
friction velocities.</p>

      <?xmltex \floatpos{t}?><fig id="Ch1.F16"><label>Figure 16</label><caption><p id="d1e6036"><bold>(a)</bold>–<bold>(c)</bold> As Fig. <xref ref-type="fig" rid="Ch1.F14"/>a–c for the Amundsen
Sea ice-shelf cavities. The dashed black lines in panels <bold>(b)</bold>
and <bold>(c)</bold> show the 1500 m isobath. <bold>(d)</bold> As
Fig. <xref ref-type="fig" rid="Ch1.F14"/>d, zoomed into the Pine Island Ice Shelf cavity
(region outlined in the rightmost panel of panel <bold>a</bold>). PI indicates Pine Island
Ice Shelf, Th indicates Thwaites Ice Shelf, Do indicates Dotson Ice Shelf,
and Gz indicates Getz Ice Shelf.</p></caption>
            <?xmltex \igopts{width=241.848425pt}?><graphic xlink:href="https://gmd.copernicus.org/articles/11/1257/2018/gmd-11-1257-2018-f16.png"/>

          </fig>

<?xmltex \hack{\newpage}?>
</sec>
<sec id="Ch1.S4.SS3.SSS7">
  <title>Bellingshausen Sea</title>
      <p id="d1e6075">Similarly to the Amundsen Sea, observations of the nearby Bellingshausen Sea
show intrusions of unmodified CDW, particularly beneath George VI Ice Shelf
in the east <xref ref-type="bibr" rid="bib1.bibx76" id="paren.198"/> as well as into Marguerite Bay
<xref ref-type="bibr" rid="bib1.bibx109" id="paren.199"/>. Again, all three model simulations largely fail to
capture the observed CDW intrusions, and as a result underestimate basal mass
loss for the Stange and George VI ice shelves (Table <xref ref-type="table" rid="Ch1.T1"/>,
point 7). MetROMS and low-resolution FESOM also produce an underestimate for
the Abbot Ice Shelf, while high-resolution FESOM agrees with
<xref ref-type="bibr" rid="bib1.bibx134" id="text.200"/>'s observations. All three simulations agree with
observations for the Wilkins Ice Shelf. The increase in melting between the
low-resolution and high-resolution FESOM simulations is substantial, with
basal mass loss more than doubling for the Stange and George VI ice shelves.
However, this still falls below the range of observations.</p>
      <p id="d1e6089">As in the Amundsen Sea, insufficient resolution as well as the depth of the
<inline-formula><mml:math id="M354" display="inline"><mml:mi mathvariant="italic">σ</mml:mi></mml:math></inline-formula>–<inline-formula><mml:math id="M355" display="inline"><mml:mi>z</mml:mi></mml:math></inline-formula> transition in FESOM could be playing a role in the simulated lack
of CDW intrusions. Observations suggest that the dominant mechanism for
cross-shelf CDW transport in this region is eddies shed from the ACC
<xref ref-type="bibr" rid="bib1.bibx100" id="paren.201"/>, which none of our simulations fully resolve. The
sensitivity of Bellingshausen Sea temperatures to model resolution was
further demonstrated by <xref ref-type="bibr" rid="bib1.bibx43" id="text.202"/>, whose ROMS simulations exhibited
greater onshore heat transport at 1.5 km resolution compared to 4 km
resolution due to increased eddy activity. However, the FESOM simulations of
<xref ref-type="bibr" rid="bib1.bibx164" id="text.203"/> had much higher melt rates in the Bellingshausen Sea
than our FESOM simulations, despite similar resolution and the same <inline-formula><mml:math id="M356" display="inline"><mml:mi mathvariant="italic">σ</mml:mi></mml:math></inline-formula>–<inline-formula><mml:math id="M357" display="inline"><mml:mi>z</mml:mi></mml:math></inline-formula>
transition. Since <xref ref-type="bibr" rid="bib1.bibx164" id="text.204"/> forced FESOM with the National Centers
for Environmental Prediction (NCEP)/National Center for Atmospheric Research
(NCAR) atmospheric reanalysis rather than ERA-Interim, differences in atmospheric
forcing could also be a factor.</p>
      <?pagebreak page1282?><p id="d1e6133">MetROMS has generally warmer bottom water in the Bellingshausen Sea than
either FESOM simulation (Fig. <xref ref-type="fig" rid="Ch1.F17"/>b), leading to stronger
melting in some regions (Fig. <xref ref-type="fig" rid="Ch1.F17"/>a). In Ronne Entrance,
extending into the southern end of the channel-shaped George VI Ice Shelf,
MetROMS displays an intrusion of MCDW (approximately <inline-formula><mml:math id="M358" display="inline"><mml:mn mathvariant="normal">0</mml:mn></mml:math></inline-formula> <inline-formula><mml:math id="M359" display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula>C).
High-resolution FESOM also has warmer bottom water here than low-resolution
FESOM (approximately <inline-formula><mml:math id="M360" display="inline"><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">0.25</mml:mn></mml:mrow></mml:math></inline-formula> and <inline-formula><mml:math id="M361" display="inline"><mml:mrow><mml:mo>-</mml:mo><mml:msup><mml:mn mathvariant="normal">0.75</mml:mn><mml:mo>∘</mml:mo></mml:msup></mml:mrow></mml:math></inline-formula>, respectively). For all three
simulations, however, bottom water temperatures in this region and in
Marguerite Bay fall well below the observed <inline-formula><mml:math id="M362" display="inline"><mml:mn mathvariant="normal">1</mml:mn></mml:math></inline-formula> <inline-formula><mml:math id="M363" display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula>C
<xref ref-type="bibr" rid="bib1.bibx76" id="paren.205"/>. As in the Amundsen Sea, the increased meltwater in
high-resolution FESOM compared to low-resolution FESOM leads to lower
salinities in most regions of the Bellingshausen Sea
(Fig. <xref ref-type="fig" rid="Ch1.F17"/>c). MetROMS displays pockets of exceptionally
cold and fresh water (approximately <inline-formula><mml:math id="M364" display="inline"><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">2</mml:mn></mml:mrow></mml:math></inline-formula> <inline-formula><mml:math id="M365" display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula>C and 33.6 psu) in some
ice-shelf cavities, which is the result of poor resolution preventing meltwater
from efficiently circulating out of semi-isolated regions, as discussed in
Sect. <xref ref-type="sec" rid="Ch1.S4.SS1.SSS3"/>. All three simulations show a dominance of MCDW in
the Bellingshausen Sea region (Fig. <xref ref-type="fig" rid="Ch1.F10"/>g), with a smaller presence
of AASW (<inline-formula><mml:math id="M366" display="inline"><mml:mo lspace="0mm">&lt;</mml:mo></mml:math></inline-formula> 5 %).</p>
      <p id="d1e6232">Beneath George VI Ice Shelf (Fig. <xref ref-type="fig" rid="Ch1.F17"/>d), FESOM displays
inflow at the southern end of the channel in Ronne Entrance, and outflow at
the northern end into Marguerite Bay. This circulation agrees with the
direction of net transport inferred from observations <xref ref-type="bibr" rid="bib1.bibx76" id="paren.206"/>.
Some outflow is also apparent in Ronne Entrance, as part of an anticyclonic
gyre, but the rest of the gyre splits off and flows through the cavity. These
circulation patterns are similar in both FESOM simulations, but they are
stronger at high resolution. In MetROMS, south-to-north transport is apparent
along the western edge of the cavity but also north-to-south transport along
the eastern edge. This channel-shaped cavity is narrower and not well
resolved by MetROMS, sometimes only two grid boxes wide.</p>

      <?xmltex \floatpos{t}?><fig id="Ch1.F17"><label>Figure 17</label><caption><p id="d1e6243"><bold>(a)</bold>–<bold>(c)</bold> As Fig. <xref ref-type="fig" rid="Ch1.F14"/>a–c for the
Bellingshausen Sea ice-shelf cavities. The dashed black lines in panels <bold>(b)</bold>
and <bold>(c)</bold> show the 1500 m isobath. <bold>(d)</bold> As
Fig. <xref ref-type="fig" rid="Ch1.F14"/>d, zoomed into the George VI Ice Shelf cavity
(region outlined in the rightmost panel of panel <bold>a</bold>). GVI indicates George VI
Ice Shelf, Wi indicates Wilkins Ice Shelf, St indicates Stange Ice Shelf,
Ab indicates Abbot Ice Shelf, MB indicates Marguerite Bay, and RE indicates Ronne Entrance.</p></caption>
            <?xmltex \igopts{width=241.848425pt}?><graphic xlink:href="https://gmd.copernicus.org/articles/11/1257/2018/gmd-11-1257-2018-f17.png"/>

          </fig>

</sec>
<sec id="Ch1.S4.SS3.SSS8">
  <title>Larsen ice shelves</title>
      <p id="d1e6280">The Larsen ice shelves on the eastern coast of the Antarctic Peninsula are
undergoing a period of dramatic change, with the collapse of Larsen A in 1995
and Larsen B in 2002 <xref ref-type="bibr" rid="bib1.bibx140 bib1.bibx130" id="paren.207"/>, followed by a major calving
of Larsen C in July 2017 <xref ref-type="bibr" rid="bib1.bibx61" id="paren.208"/>. However, these breakup events are
thought to be mainly driven by atmospheric processes rather than basal
melting <xref ref-type="bibr" rid="bib1.bibx129" id="paren.209"/>. Basal processes are actually thought to
stabilise the Larsen C Ice Shelf through the production of marine ice by
refreezing <xref ref-type="bibr" rid="bib1.bibx63" id="paren.210"/>. Additionally, tidal forcing is likely to be
important for the spatial distribution of melting <xref ref-type="bibr" rid="bib1.bibx110" id="paren.211"/>.</p>
      <p id="d1e6298">For all three simulations presented here, simulated basal mass loss for the
Larsen C and Larsen D ice shelves falls within the range of estimates given
by <xref ref-type="bibr" rid="bib1.bibx134" id="text.212"/> (Table <xref ref-type="table" rid="Ch1.T1"/>, point 8). For the larger
Larsen C Ice Shelf, basal mass loss is approximately doubled in the
low-resolution FESOM simulation compared to MetROMS, with a further increase
of approximately 50 % in high-resolution FESOM. Note that the RTopo-1
dataset used to generate the model domains does not include Larsen A or
Larsen B, as it was published following their collapse.</p>
      <p id="d1e6306">Ice-shelf melt rate (a), vertically averaged velocity (b), and bottom water
temperature (c) are shown for all three simulations in Fig. <xref ref-type="fig" rid="Ch1.F18"/>.
The stronger melting in FESOM corresponds to stronger circulation, travelling
from north to south beneath the ice shelf. These differences are due to the
lower summer sea-ice concentration along the peninsula in FESOM compared to
MetROMS, as seen in Fig. <xref ref-type="fig" rid="Ch1.F6"/>. The absence of sea ice means the
ocean surface is less sheltered from wind stress, and develops stronger
seasonal currents which extend into the cavity. Additionally, more AASW is
allowed to develop in FESOM compared to MetROMS, as seen in
Fig. <xref ref-type="fig" rid="Ch1.F10"/>h. By contrast, wintertime area-averaged melt rates are
similar in all three simulations (not shown). The low-resolution FESOM
simulation also has warmer bottom water and more MCDW in the cavities than
either MetROMS or high-resolution FESOM, consistent with the southward
excursion of the southern boundary of the ACC as discussed in
Sect. <xref ref-type="sec" rid="Ch1.S4.SS2.SSS2"/>.</p>
      <p id="d1e6317">Observations suggest extensive areas of refreezing beneath the southern
Larsen C Ice Shelf <xref ref-type="bibr" rid="bib1.bibx63 bib1.bibx134" id="paren.213"/>. Both FESOM simulations
show small regions of refreezing here but not to the same extent as
observations. MetROMS shows no regions of net refreezing at all.<?pagebreak page1283?> This
discrepancy may be due to the lack of small-scale ice-shelf thickness
variability in the smoothed ice-shelf drafts.</p>

      <?xmltex \floatpos{t}?><fig id="Ch1.F18"><label>Figure 18</label><caption><p id="d1e6326"><bold>(a)</bold>–<bold>(c)</bold> As Figs. <xref ref-type="fig" rid="Ch1.F11"/>a, b,
and <xref ref-type="fig" rid="Ch1.F14"/>b for the Larsen ice-shelf cavities. The dashed black
line in panel <bold>(c)</bold> shows the 1500 m isobath. LrC indicates Larsen C Ice Shelf,
LrD indicates Larsen D Ice Shelf, and AAP indicates Antarctic Peninsula.</p></caption>
            <?xmltex \igopts{width=241.848425pt}?><graphic xlink:href="https://gmd.copernicus.org/articles/11/1257/2018/gmd-11-1257-2018-f18.png"/>

          </fig>

</sec>
</sec>
</sec>
<sec id="Ch1.S5">
  <title>Discussion</title>
      <p id="d1e6356">Despite large variations in ocean/ice-shelf interaction, sub-ice-shelf
circulation, and continental shelf processes across different regions of the
Antarctic coastline, several consistent themes have emerged in the simulation
of these regions by the MetROMS and FESOM models. In some cases, these
patterns can be directly linked to model design and can therefore provide
guidance for future model development. In other cases, the interaction of
several different model design choices makes this attribution more difficult.</p>
      <p id="d1e6359">Apparent in nearly all regions is the influence of sea ice on ocean/ice-shelf
interactions. Since many of the prominent Southern Ocean water masses are
governed by sea-ice formation and melt, it is not surprising that simulated
sea ice would have a discernible effect on the processes in ice-shelf
cavities. First, the location and rate of sea-ice formation impacts the
properties of shelf water masses flowing into the ice-shelf cavities. MetROMS
generally exhibits weaker dense water formation than either FESOM simulation,
as evidenced by shallower mixed layers on the Antarctic continental shelf
(Fig. <xref ref-type="fig" rid="Ch1.F3"/>), lower sea-ice production (Fig. <xref ref-type="fig" rid="Ch1.F8"/>),
and a reduced presence of HSSW in ice-shelf cavities (Fig. <xref ref-type="fig" rid="Ch1.F10"/>). For
the Filchner–Ronne and Ross ice shelves, this results in colder, fresher
cavities with generally lower basal melt rates. Depending on the ice shelf,
this may lead to better or worse agreement with observed basal mass loss. In
FESOM, the excessive volume of HSSW produced by sea-ice formation in the
Filchner Depression is likely the culprit for its reversed direction of
transport through the FRIS cavity. In the Amundsen and Bellingshausen seas,
FESOM has unrealistically deep mixed layers driven by sea-ice formation,
which fill the shelf with LSSW and extract the heat of any warm bottom layer.
This contributes to the cold bias in these regions and the underestimation of
basal melt rates. This mechanism is known to be sensitive to the atmospheric
forcing <xref ref-type="bibr" rid="bib1.bibx127 bib1.bibx128 bib1.bibx113" id="paren.214"/>, although there is clearly
some model dependence since these deep mixed layers are not present in
MetROMS. The relationship between sea-ice production, cross-shelf CDW
transport, and ice-shelf basal melting was also investigated by
<xref ref-type="bibr" rid="bib1.bibx161" id="text.215"/> in the context of future climate projections.</p>
      <p id="d1e6374">Simulated summer sea-ice extent is too low in all three simulations, which
exposes a larger area of the ocean to surface heating and drives increased
summertime melting of ice-shelf fronts. This behaviour is exacerbated by
smoothing of the ice-shelf front which allows the warm surface waters to
slide further back into the cavity, as seen for the Ross Ice Shelf in both
models and the Amery Ice Shelf in FESOM. The question of ice-shelf smoothing
will be discussed later in this section, but it is also worth considering the
potential causes of low summer sea ice and how this might be ameliorated. The
fact that both models underestimate the sea-ice minimum by a similar amount,
despite MetROMS' more sophisticated sea-ice physics, suggests that
atmospheric forcing could be part of the problem. Atmospheric GCMs consistently overestimate the amount of solar
radiation reaching the Southern Ocean in summertime due to biases in cloud
cover <xref ref-type="bibr" rid="bib1.bibx165 bib1.bibx175 bib1.bibx4" id="paren.216"/>, and while
atmospheric reanalyses such as ERA-Interim perform better than GCMs, some
biases persist <xref ref-type="bibr" rid="bib1.bibx114" id="paren.217"/>. For future work, it would be worthwhile to
force MetROMS or FESOM with the output of a high-resolution regional
atmospheric model, such as RACMO (Regional Atmospheric Climate Model) which
has previously been used for downscaling ECMWF reanalyses over Antarctica
<xref ref-type="bibr" rid="bib1.bibx92" id="paren.218"/>. The resulting impact (if any) on cloud cover,
radiation, and summer sea-ice extent would be useful to quantify. Another
potential contributor to low summer sea ice in our simulations is the fact
that grounded icebergs are not considered. In addition to enhancing polynya
activity (as described in Sect. <xref ref-type="sec" rid="Ch1.S4.SS3.SSS4"/>), grounded icebergs
increase the extent of fast ice, which is anchored to the coast and better
survives the summer melt <xref ref-type="bibr" rid="bib1.bibx30" id="paren.219"/>.</p>
      <p id="d1e6391">Another model characteristic which influences ice-shelf melt rates is the
degree of smoothing of the bathymetry and the ice-shelf draft. This problem
is more severe in FESOM, which for a given horizontal resolution requires
more smoothing than MetROMS to ensure numerical stability, but it is also
apparent for MetROMS at the Ross Ice Shelf front. Indeed, steep ice-shelf
fronts are some of the most affected regions, as smoothing allows warm AASW
to slide into the ice-shelf cavity and cause seasonal spikes in melt<?pagebreak page1284?> rates.
However, a lack of observations means that the true amplitude of this
seasonal cycle is unknown, and strong summertime melting is not necessarily
unrealistic. This smoothing might also partially compensate for the absence
of tides in both models, given that tidal processes similarly drive water
mass exchanges across the ice-shelf front. Another environment in which
smoothing is problematic is the continental shelf break, which is too gently
sloping in some regions of the FESOM mesh, such as near the Fimbul Ice Shelf
(Fig. <xref ref-type="fig" rid="Ch1.F12"/>c). This allows MCDW to mix up the slope more easily,
and indeed FESOM has a greater presence of MCDW than MetROMS in many
ice-shelf cavities (Fig. <xref ref-type="fig" rid="Ch1.F10"/>). Finally, the steeply sloping Amery Ice
Shelf draft is a particularly challenging feature for the FESOM mesh to
adequately represent without considerable computational expense. In order to
satisfy steepness limitations, the deep ice near the back of the cavity
shoals, particularly in the coarse-resolution mesh, which raises the in situ
freezing point at the ice-shelf base and contributes to reduced melting.</p>
      <p id="d1e6399">All three of these problems improve with increased resolution, as less
smoothing is required. High-resolution FESOM represents steep ice-shelf
drafts more accurately than low-resolution FESOM, leading to fewer AASW
intrusions beneath the Ross and Amery ice shelf fronts, and higher melt rates
near the back of the Amery. The continental shelf break is also better
preserved, and as a result the proportion of MCDW in ice-shelf cavities
decreases in almost every sector (Fig. <xref ref-type="fig" rid="Ch1.F10"/>). Increased resolution is
a straightforward solution to oversmoothing, and FESOM's unstructured mesh is
ideal for targeting problematic regions without requiring increased
resolution everywhere. However, since the maximum stable time step of the
model is a function of the smallest element rather than the average element,
the impact of this approach on computational cost can still be substantial,
as we found when experimenting with resolution in the Amery Ice Shelf cavity.
Therefore, in the future, it may be worthwhile to experiment with different
topographic smoothing methods, which may uncover options to minimise the
trade-off between numerical stability and geometric accuracy. Another
worthwhile approach would be to investigate alternative methods for the
calculation of the horizontal pressure gradient force, such as that of
<xref ref-type="bibr" rid="bib1.bibx24" id="text.220"/>, which may permit more steeply sloping layers and
therefore less topographic smoothing.</p>
      <p id="d1e6407">Another way in which resolution impacts ice-shelf melt rates in FESOM is by
affecting the strength of sub-ice-shelf circulation and the friction velocity
at the ice-shelf base. The high-resolution FESOM simulation is able to
resolve circulation patterns beneath small ice shelves, such as the Totten,
Pine Island, and George VI ice shelves, whereas the low-resolution simulation
shows more or less stagnant cavities with lower melt rates. Furthermore, the
well-studied impact of resolution on Amundsen Sea CDW intrusions
<xref ref-type="bibr" rid="bib1.bibx113" id="paren.221"/> is apparent in our simulations, with bottom water
temperature on the continental shelf increasing by up to <inline-formula><mml:math id="M367" display="inline"><mml:mn mathvariant="normal">1</mml:mn></mml:math></inline-formula> <inline-formula><mml:math id="M368" display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula>C as a
result of better-resolved troughs in the bathymetry. However, both FESOM
simulations still have a cold bias in the Amundsen Sea due to deep
wintertime mixed layers in this region. Convection fills the continental
shelf with LSSW and erodes any CDW intruding into the bottom layer. This
behaviour is driven by sea-ice formation and has been shown to be sensitive
to the atmospheric dataset used <xref ref-type="bibr" rid="bib1.bibx113" id="paren.222"/>. The relatively deep
transition between <inline-formula><mml:math id="M369" display="inline"><mml:mi mathvariant="italic">σ</mml:mi></mml:math></inline-formula> and <inline-formula><mml:math id="M370" display="inline"><mml:mi>z</mml:mi></mml:math></inline-formula> coordinates in our FESOM simulations is also
known to inhibit CDW intrusions. Finally, <xref ref-type="bibr" rid="bib1.bibx154" id="text.223"/> found that
mesoscale eddies are vital to onshore CDW transport. None of our simulations
fully resolve eddies on the Antarctic continental shelf, which would require
resolution of approximately 1 km <xref ref-type="bibr" rid="bib1.bibx153" id="paren.224"/>, although
high-resolution FESOM is eddy-permitting (2–4 km) in the Amundsen and
Bellingshausen seas.</p>
      <p id="d1e6453">FESOM's hybrid vertical coordinate system, with <inline-formula><mml:math id="M371" display="inline"><mml:mi>z</mml:mi></mml:math></inline-formula> coordinates in most of
the domain, is advantageous for the accurate simulation of the interior
Southern Ocean and the ACC. Both FESOM and MetROMS exhibit some erosion of
AAIW during the simulation, but the erosion is more severe in MetROMS due to
spurious diapycnal mixing associated with terrain-following coordinates in
the deep ocean. This degradation of deep water masses could explain the
relatively weak Drake Passage transport simulated by MetROMS. However,
terrain-following coordinates as used by both MetROMS and FESOM have other
benefits on the continental shelf/slope and within ice-shelf cavities. In
particular, they avoid the considerable sensitivity of ice-shelf melt rates
to vertical resolution which is seen in <inline-formula><mml:math id="M372" display="inline"><mml:mi>z</mml:mi></mml:math></inline-formula>-coordinate models of ice-shelf
cavities <xref ref-type="bibr" rid="bib1.bibx47 bib1.bibx101" id="paren.225"/>. On the other hand, <inline-formula><mml:math id="M373" display="inline"><mml:mi>z</mml:mi></mml:math></inline-formula>-coordinate
cavities are not susceptible to pressure gradient errors at the ice-shelf
front and do not introduce time step limitations at the grounding line
<xref ref-type="bibr" rid="bib1.bibx101" id="paren.226"/>.</p>
      <p id="d1e6483">None of our simulations include tides, which would increase ice-shelf melt
rates through enhanced mixing and higher friction velocities. Tidal
amplitudes are particularly large in the Weddell Sea, and the inclusion of
tides has been shown to have a significant effect on simulated FRIS melt
rates <xref ref-type="bibr" rid="bib1.bibx96 bib1.bibx111" id="paren.227"/>. While tides do exist as an option in
the ROMS code, their implementation requires specifying tidal elevation
and/or tidal currents at lateral boundaries from which the tidal signal
propagates throughout the domain. In our MetROMS domain, the northern
boundary at <inline-formula><mml:math id="M374" display="inline"><mml:mn mathvariant="normal">30</mml:mn></mml:math></inline-formula><inline-formula><mml:math id="M375" display="inline"><mml:msup><mml:mi/><mml:mo>∘</mml:mo></mml:msup></mml:math></inline-formula> S requires a strong sponge layer of increased
diffusivity and viscosity to prevent numerical instabilities, which is likely
to modify any tidal signal specified at the boundary. Additionally, the ice-shelf cavities are much further from the boundary than is typical for
simulations investigating tide/ice-shelf interactions, which generally have
smaller domains focusing on a single ice shelf
<xref ref-type="bibr" rid="bib1.bibx123 bib1.bibx111" id="paren.228"/>. We experimented with tides in our MetROMS
setup, but they triggered northern boundary instabilities leading to large
oscillations in ACC transport<?pagebreak page1285?> and subsequently CDW upwelling. In FESOM, tides
have only been applied in regional configurations such as the Ross Sea
<xref ref-type="bibr" rid="bib1.bibx172" id="paren.229"/> using a similar method to ROMS. Nonetheless, future model
development to successfully implement tides in our MetROMS and FESOM domains
would be valuable, and in particular may improve MetROMS' simulated basal
mass loss for FRIS. An alternative approach better suited to large domains
could be to implement the complete lunisolar tides of <xref ref-type="bibr" rid="bib1.bibx158" id="text.230"/>,
which have been successfully incorporated into two global ocean models
<xref ref-type="bibr" rid="bib1.bibx112" id="paren.231"/>.</p>
      <p id="d1e6517">Finally, we must acknowledge that while the three-equation parameterisation
for ice-shelf/ocean interaction is widely used, it relies on turbulent
transfer coefficients which are largely unconstrained by observations. Little
is known about the basal roughness of ice shelves, and thus the spatial and
temporal variations in the corresponding drag coefficient are not generally
included in models. The thermodynamics of the boundary layer at the
ice-shelf/ocean interface are also influenced by the choice of vertical
mixing scheme <xref ref-type="bibr" rid="bib1.bibx49" id="paren.232"/>. A more sophisticated treatment of marine
ice formation, such as the explicit frazil ice models of <xref ref-type="bibr" rid="bib1.bibx150" id="text.233"/>
and <xref ref-type="bibr" rid="bib1.bibx35" id="text.234"/>, may improve simulated patterns of refreezing.
Furthermore, alternative parameterisations of ice-shelf basal melt are being
explored by the community <xref ref-type="bibr" rid="bib1.bibx74 bib1.bibx75" id="paren.235"/>, which may provide
valuable intercomparisons with the three-equation parameterisation in the
future.</p>
</sec>
<sec id="Ch1.S6" sec-type="conclusions">
  <title>Conclusions</title>
      <p id="d1e6538">We have presented the first published model intercomparison of circumpolar
Antarctic ocean/sea-ice/ice-shelf interactions over a realistic domain. While
we find that both MetROMS and FESOM underestimate total basal mass loss from
ice shelves, this is a regional bias largely confined to small, warm-cavity
ice shelves which are not well resolved by the model configurations
considered here. With respect to simulated sub-ice-shelf circulation, some
ice-shelf cavities show agreement with the direction of transport inferred
from observations (such as the Ross Ice Shelf in both models, the Amery Ice
Shelf in MetROMS, and the George VI Ice Shelf in FESOM) and others show
disagreement (such as the George VI Ice Shelf in MetROMS, and the Amery and
Filchner–Ronne ice shelves in FESOM). FESOM's simulation of ice-shelf
cavities improves at higher resolution, suggesting that further refinement of
resolution is justified as available computational power continues to
increase. Sea-ice extent in both MetROMS and FESOM mostly agrees with
observations, although both models underestimate the summer sea-ice minimum,
and MetROMS requires surface salinity restoring to prevent a spurious
open-ocean polynya from forming in the Weddell Sea. Sea-ice production is too
strong in the Ross and Weddell seas compared to observations, and too weak in
the small coastal polynyas of the Australian sector. In the interior Southern
Ocean and the ACC, FESOM has an advantage due to its vertical coordinate
system, which is locally <inline-formula><mml:math id="M376" display="inline"><mml:mi>z</mml:mi></mml:math></inline-formula> coordinate compared to MetROMS'
terrain-following coordinate which covers the entire domain. Our results are
dependent on the ERA-Interim atmospheric reanalysis and are influenced by any
biases it may contain over the Southern Ocean, including its known
underestimation of summertime cloud cover which leads to excessive sea-ice
melt. We conclude that realistic intercomparisons of simulated ice-shelf
cavities are valuable for guiding model development. Future studies including
a greater variety of models, alternative atmospheric forcing datasets, and
ideally more observations would be worthwhile as these coupled
ocean/sea-ice/ice-shelf models continue to be developed by the community.</p>
</sec>

      
      </body>
    <back><notes notes-type="codeavailability"><title>Code availability</title>

      <p id="d1e6552">The source codes used for the simulations described here
are archived at <ext-link xlink:href="https://doi.org/10.5281/zenodo.1157229" ext-link-type="DOI">10.5281/zenodo.1157229</ext-link> (Naughten et al., 2018a) for MetROMS-iceshelf and
<ext-link xlink:href="https://doi.org/10.5281/zenodo.1157227" ext-link-type="DOI">10.5281/zenodo.1157227</ext-link> (Naughten et al., 2018b) for FESOM 1.4. These repositories also include
the grid/mesh files and the configuration files. The atmospheric forcing,
initial conditions, and ROMS northern boundary conditions can be obtained
from the authors upon request. Additionally, the maintained version of
MetROMS-iceshelf is publicly available at
<uri>https://github.com/knaughten/metroms_iceshelf</uri>, and the maintained
version of FESOM is available at
<uri>https://swrepo1.awi.de/svn/awi-cm/trunk</uri> following registration at
<uri>https://swrepo1.awi.de/projects/fesom/</uri>.</p>
  </notes><notes notes-type="competinginterests"><title>Competing interests</title>

      <p id="d1e6573">The authors declare that they have no conflict of
interest.</p>
  </notes><ack><title>Acknowledgements</title><p id="d1e6579">We would like to thank Michael Dinniman, Paul Holland, Petra Heil,
Elizabeth Hunke, Matt Mazloff, and Xylar Asay-Davis for their advice while
tuning and debugging MetROMS, and Nicholas Hannah for technical support.
Marta Kasper, Yoshihiro Nakayama, and Lukrecia Stulic provided assistance
with installing and configuring FESOM, and Dmitry Sein with generating the
FESOM mesh. We are also grateful to Sergey Danilov and Qiang Wang for helpful
discussions regarding effective resolution and pressure gradient calculation
in FESOM. Finally, Jean-Baptiste Sallée and Violaine Pellichero provided
observations of climatological mixed layer depth, and Takeshi Tamura provided
observations of sea-ice production. This research was supported by an
Australian Government scholarship under the Australian Postgraduate Award and
Research Training Program schemes, a UNSW Research Excellence Award, and UNSW
Science Silver Star and Gold Star awards. Support was also provided by the
Australian Research Council's Centre of Excellence for Climate System
Science, and the Australian Government's Cooperative Research Centre
Programme through the Antarctic Climate &amp; Ecosystems Cooperative Research
Centre.<?pagebreak page1286?> Computational resources were provided by the NCI National Facility at
the Australian National University, through awards under the Merit Allocation
Scheme, the Intersect allocation scheme, and the UNSW HPC at NCI Scheme. Ben
Galton-Fenzi is supported under the Australian Research Council's Special
Research Initiative for the Antarctic Gateway Partnership SRI40300001, which
also contributes to the World Climate Research Programme (WCRP) Climate and
Cryosphere (CliC) project Targeted Activity “Linkages Between Cryosphere
Elements”. Ralph Timmermann acknowledges funding by the Helmholtz Climate
Initiative REKLIM (Regional Climate Change), a joint research project of the
Helmholtz Association of German Research Centres (HGF). We are grateful to
Xylar Asay-Davis and one anonymous reviewer for their helpful comments on an
earlier version of the manuscript.<?xmltex \hack{\newline}?><?xmltex \hack{\newline}?> Edited by:
Jeremy Fyke<?xmltex \hack{\newline}?> Reviewed by: Xylar Asay-Davis and one anonymous
referee</p></ack><ref-list>
    <title>References</title>

      <ref id="bib1.bibx1"><label>Arakawa and Lamb(1977)</label><mixed-citation>
Arakawa, A. and Lamb, V. R.: Computational design of the basic dynamical
processes of the UCLA general circulation model, Methods in Computational
Physics: Advances in Research and Applications, 17, 173–265, 1977.</mixed-citation></ref>
      <ref id="bib1.bibx2"><label>Asay-Davis et al.(2016)</label><mixed-citation>Asay-Davis, X. S., Cornford, S. L., Durand, G., Galton-Fenzi, B. K.,
Gladstone, R. M., Gudmundsson, G. H., Hattermann, T., Holland, D. M.,
Holland, D., Holland, P. R., Martin, D. F., Mathiot, P., Pattyn, F., and
Seroussi, H.: Experimental design for three interrelated marine ice sheet and
ocean model intercomparison projects: MISMIP v. 3 (MISMIP <inline-formula><mml:math id="M377" display="inline"><mml:mo>+</mml:mo></mml:math></inline-formula>), ISOMIP v. 2
(ISOMIP <inline-formula><mml:math id="M378" display="inline"><mml:mo>+</mml:mo></mml:math></inline-formula>) and MISOMIP v. 1 (MISOMIP1), Geosci. Model Dev., 9, 2471–2497,
https://doi.org/10.5194/gmd-9-2471-2016, 2016.</mixed-citation></ref>
      <ref id="bib1.bibx3"><label>AVISO(2011)</label><mixed-citation>AVISO: AVISO Level 4 Absolute Dynamic Topography for Climate Model
Comparison, Version 1, <ext-link xlink:href="https://doi.org/10.5067/DYNTO-1D1M1" ext-link-type="DOI">10.5067/DYNTO-1D1M1</ext-link>, 2011.</mixed-citation></ref>
      <ref id="bib1.bibx4"><label>Bodas-Salcedo et al.(2016)</label><mixed-citation>Bodas-Salcedo, A., Andrews, T., Karmalkar, A. V., and Ringer, M. A.: Cloud
liquid water path and radiative feedbacks over the Southern Ocean, Geophys.
Res. Lett., 43, 10938–10946, <ext-link xlink:href="https://doi.org/10.1002/2016GL070770" ext-link-type="DOI">10.1002/2016GL070770</ext-link>, 2016.</mixed-citation></ref>
      <ref id="bib1.bibx5"><label>Bouillon et al.(2013)</label><mixed-citation>Bouillon, S., Fichefet, T., Legat, V., and Madec, G.: The
elastic-viscous-plastic method revisited, Ocean Model., 71, 2–12,
<ext-link xlink:href="https://doi.org/10.1016/j.ocemod.2013.05.013" ext-link-type="DOI">10.1016/j.ocemod.2013.05.013</ext-link>, 2013.</mixed-citation></ref>
      <ref id="bib1.bibx6"><label>Briegleb and Light(2007)</label><mixed-citation>
Briegleb, B. P. and Light, B.: A Delta-Eddington multiple scattering
parameterization for solar radiation in the sea ice component of the
Community Climate System Model: NCAR Tech. Note NCAR/TN-472+STR, Tech. rep.,
National Center for Atmospheric Research, 2007.</mixed-citation></ref>
      <ref id="bib1.bibx7"><label>Chapman(1985)</label><mixed-citation>Chapman, D. C.: Numerical Treatment of Cross-Shelf Open Boundaries in a
Barotropic Coastal Ocean Model, J. Phys. Oceanogr., 15, 1060–1075,
<ext-link xlink:href="https://doi.org/10.1175/1520-0485(1985)015&lt;1060:NTOCSO&gt;2.0.CO;2" ext-link-type="DOI">10.1175/1520-0485(1985)015&lt;1060:NTOCSO&gt;2.0.CO;2</ext-link>, 1985.</mixed-citation></ref>
      <ref id="bib1.bibx8"><label>Cook et al.(2013)</label><mixed-citation>Cook, C. P., van de Flierdt, T., Williams, T., Hemming, S. R., Iwai, M.,
Kobayashi, M., Jimenez-Espejo, F. J., Escutia, C., González, J. J.,
Khim, B.-K., McKay, R. M., Passchier, S., Bohaty, S. M., Riesselman, C. R.,
Tauxe, L., Sugisaki, S., Galindo, A. L., Patterson, M. O., Sangiorgi, F.,
Pierce, E. L., Brinkhuis, H., Klaus, A., Fehr, A., Bendle, J. A. P., Bijl,
P. K., Carr, S. A., Dunbar, R. B., Flores, J. A., Hayden, T. G., Katsuki, K.,
Kong, G. S., Nakai, M., Olney, M. P., Pekar, S. F., Pross, J., Röhl,
U., Sakai, T., Shrivastava, P. K., Stickley, C. E., Tuo, S., Welsh, K., and
Yamane, M.: Dynamic behaviour of the East Antarctic ice sheet during
Pliocene warmth, Nat. Geosci., 6, 765–769, <ext-link xlink:href="https://doi.org/10.1038/ngeo1889" ext-link-type="DOI">10.1038/ngeo1889</ext-link>, 2013.</mixed-citation></ref>
      <ref id="bib1.bibx9"><label>Cougnon et al.(2013)</label><mixed-citation>Cougnon, E. A., Galton-Fenzi, B. K., Meijers, A. J. S., and Legrésy,
B.: Modeling interannual dense shelf water export in the region of the Mertz
Glacier Tongue (1992–2007), J. Geophys. Res.-Oceans, 118, 5858–5872,
<ext-link xlink:href="https://doi.org/10.1002/2013JC008790" ext-link-type="DOI">10.1002/2013JC008790</ext-link>, 2013.</mixed-citation></ref>
      <ref id="bib1.bibx10"><label>Cunningham et al.(2003)</label><mixed-citation>Cunningham, S. A., Alderson, S. G., King, B. A., and Brandon, M. A.:
Transport and variability of the Antarctic Circumpolar Current in Drake
Passage, J. Geophys. Res., 108, 8084, <ext-link xlink:href="https://doi.org/10.1029/2001JC001147" ext-link-type="DOI">10.1029/2001JC001147</ext-link>, 2003.</mixed-citation></ref>
      <ref id="bib1.bibx11"><label>Danilov et al.(2015)</label><mixed-citation>
Danilov, S., Wang, Q., Timmermann, R., Iakovlev, N., Sidorenko, D., Kimmritz,
M., Jung, T., and Schröter, J.: Finite-Element Sea Ice Model (FESIM),
version 2, Geosci. Model Dev., 8, 1747-1761,
https://doi.org/10.5194/gmd-8-1747-2015, 2015.</mixed-citation></ref>
      <ref id="bib1.bibx12"><label>Debernard et al.(2017)</label><mixed-citation>Debernard, J. B., Kristensen, N. M., Maartensson, S., Wang, K., and Waagbo,
G. A.: metno/metroms: Intermediate release, <ext-link xlink:href="https://doi.org/10.5281/zenodo.290667" ext-link-type="DOI">10.5281/zenodo.290667</ext-link>,
2017.</mixed-citation></ref>
      <ref id="bib1.bibx13"><label>Deconto and Pollard(2016)</label><mixed-citation>Deconto, R. M. and Pollard, D.: Contribution of Antarctica to past and
future sea-level rise, Nature, 531, 591–597, <ext-link xlink:href="https://doi.org/10.1038/nature17145" ext-link-type="DOI">10.1038/nature17145</ext-link>,
2016.</mixed-citation></ref>
      <ref id="bib1.bibx14"><label>Dee et al.(2011)</label><mixed-citation>
Dee, D. P., Uppala, S. M., Simmons, A. J., Berrisford, P., Poli, P.,
Kobayashi, S., Andrae, U., Balmaseda, M. A., Balsamo, G., Bauer, P.,
Bechtold, P., Beljaars, A. C. M., van de Berg, I., Biblot, J., Bormann, N.,
Delsol, C., Dragani, R., Fuentes, M., Greer, A. J., Haimberger, L., Healy, S.
B., Hersbach, H., Holm, E. V., Isaksen, L., Kallberg, P., Kohler, M.,
Matricardi, M., McNally, A. P., Mong-Sanz, B. M., Morcette, J.-J., Park,
B.-K., Peubey, C., de Rosnay, P., Tavolato, C., Thepaut, J. N., and Vitart,
F.: The ERA-Interim reanalysis: Configuration and performance of the data
assimilation system, Q. J. Roy. Meteorol. Soc., 137, 553–597,
doi:10.1002/qj.828, 2011.</mixed-citation></ref>
      <ref id="bib1.bibx15"><label>Depoorter et al.(2013)</label><mixed-citation>Depoorter, M. A., Bamber, J. L., Griggs, J. A., Lenaerts, J. T. M.,
Ligtenberg, S. R. M., van den Broeke, M. R., and Moholdt, G.: Calving fluxes
and basal melt rates of Antarctic ice shelves, Nature, 502, 89–92,
<ext-link xlink:href="https://doi.org/10.1038/nature12567" ext-link-type="DOI">10.1038/nature12567</ext-link>, 2013.</mixed-citation></ref>
      <ref id="bib1.bibx16"><label>Dinniman et al.(2011)</label><mixed-citation>Dinniman, M. S., Klinck, J. M., and Smith, W. O.: A model study of
Circumpolar Deep Water on the West Antarctic Peninsula and Ross Sea
continental shelves, Deep-Sea Res. Pt. II, 58, 1508–1523,
<ext-link xlink:href="https://doi.org/10.1016/j.dsr2.2010.11.013" ext-link-type="DOI">10.1016/j.dsr2.2010.11.013</ext-link>, 2011.</mixed-citation></ref>
      <ref id="bib1.bibx17"><label>Dinniman et al.(2016)</label><mixed-citation>Dinniman, M. S., Asay-Davis, X. S., Galton-Fenzi, B. K., Holland, P. R., and
Jenkins, A.: Modeling Ice Shelf/Ocean Interaction in Antarctica: a review,
Oceanography, 29, 144–153, <ext-link xlink:href="https://doi.org/10.5670/oceanog.2016.106" ext-link-type="DOI">10.5670/oceanog.2016.106</ext-link>, 2016.</mixed-citation></ref>
      <ref id="bib1.bibx18"><label>Donohue et al.(2016)</label><mixed-citation>Donohue, K. A., Tracey, K. L., Watts, D. R., Chidichimo, M. P., and
Chereskin, T. K.: Mean Antarctic Circumpolar Current transport measured in
Drake Passage, Geophys. Res. Lett., 43, 11,760–11,767,
<ext-link xlink:href="https://doi.org/10.1002/2016GL070319" ext-link-type="DOI">10.1002/2016GL070319</ext-link>, 2016.</mixed-citation></ref>
      <ref id="bib1.bibx19"><label>Dupont(2005)</label><mixed-citation>Dupont, T. K.: Assessment of the importance of ice-shelf buttressing to
ice-sheet flow, Geophys. Res. Letters, 32, L04503,
<ext-link xlink:href="https://doi.org/10.1029/2004GL022024" ext-link-type="DOI">10.1029/2004GL022024</ext-link>, 2005.</mixed-citation></ref>
      <ref id="bib1.bibx20"><label>Dutrieux et al.(2014)</label><mixed-citation>Dutrieux, P., De Rydt, J., Jenkins, A., Holland, P. R., Ha, H. K., Lee,
S. H., Steig, E. J., Ding, Q., Abrahamsen, E. P., and Schröder, M.:
Strong sensitivity of Pine Island ice-shelf melting to climatic
variability, Science, 343, 174–178, <ext-link xlink:href="https://doi.org/10.1126/science.1244341" ext-link-type="DOI">10.1126/science.1244341</ext-link>, 2014.</mixed-citation></ref>
      <?pagebreak page1287?><ref id="bib1.bibx21"><label>Dutton et al.(2015)</label><mixed-citation>Dutton, A., Carlson, A. E., Long, A. J., Milne, G. A., Clark, P. U., Deconto,
R. M., Horton, B. P., Rahmstorf, S., and Raymo, M. E.: Sea-level rise due to
polar ice-sheet mass loss during past warm periods, Science, 349, aaa4019,
<ext-link xlink:href="https://doi.org/10.1126/science.aaa4019" ext-link-type="DOI">10.1126/science.aaa4019</ext-link>, 2015.</mixed-citation></ref>
      <ref id="bib1.bibx22"><label>England(1993)</label><mixed-citation>England, M. H.: Representing the Global-Scale Water Masses in Ocean General
Circulation Models, J. Phys. Oceanogr., 23, 1523–1552,
<ext-link xlink:href="https://doi.org/10.1175/1520-0485(1993)023&lt;1523:RTGSWM&gt;2.0.CO;2" ext-link-type="DOI">10.1175/1520-0485(1993)023&lt;1523:RTGSWM&gt;2.0.CO;2</ext-link>, 1993.</mixed-citation></ref>
      <ref id="bib1.bibx23"><label>England et al.(1993)</label><mixed-citation>England, M. H., Godfrey, J. S., Hirst, A. C., and Tomczak, M.: The Mechanism
for Antarctic Intermediate Water Renewal in a World Ocean Model, J. Phys.
Oceanogr., 23, 1553–1560,
<ext-link xlink:href="https://doi.org/10.1175/1520-0485(1993)023&lt;1553:TMFAIW&gt;2.0.CO;2" ext-link-type="DOI">10.1175/1520-0485(1993)023&lt;1553:TMFAIW&gt;2.0.CO;2</ext-link>, 1993.</mixed-citation></ref>
      <ref id="bib1.bibx24"><label>Engwirda et al.(2017)</label><mixed-citation>Engwirda, D., Kelley, M., and Marshall, J.: High-order accurate
finite-volume formulations for the pressure gradient force in layered ocean
models, Ocean Model., 116, 1–15, <ext-link xlink:href="https://doi.org/10.1016/j.ocemod.2017.05.003" ext-link-type="DOI">10.1016/j.ocemod.2017.05.003</ext-link>, 2017.</mixed-citation></ref>
      <ref id="bib1.bibx25"><label>Fairall et al.(1996)</label><mixed-citation>Fairall, C. W., Bradley, E. F., Rogers, D. P., Edson, J. B., and Young,
G. S.: Bulk parameterization of air-sea fluxes for Tropical Ocean-Global
Atmosphere Coupled-Ocean Atmosphere Response Experiment, J. Geophys.
Res.-Oceans, 101, 3747–3764, <ext-link xlink:href="https://doi.org/10.1029/95JC03205" ext-link-type="DOI">10.1029/95JC03205</ext-link>, 1996.</mixed-citation></ref>
      <ref id="bib1.bibx26"><label>Fetterrer et al.(2016)</label><mixed-citation>Fetterrer, F., Knowles, K., Meier, W., and Savoie, M.: Sea Ice Index,
Version 2, National Snow &amp; Ice Data Center, <ext-link xlink:href="https://doi.org/10.7265/N5736NV7" ext-link-type="DOI">10.7265/N5736NV7</ext-link>, 2016.</mixed-citation></ref>
      <ref id="bib1.bibx27"><label>Flather(1976)</label><mixed-citation>
Flather, R.: A tidal model of the northwest European continental shelf,
Memoires de la Societe Royale de Sciences de Liege, 6, 141–164, 1976.</mixed-citation></ref>
      <ref id="bib1.bibx28"><label>Foldvik et al.(1990)</label><mixed-citation>Foldvik, A., Middleton, J. H., and Foster, T. D.: The tides of the southern
Weddell Sea, Deep-Sea Res. Pt. I, 37, 1345–1362,
<ext-link xlink:href="https://doi.org/10.1016/0198-0149(90)90047-Y" ext-link-type="DOI">10.1016/0198-0149(90)90047-Y</ext-link>, 1990.</mixed-citation></ref>
      <ref id="bib1.bibx29"><label>Foldvik et al.(2004)</label><mixed-citation>Foldvik, A., Gammelsrød, T., Østerhus, S., Fahrbach, E., Rohart, G.,
Schröder, M., Nicholls, K. W., Padman, L., and Woodgate, R. A.: Ice
shelf water overflow and bottom water formation in the southern Weddell Sea,
J. Geophys. Res., 109, C02015, <ext-link xlink:href="https://doi.org/10.1029/2003JC002008" ext-link-type="DOI">10.1029/2003JC002008</ext-link>, 2004.</mixed-citation></ref>
      <ref id="bib1.bibx30"><label>Fraser et al.(2012)</label><mixed-citation>Fraser, A. D., Massom, R. A., Michael, K. J., Galton-Fenzi, B. K., and
Lieser, J. L.: East Antarctic Landfast Sea Ice Distribution and Variability,
2000-08, J. Climate, 25, 1137–1156, <ext-link xlink:href="https://doi.org/10.1175/JCLI-D-10-05032.1" ext-link-type="DOI">10.1175/JCLI-D-10-05032.1</ext-link>, 2012.</mixed-citation></ref>
      <ref id="bib1.bibx31"><label>Fretwell et al.(2013)</label><mixed-citation>
Fretwell, P., Pritchard, H. D., Vaughan, D. G., Bamber, J. L., Barrand, N.
E., Bell, R., Bianchi, C., Bingham, R. G., Blankenship, D. D., Casassa, G.,
Catania, G., Callens, D., Conway, H., Cook, A. J., Corr, H. F. J., Damaske,
D., Damm, V., Ferraccioli, F., Forsberg, R., Fujita, S., Gim, Y., Gogineni,
P., Griggs, J. A., Hindmarsh, R. C. A., Holmlund, P., Holt, J. W., Jacobel,
R. W., Jenkins, A., Jokat, W., Jordan, T., King, E. C., Kohler, J., Krabill,
W., Riger-Kusk, M., Langley, K. A., Leitchenkov, G., Leuschen, C., Luyendyk,
B. P., Matsuoka, K., Mouginot, J., Nitsche, F. O., Nogi, Y., Nost, O. A.,
Popov, S. V., Rignot, E., Rippin, D. M., Rivera, A., Roberts, J., Ross, N.,
Siegert, M. J., Smith, A. M., Steinhage, D., Studinger, M., Sun, B., Tinto,
B. K., Welch, B. C., Wilson, D., Young, D. A., Xiangbin, C., and Zirizzotti,
A.: Bedmap2: improved ice bed, surface and thickness datasets for Antarctica,
The Cryosphere, 7, 375–393, https://doi.org/10.5194/tc-7-375-2013, 2013.</mixed-citation></ref>
      <ref id="bib1.bibx32"><label>Gade(1979)</label><mixed-citation>Gade, H. G.: Melting of Ice in Sea Water: A Primitive Model with Application
to the Antarctic Ice Shelf and Icebergs, J. Phys. Oceanogr., 9, 189–198,
<ext-link xlink:href="https://doi.org/10.1175/1520-0485(1979)009&lt;0189:MOIISW&gt;2.0.CO;2" ext-link-type="DOI">10.1175/1520-0485(1979)009&lt;0189:MOIISW&gt;2.0.CO;2</ext-link>, 1979.</mixed-citation></ref>
      <ref id="bib1.bibx33"><label>Galton-Fenzi(2009)</label><mixed-citation>
Galton-Fenzi, B. K.: Modelling Ice-Shelf/Ocean Interaction, PhD thesis,
University of Tasmania, 2009.</mixed-citation></ref>
      <ref id="bib1.bibx34"><label>Galton-Fenzi et al.(2008)</label><mixed-citation>Galton-Fenzi, B. K., Maraldi, C., Coleman, R., and Hunter, J. R.: The cavity
under the Amery Ice Shelf, East Antarctica, J. Glaciol., 54, 881–887,
<ext-link xlink:href="https://doi.org/10.3189/002214308787779898" ext-link-type="DOI">10.3189/002214308787779898</ext-link>, 2008.</mixed-citation></ref>
      <ref id="bib1.bibx35"><label>Galton-Fenzi et al.(2012)</label><mixed-citation>Galton-Fenzi, B. K., Hunter, J. R., Coleman, R., Marsland, S. J., and Warner,
R. C.: Modeling the basal melting and marine ice accretion of the Amery Ice
Shelf, J. Geophys. Res.-Oceans, 117, C09031, <ext-link xlink:href="https://doi.org/10.1029/2012JC008214" ext-link-type="DOI">10.1029/2012JC008214</ext-link>,
2012.</mixed-citation></ref>
      <ref id="bib1.bibx36"><label>Gent and McWilliams(1990)</label><mixed-citation>Gent, P. R. and McWilliams, J. C.: Isopycnal Mixing in Ocean Circulation
Models, J. Phys. Oceanogr., 20, 150–155,
<ext-link xlink:href="https://doi.org/10.1175/1520-0485(1990)020&lt;0150:IMIOCM&gt;2.0.CO;2" ext-link-type="DOI">10.1175/1520-0485(1990)020&lt;0150:IMIOCM&gt;2.0.CO;2</ext-link>, 1990.</mixed-citation></ref>
      <ref id="bib1.bibx37"><label>Gent et al.(1995)</label><mixed-citation>Gent, P. R., Willebrand, J., McDougall, T. J., and McWilliams, J. C.:
Parameterizing Eddy-Induced Tracer Transports in Ocean Circulation Models,
J. Phys. Oceanogr., 25, 463–474,
<ext-link xlink:href="https://doi.org/10.1175/1520-0485(1995)025&lt;0463:PEITTI&gt;2.0.CO;2" ext-link-type="DOI">10.1175/1520-0485(1995)025&lt;0463:PEITTI&gt;2.0.CO;2</ext-link>, 1995.</mixed-citation></ref>
      <ref id="bib1.bibx38"><label>Gladstone et al.(2017)</label><mixed-citation>
Gladstone, R. M., Jong, L., Galton-Fenzi, B. K., Gwyther, D. E., and Moore,
J. C.: Simulating the interactions between marine ice sheets and their
surrounding oceans, in: Forum for Research into Ice Shelf Processes, Bergen,
Norway, 2017.</mixed-citation></ref>
      <ref id="bib1.bibx39"><label>Golledge et al.(2015)</label><mixed-citation>Golledge, N. R., Kowalewski, D. E., Naish, T. R., Levy, R. H., Fogwill,
C. J., and Gasson, E. G. W.: The multi-millennial Antarctic commitment to
future sea-level rise, Nature, 526, 421–425, <ext-link xlink:href="https://doi.org/10.1038/nature15706" ext-link-type="DOI">10.1038/nature15706</ext-link>,
2015.</mixed-citation></ref>
      <ref id="bib1.bibx40"><label>Golledge et al.(2017)</label><mixed-citation>Golledge, N. R., Levy, R. H., McKay, R. M., and Naish, T. R.: East Antarctic
ice sheet most vulnerable to Weddell Sea warming, Geophys. Res. Lett., 44,
1–9, <ext-link xlink:href="https://doi.org/10.1002/2016GL072422" ext-link-type="DOI">10.1002/2016GL072422</ext-link>, 2017.</mixed-citation></ref>
      <ref id="bib1.bibx41"><label>Goosse and Fichefet(2001)</label><mixed-citation>Goosse, H. and Fichefet, T.: Open-ocean convection and polynya formation in
a large-scale ice-ocean model, Tellus A, 53, 94–111,
<ext-link xlink:href="https://doi.org/10.1034/j.1600-0870.2001.01061.x" ext-link-type="DOI">10.1034/j.1600-0870.2001.01061.x</ext-link>, 2001.</mixed-citation></ref>
      <ref id="bib1.bibx42"><label>Gordon et al.(2015)</label><mixed-citation>Gordon, A. L., Huber, B. A., and Busecke, J.: Bottom water export from the
western Ross Sea, 2007 through 2010, Geophys. Res. Lett., 42, 5387–5394,
<ext-link xlink:href="https://doi.org/10.1002/2015GL064457" ext-link-type="DOI">10.1002/2015GL064457</ext-link>, 2015.</mixed-citation></ref>
      <ref id="bib1.bibx43"><label>Graham et al.(2016)</label><mixed-citation>Graham, J. A., Dinniman, M. S., and Klinck, J. M.: Impact of model
resolution for on-shelf heat transport along the West Antarctic Peninsula,
J. Geophys. Res.-Oceans, 121, 7880–7897, <ext-link xlink:href="https://doi.org/10.1002/2016JC011875" ext-link-type="DOI">10.1002/2016JC011875</ext-link>, 2016.</mixed-citation></ref>
      <ref id="bib1.bibx44"><label>Greenbaum et al.(2015)</label><mixed-citation>Greenbaum, J. S., Blankenship, D. D., Young, D. A., Richter, T. G., Roberts,
J. L., Aitken, A. R. A., Legrésy, B., Schroeder, D. M., Warner, R. C.,
van Ommen, T. D., and Siegert, M. J.: Ocean access to a cavity beneath
Totten Glacier in East Antarctica, Nat. Geosci., 8, 294–298,
<ext-link xlink:href="https://doi.org/10.1038/NGEO2388" ext-link-type="DOI">10.1038/NGEO2388</ext-link>, 2015.</mixed-citation></ref>
      <ref id="bib1.bibx45"><label>Griffies et al.(2000)</label><mixed-citation>Griffies, S. M., Böning, C., Bryan, F. O., Chassignet, E. P., Gerdes,
R., Hasumi, H., Hirst, A. C., Treguier, A.-M., and Webb, D.: Developments in
ocean climate modelling, J. Comput. Phys., 2, 123–192,
<ext-link xlink:href="https://doi.org/10.1016/S1463-5003(00)00014-7" ext-link-type="DOI">10.1016/S1463-5003(00)00014-7</ext-link>, 2000.</mixed-citation></ref>
      <ref id="bib1.bibx46"><label>Griffies et al.(2009)</label><mixed-citation>Griffies, S. M., Biastoch, A., Böning, C., Bryan, F. O., Danabasoglu,
G., Chassignet, E. P., England, M. H., Gerdes, R., Haak, H., Hallberg, R. W.,
Hazeleger, W., Jungclaus, J. H., Large, W. G., Madec, G., Pirani, A.,
Samuels, B. L., Scheinert, M., Gupta, A. S., Severijns, C. A., Simmons,
H. L., Treguier,<?pagebreak page1288?> A.-M., Winton, M., Yeager, S. G., and Yin, J.: Coordinated
Ocean-ice Reference Experiments (COREs), Ocean Model., 26, 1–46,
<ext-link xlink:href="https://doi.org/10.1016/j.ocemod.2008.08.007" ext-link-type="DOI">10.1016/j.ocemod.2008.08.007</ext-link>, 2009.</mixed-citation></ref>
      <ref id="bib1.bibx47"><label>Gwyther(2016)</label><mixed-citation>
Gwyther, D. E.: Modelling regional ice shelf/ocean interaction with ROMS,
in: ROMS Asia-Pacific Workshop, Hobart, Australia, 2016.</mixed-citation></ref>
      <ref id="bib1.bibx48"><label>Gwyther et al.(2014)</label><mixed-citation>
Gwyther, D. E., Galton-Fenzi, B. K., Hunter, J. R., and Roberts, J. L.:
Simulated melt rates for the Totten and Dalton ice shelves, Ocean Sci., 10,
267–279, https://doi.org/10.5194/os-10-267-2014, 2014.</mixed-citation></ref>
      <ref id="bib1.bibx49"><label>Gwyther et al.(2015)</label><mixed-citation>Gwyther, D. E., Galton-Fenzi, B. K., Dinniman, M. S., Roberts, J. L., and
Hunter, J. R.: The effect of basal friction on melting and freezing in ice
shelf-ocean models, Ocean Model., 95, 38–52,
<ext-link xlink:href="https://doi.org/10.1016/j.ocemod.2015.09.004" ext-link-type="DOI">10.1016/j.ocemod.2015.09.004</ext-link>, 2015.</mixed-citation></ref>
      <ref id="bib1.bibx50"><label>Gwyther et al.(2016)</label><mixed-citation>Gwyther, D. E., Cougnon, E. A., Galton-Fenzi, B. K., Roberts, J. L., Hunter,
J. R., and Dinniman, M. S.: Modelling the response of ice shelf basal
melting to different ocean cavity environmental regimes, Ann. Glaciol., 57,
131–141, <ext-link xlink:href="https://doi.org/10.1017/aog.2016.31" ext-link-type="DOI">10.1017/aog.2016.31</ext-link>, 2016.</mixed-citation></ref>
      <ref id="bib1.bibx51"><label>Haas et al.(2008)</label><mixed-citation>Haas, C., Nicolaus, M., Willmes, S., Worby, A., and Flinspach, D.: Sea ice
and snow thickness and physical properties of an ice floe in the western
Weddell Sea and their changes during spring warming, Deep-Sea Res. Pt. II,
55, 963–974, <ext-link xlink:href="https://doi.org/10.1016/j.dsr2.2007.12.020" ext-link-type="DOI">10.1016/j.dsr2.2007.12.020</ext-link>, 2008.</mixed-citation></ref>
      <ref id="bib1.bibx52"><label>Haidvogel and Beckmann(1999)</label><mixed-citation>
Haidvogel, D. B. and Beckmann, A.: Numerical Ocean Circulation Modeling,
Imperial College Press, London, 1999.</mixed-citation></ref>
      <ref id="bib1.bibx53"><label>Haney(1991)</label><mixed-citation>Haney, R. L.: On the Pressure Gradient Force over Steep Topography in Sigma
Coordinate Ocean Models, J. Phys. Oceanogr., 21, 610–619,
<ext-link xlink:href="https://doi.org/10.1175/1520-0485(1991)021&lt;0610:OTPGFO&gt;2.0.CO;2" ext-link-type="DOI">10.1175/1520-0485(1991)021&lt;0610:OTPGFO&gt;2.0.CO;2</ext-link>, 1991.</mixed-citation></ref>
      <ref id="bib1.bibx54"><label>Hattermann et al.(2012)</label><mixed-citation>Hattermann, T., Nøst, O. A., Lilly, J. M., and Smedsrud, L. H.: Two years
of oceanic observations below the Fimbul Ice Shelf, Antarctica, Geophys.
Res. Lett., 39, L12605, <ext-link xlink:href="https://doi.org/10.1029/2012GL051012" ext-link-type="DOI">10.1029/2012GL051012</ext-link>, 2012.</mixed-citation></ref>
      <ref id="bib1.bibx55"><label>Hellmer and Olbers(1989)</label><mixed-citation>Hellmer, H. H. and Olbers, D. J.: A two-dimensional model for the
thermohaline circulation under an ice shelf, Antarct. Sci., 1, 325–336,
<ext-link xlink:href="https://doi.org/10.1017/S0954102089000490" ext-link-type="DOI">10.1017/S0954102089000490</ext-link>, 1989.</mixed-citation></ref>
      <ref id="bib1.bibx56"><label>Hellmer et al.(1998)</label><mixed-citation>Hellmer, H. H., Jacobs, S. S., and Jenkins, A.: Oceanic erosion of a
floating Antarctic glacier in the Amundsen Sea, in: Ocean, Ice, and
Atmosphere: Interactions at the Antarctic Continental Margin, edited by:
Jacobs, S. S. and Weiss, R. F., American Geophysical Union, Washington, DC,
319–339, <ext-link xlink:href="https://doi.org/10.1029/AR075p0341" ext-link-type="DOI">10.1029/AR075p0341</ext-link>, 1998.</mixed-citation></ref>
      <ref id="bib1.bibx57"><label>Hellmer et al.(2012)</label><mixed-citation>Hellmer, H. H., Kauker, F., Timmermann, R., Determann, J., and Rae, J.:
Twenty-first-century warming of a large Antarctic ice-shelf cavity by a
redirected coastal current, Nature, 485, 225–228,
<ext-link xlink:href="https://doi.org/10.1038/nature11064" ext-link-type="DOI">10.1038/nature11064</ext-link>, 2012.</mixed-citation></ref>
      <ref id="bib1.bibx58"><label>Hellmer et al.(2017)</label><mixed-citation>Hellmer, H. H., Kauker, F., Timmermann, R., and Hattermann, T.: The fate of
the southern Weddell Sea continental shelf in a warming climate, J. Climate,
30, 4337–4350, <ext-link xlink:href="https://doi.org/10.1175/JCLI-D-16-0420.1" ext-link-type="DOI">10.1175/JCLI-D-16-0420.1</ext-link>, 2017.</mixed-citation></ref>
      <ref id="bib1.bibx59"><label>Herraiz-Borreguero et al.(2016)</label><mixed-citation>Herraiz-Borreguero, L., Church, J. A., Allison, I., Pena-Molino, B., Coleman,
R., Tomczak, M., and Craven, M.: Basal melt, seasonal water mass
transformation, ocean current variability, and deep convection processes
along the Amery Ice Shelf, J. Geophys. Res.-Oceans, 121, 4946–4965,
<ext-link xlink:href="https://doi.org/10.1002/2013JC008846" ext-link-type="DOI">10.1002/2013JC008846</ext-link>, 2016.</mixed-citation></ref>
      <ref id="bib1.bibx60"><?xmltex \def\ref@label{{Heuz{\'{e}} et~al.(2015)}}?><label>Heuzé et al.(2015)</label><mixed-citation>
Heuzé, C., Ridley, J. K., Calvert, D., Stevens, D. P., and Heywood, K.
J.: Increasing vertical mixing to reduce Southern Ocean deep convection in
NEMO3.4, Geosci. Model Dev., 8, 3119–3130,
https://doi.org/10.5194/gmd-8-3119-2015, 2015.</mixed-citation></ref>
      <ref id="bib1.bibx61"><label>Hogg and Gudmundsson(2017)</label><mixed-citation>Hogg, A. E. and Gudmundsson, G. H.: Impacts of the Larsen-C Ice Shelf
calving event, Nat. Clim. Change, 7, 540–542, <ext-link xlink:href="https://doi.org/10.1038/nclimate3359" ext-link-type="DOI">10.1038/nclimate3359</ext-link>,
2017.</mixed-citation></ref>
      <ref id="bib1.bibx62"><label>Holland and Jenkins(1999)</label><mixed-citation>Holland, D. M. and Jenkins, A.: Modeling Thermodynamic Ice-Ocean
Interactions at the Base of an Ice Shelf, J. Phys. Oceanogr., 29,
1787–1800, <ext-link xlink:href="https://doi.org/10.1175/1520-0485(1999)029&lt;1787:MTIOIA&gt;2.0.CO;2" ext-link-type="DOI">10.1175/1520-0485(1999)029&lt;1787:MTIOIA&gt;2.0.CO;2</ext-link>, 1999.</mixed-citation></ref>
      <ref id="bib1.bibx63"><label>Holland et al.(2009)</label><mixed-citation>Holland, P. R., Corr, H. F. J., Vaughan, D. G., Jenkins, A., and Skvarca, P.:
Marine ice in Larsen Ice Shelf, Geophys. Res. Lett., 36, 1–6,
<ext-link xlink:href="https://doi.org/10.1029/2009GL038162" ext-link-type="DOI">10.1029/2009GL038162</ext-link>, 2009.</mixed-citation></ref>
      <ref id="bib1.bibx64"><label>Holland et al.(2014)</label><mixed-citation>Holland, P. R., Bruneau, N., Enright, C., Losch, M., Kurtz, N. T., and Kwok,
R.: Modeled trends in Antarctic Sea Ice Thickness, J. Climate, 27,
3784–3801, <ext-link xlink:href="https://doi.org/10.1175/JCLI-D-13-00301.1" ext-link-type="DOI">10.1175/JCLI-D-13-00301.1</ext-link>, 2014.</mixed-citation></ref>
      <ref id="bib1.bibx65"><label>Hunke et al.(2013)</label><mixed-citation>Hunke, E. C., Hebert, D. A., and Lecomte, O.: Level-ice melt ponds in the
Los Alamos sea ice model, CICE, Ocean Model., 71, 26–42,
<ext-link xlink:href="https://doi.org/10.1016/j.ocemod.2012.11.008" ext-link-type="DOI">10.1016/j.ocemod.2012.11.008</ext-link>, 2013.</mixed-citation></ref>
      <ref id="bib1.bibx66"><label>Hunke et al.(2015)</label><mixed-citation>
Hunke, E. C., Lipscomb, W. H., Turner, A. K., Jeffery, N., and Elliott, S.:
CICE : the Los Alamos Sea Ice Model Documentation and Software User's Manual
LA-CC-06-012, Tech. rep., Los Alamos National Laboratory, 2015.</mixed-citation></ref>
      <ref id="bib1.bibx67"><label>Hunter(2006)</label><mixed-citation>
Hunter, J. R.: Specification for Test Models of Ice Shelf Cavities, Tech.
rep., Antarctic Climate &amp; Ecosystems Cooperative Research Centre, 2006.</mixed-citation></ref>
      <ref id="bib1.bibx68"><label>Jacob et al.(2005)</label><mixed-citation>Jacob, R., Larson, J., and Ong, E.: M <inline-formula><mml:math id="M379" display="inline"><mml:mo>×</mml:mo></mml:math></inline-formula> N Communication and
Parallel Interpolation in Community Climate System Model Version 3 Using the
Model Coupling Toolkit, Int. J. High Perform. C, 19, 293–307,
<ext-link xlink:href="https://doi.org/10.1177/1094342005056116" ext-link-type="DOI">10.1177/1094342005056116</ext-link>, 2005.</mixed-citation></ref>
      <ref id="bib1.bibx69"><label>Jacobs and Comiso(1989)</label><mixed-citation>Jacobs, S. S. and Comiso, J. C.: Sea ice and oceanic processes on the Ross
Sea continental shelf, J. Geophys. Res., 94, 18195–18211,
<ext-link xlink:href="https://doi.org/10.1029/JC094iC12p18195" ext-link-type="DOI">10.1029/JC094iC12p18195</ext-link>, 1989.</mixed-citation></ref>
      <ref id="bib1.bibx70"><label>Jacobs et al.(1979)</label><mixed-citation>Jacobs, S. S., Gordon, A. L., and Ardai, J. L.: Circulation and melting
beneath the Ross Ice Shelf, Science, 203, 439–443,
<ext-link xlink:href="https://doi.org/10.1126/science.203.4379.439" ext-link-type="DOI">10.1126/science.203.4379.439</ext-link>, 1979.</mixed-citation></ref>
      <ref id="bib1.bibx71"><label>Jacobs et al.(1992)</label><mixed-citation>Jacobs, S. S., Hellmer, H. H., Doake, C. S. M., Jenkins, A., and Frolich,
R. M.: Melting of ice shelves and the mass balance of Antarctica, J.
Glaciol., 38, 375–387, <ext-link xlink:href="https://doi.org/10.1038/228047a0" ext-link-type="DOI">10.1038/228047a0</ext-link>, 1992.</mixed-citation></ref>
      <ref id="bib1.bibx72"><label>Jacobs et al.(2011)</label><mixed-citation>Jacobs, S. S., Jenkins, A., Giulivi, C. F., and Dutrieux, P.: Stronger ocean
circulation and increased melting under Pine Island Glacier ice shelf, Nat.
Geosci., 4, 519–523, <ext-link xlink:href="https://doi.org/10.1038/ngeo1188" ext-link-type="DOI">10.1038/ngeo1188</ext-link>, 2011.</mixed-citation></ref>
      <ref id="bib1.bibx73"><label>Jenkins(1991)</label><mixed-citation>Jenkins, A.: A One-Dimensional Model of Ice Shelf-Ocean Interaction, J.
Geophys. Res.-Oceans, 96, 20671–20677, <ext-link xlink:href="https://doi.org/10.1029/91JC01842" ext-link-type="DOI">10.1029/91JC01842</ext-link>, 1991.</mixed-citation></ref>
      <ref id="bib1.bibx74"><label>Jenkins(2011)</label><mixed-citation>Jenkins, A.: Convection-Driven Melting near the Grounding Lines of Ice
Shelves and Tidewater Glaciers, J. Phys. Oceanogr., 41, 2279–2294,
<ext-link xlink:href="https://doi.org/10.1175/JPO-D-11-03.1" ext-link-type="DOI">10.1175/JPO-D-11-03.1</ext-link>, 2011.</mixed-citation></ref>
      <ref id="bib1.bibx75"><label>Jenkins(2016)</label><mixed-citation>Jenkins, A.: A Simple Model of the Ice Shelf–Ocean Boundary Layer and
Current, J. Phys. Oceanogr., 46, 1785–1803, <ext-link xlink:href="https://doi.org/10.1175/JPO-D-15-0194.1" ext-link-type="DOI">10.1175/JPO-D-15-0194.1</ext-link>,
2016.</mixed-citation></ref>
      <ref id="bib1.bibx76"><label>Jenkins and Jacobs(2008)</label><mixed-citation>Jenkins, A. and Jacobs, S. S.: Circulation and melting beneath George VI Ice
Shelf, Antarctica, J. Geophys. Res., 113, 1–18, <ext-link xlink:href="https://doi.org/10.1029/2007JC004449" ext-link-type="DOI">10.1029/2007JC004449</ext-link>,
2008.</mixed-citation></ref>
      <ref id="bib1.bibx77"><label>Jenkins et al.(2010)</label><mixed-citation>Jenkins, A., Dutrieux, P., Jacobs, S. S., McPhail, S. D., Perrett, J. R.,
Webb, A. T., and White, D.: Observations beneath Pine Island<?pagebreak page1289?> Glacier in West
Antarctica and implications for its retreat, Nat. Geosci., 3, 468–472,
<ext-link xlink:href="https://doi.org/10.1038/ngeo890" ext-link-type="DOI">10.1038/ngeo890</ext-link>, 2010.</mixed-citation></ref>
      <ref id="bib1.bibx78"><label>Joughin and Alley(2011)</label><mixed-citation>Joughin, I. and Alley, R. B.: Stability of the West Antarctic ice sheet in a
warming world, Nat. Geosci., 4, 506–513, <ext-link xlink:href="https://doi.org/10.1038/ngeo1194" ext-link-type="DOI">10.1038/ngeo1194</ext-link>, 2011.</mixed-citation></ref>
      <ref id="bib1.bibx79"><label>Joughin and Padman(2003)</label><mixed-citation>Joughin, I. and Padman, L.: Melting and freezing beneath Filchner-Ronne Ice
Shelf, Antarctica, Geophys. Res. Lett., 30, 1477,
<ext-link xlink:href="https://doi.org/10.1029/2003GL016941" ext-link-type="DOI">10.1029/2003GL016941</ext-link>, 2003.</mixed-citation></ref>
      <ref id="bib1.bibx80"><label>Khazendar et al.(2013)</label><mixed-citation>Khazendar, A., Schodlok, M. P., Fenty, I. G., Ligtenberg, S. R. M., Rignot,
E. J., and van den Broeke, M. R.: Observed thinning of Totten Glacier is
linked to coastal polynya variability, Nat. Commun., 4, 2857,
<ext-link xlink:href="https://doi.org/10.1038/ncomms3857" ext-link-type="DOI">10.1038/ncomms3857</ext-link>, 2013.</mixed-citation></ref>
      <ref id="bib1.bibx81"><label>Kjellsson et al.(2015)</label><mixed-citation>Kjellsson, J., Holland, P. R., Marshall, G. J., Mathiot, P., Aksenov, Y.,
Coward, A. C., Bacon, S., Megann, A. P., and Ridley, J. K.: Model
sensitivity of the Weddell and Ross seas, Antarctica, to vertical mixing and
freshwater forcing, Ocean Model., 94, 141–152,
<ext-link xlink:href="https://doi.org/10.1016/j.ocemod.2015.08.003" ext-link-type="DOI">10.1016/j.ocemod.2015.08.003</ext-link>, 2015.</mixed-citation></ref>
      <ref id="bib1.bibx82"><label>Koltermann et al.(2011)</label><mixed-citation>
Koltermann, K. P., Gouretski, V. V., and Jancke, K.: Hydrographic Atlas of
the World Ocean Circulation Experiment (WOCE), vol. 3, in: Atlantic Ocean
edited by: Sparrow, M., Chapman, P., and Gould, J., Tech. rep., International
WOCE Project Office, Southampton, UK, 2011.</mixed-citation></ref>
      <ref id="bib1.bibx83"><label>Kurtz and Markus(2012)</label><mixed-citation>Kurtz, N. T. and Markus, T.: Satellite observations of Antarctic sea ice
thickness and volume, J. Geophys. Res.-Oceans, 117, 1–9,
<ext-link xlink:href="https://doi.org/10.1029/2012JC008141" ext-link-type="DOI">10.1029/2012JC008141</ext-link>, 2012.</mixed-citation></ref>
      <ref id="bib1.bibx84"><label>Kusahara et al.(2010)</label><mixed-citation>Kusahara, K., Hasumi, H., and Tamura, T.: Modeling sea ice production and
dense shelf water formation in coastal polynyas around East Antarctica, J.
Geophys. Res.-Oceans, 115, C10006, <ext-link xlink:href="https://doi.org/10.1029/2010JC006133" ext-link-type="DOI">10.1029/2010JC006133</ext-link>, 2010.</mixed-citation></ref>
      <ref id="bib1.bibx85"><label>Kusahara et al.(2017)</label><mixed-citation>Kusahara, K., Williams, G. D., Massom, R., Reid, P., and Hasumi, H.: Roles
of wind stress and thermodynamic forcing in recent trends in Antarctic sea
ice and Southern Ocean SST: An ocean-sea ice model study, Global Planet.
Change, 158, 103–118, <ext-link xlink:href="https://doi.org/10.1016/j.gloplacha.2017.09.012" ext-link-type="DOI">10.1016/j.gloplacha.2017.09.012</ext-link>, 2017.</mixed-citation></ref>
      <ref id="bib1.bibx86"><label>Kwok and Cunningham(2008)</label><mixed-citation>Kwok, R. and Cunningham, G. F.: ICESat over Arctic sea ice: Estimation of
snow depth and ice thickness, J. Geophys. Res.-Oceans, 113, 1–17,
<ext-link xlink:href="https://doi.org/10.1029/2008JC004753" ext-link-type="DOI">10.1029/2008JC004753</ext-link>, 2008.</mixed-citation></ref>
      <ref id="bib1.bibx87"><label>Langhorne et al.(2015)</label><mixed-citation>Langhorne, P. J., Hughes, K. G., Gough, A. J., Smith, I. J., Williams, M.
J. M., Robinson, N. J., Stevens, C. L., Rack, W., Price, D., Leonard, G. H.,
Mahoney, A. R., Haas, C., and Haskell, T. G.: Observed platelet ice
distributions in Antarctic sea ice: An index for ocean-ice shelf heat flux,
Geophys. Res. Lett., 42, 5442–5451, <ext-link xlink:href="https://doi.org/10.1002/2015GL064508" ext-link-type="DOI">10.1002/2015GL064508</ext-link>, 2015.</mixed-citation></ref>
      <ref id="bib1.bibx88"><label>Langley et al.(2014)</label><mixed-citation>Langley, K. A., Kohler, J., Sinisalo, A., Øyan, M. J., Hamran, S. E.,
Hattermann, T., Matsuoka, K., Nøst, O. A., and Isaksson, E.: Low melt
rates with seasonal variability at the base of Fimbul Ice Shelf, East
Antarctica, revealed by in situ interferometric radar measurements, Geophys.
Res. Lett., 41, 8138–8146, <ext-link xlink:href="https://doi.org/10.1002/2014GL061782" ext-link-type="DOI">10.1002/2014GL061782</ext-link>, 2014.</mixed-citation></ref>
      <ref id="bib1.bibx89"><label>Large et al.(1994)</label><mixed-citation>Large, W. G., McWilliams, J. C., and Doney, S. C.: Oceanic Vertical Mixing
– a Review and a Model with a Nonlocal Boundary-Layer Parameterization,
Rev. Geophys., 32, 363–403, <ext-link xlink:href="https://doi.org/10.1029/94rg01872" ext-link-type="DOI">10.1029/94rg01872</ext-link>, 1994.</mixed-citation></ref>
      <ref id="bib1.bibx90"><label>Larson et al.(2005)</label><mixed-citation>Larson, J., Jacob, R., and Ong, E.: The Model Coupling Toolkit: A New
Fortran90 Toolkit for Building Multiphysics Parallel Coupled Models, Int. J.
High Perform. C., 19, 277–292, <ext-link xlink:href="https://doi.org/10.1177/1094342005056115" ext-link-type="DOI">10.1177/1094342005056115</ext-link>, 2005.</mixed-citation></ref>
      <ref id="bib1.bibx91"><?xmltex \def\ref@label{{Lemari{\'{e}} et~al.(2012)}}?><label>Lemarié et al.(2012)</label><mixed-citation>Lemarié, F., Kurian, J., Shchepetkin, A. F., Jeroen Molemaker, M.,
Colas, F., and McWilliams, J. C.: Are there inescapable issues prohibiting
the use of terrain-following coordinates in climate models?, Ocean Model.,
42, 57–79, <ext-link xlink:href="https://doi.org/10.1016/j.ocemod.2011.11.007" ext-link-type="DOI">10.1016/j.ocemod.2011.11.007</ext-link>, 2012.</mixed-citation></ref>
      <ref id="bib1.bibx92"><label>Lenaerts et al.(2012)</label><mixed-citation>Lenaerts, J. T. M., van den Broeke, M. R., van de Berg, W. J., van Meijgaard,
E., and Munneke, P. K.: A new, high-resolution surface mass balance map of
Antarctica (1979–2010) based on regional atmospheric climate modeling,
Geophys. Res. Lett., 39, 1–5, <ext-link xlink:href="https://doi.org/10.1029/2011GL050713" ext-link-type="DOI">10.1029/2011GL050713</ext-link>, 2012.</mixed-citation></ref>
      <ref id="bib1.bibx93"><label>Lewis and Perkin(1986)</label><mixed-citation>Lewis, E. L. and Perkin, R. G.: Ice pumps and their rates, J. Geophys.
Res., 91, 11756–11762 , <ext-link xlink:href="https://doi.org/10.1029/JC091iC10p11756" ext-link-type="DOI">10.1029/JC091iC10p11756</ext-link>, 1986.</mixed-citation></ref>
      <ref id="bib1.bibx94"><label>Lipscomb and Hunke(2004)</label><mixed-citation>Lipscomb, W. H. and Hunke, E. C.: Modeling Sea Ice Transport Using
Incremental Remapping, Mon. Weather Rev., 132, 1341–1354,
<ext-link xlink:href="https://doi.org/10.1175/1520-0493(2004)132&lt;1341:MSITUI&gt;2.0.CO;2" ext-link-type="DOI">10.1175/1520-0493(2004)132&lt;1341:MSITUI&gt;2.0.CO;2</ext-link>, 2004.</mixed-citation></ref>
      <ref id="bib1.bibx95"><label>Lipscomb et al.(2007)</label><mixed-citation>Lipscomb, W. H., Hunke, E. C., Maslowski, W., and Jakacki, J.: Ridging,
strength, and stability in high-resolution sea ice models, J. Geophys.
Res.-Oceans, 112, 1–18, <ext-link xlink:href="https://doi.org/10.1029/2005JC003355" ext-link-type="DOI">10.1029/2005JC003355</ext-link>, 2007.</mixed-citation></ref>
      <ref id="bib1.bibx96"><label>Makinson et al.(2011)</label><mixed-citation>Makinson, K., Holland, P. R., Jenkins, A., Nicholls, K. W., and Holland,
D. M.: Influence of tides on melting and freezing beneath Filchner-Ronne Ice
Shelf, Antarctica, Geophys. Res. Lett., 38, L06601,
<ext-link xlink:href="https://doi.org/10.1029/2010GL046462" ext-link-type="DOI">10.1029/2010GL046462</ext-link>, 2011.</mixed-citation></ref>
      <ref id="bib1.bibx97"><label>Marchesiello et al.(2001)</label><mixed-citation>Marchesiello, P., McWilliams, J. C., and Shchepetkin, A. F.: Open boundary
conditions for long-term integration of regional oceanic models, Ocean
Model., 3, 1–20, <ext-link xlink:href="https://doi.org/10.1016/S1463-5003(00)00013-5" ext-link-type="DOI">10.1016/S1463-5003(00)00013-5</ext-link>, 2001.</mixed-citation></ref>
      <ref id="bib1.bibx98"><label>Marchesiello et al.(2009)</label><mixed-citation>Marchesiello, P., Debreu, L., and Couvelard, X.: Spurious diapycnal mixing
in terrain-following coordinate models: The problem and a solution, Ocean
Model., 26, 156–169, <ext-link xlink:href="https://doi.org/10.1016/j.ocemod.2008.09.004" ext-link-type="DOI">10.1016/j.ocemod.2008.09.004</ext-link>, 2009.</mixed-citation></ref>
      <ref id="bib1.bibx99"><label>Martin and Adcroft(2010)</label><mixed-citation>Martin, T. and Adcroft, A.: Parameterizing the fresh-water flux from land
ice to ocean with interactive icebergs in a coupled climate model, Ocean
Model., 34, 111–124, <ext-link xlink:href="https://doi.org/10.1016/j.ocemod.2010.05.001" ext-link-type="DOI">10.1016/j.ocemod.2010.05.001</ext-link>, 2010.</mixed-citation></ref>
      <ref id="bib1.bibx100"><label>Martinson and McKee(2012)</label><mixed-citation>
Martinson, D. G. and McKee, D. C.: Transport of warm Upper Circumpolar Deep
Water onto the western Antarctic Peninsula continental shelf, Ocean Sci., 8,
433–442, https://doi.org/10.5194/os-8-433-2012, 2012.</mixed-citation></ref>
      <ref id="bib1.bibx101"><label>Mathiot et al.(2017)</label><mixed-citation>Mathiot, P., Jenkins, A., Harris, C., and Madec, G.: Explicit representation
and parametrised impacts of under ice shelf seas in the <inline-formula><mml:math id="M380" display="inline"><mml:mrow><mml:msup><mml:mi>z</mml:mi><mml:mo>*</mml:mo></mml:msup></mml:mrow></mml:math></inline-formula> coordinate
ocean model NEMO 3.6, Geosci. Model Dev., 10, 2849–2874,
https://doi.org/10.5194/gmd-10-2849-2017, 2017.</mixed-citation></ref>
      <ref id="bib1.bibx102"><label>McPhail et al.(2009)</label><mixed-citation>McPhail, S. D., Furlong, M. E., Pebody, M., Perrett, J. R., Stevenson, P.,
Webb, A., and White, D.: Exploring beneath the PIG Ice Shelf with the
Autosub3 AUV, in: OCEANS '09 IEEE Bremen: Balancing Technology with Future
Needs, Piscataway, USA, p. 6, <ext-link xlink:href="https://doi.org/10.1109/OCEANSE.2009.5278170" ext-link-type="DOI">10.1109/OCEANSE.2009.5278170</ext-link>, 2009.</mixed-citation></ref>
      <ref id="bib1.bibx103"><label>McPhee et al.(1987)</label><mixed-citation>McPhee, M. G., Maykut, G. A., and Morison, J. H.: Dynamics and
thermodynamics of the ice/upper ocean system in the marginal ice zone of the
Greenland Sea, J. Geophys. Res.-Oceans, 92, 7017–7031,
<ext-link xlink:href="https://doi.org/10.1029/JC092iC07p07017" ext-link-type="DOI">10.1029/JC092iC07p07017</ext-link>, 1987.</mixed-citation></ref>
      <ref id="bib1.bibx104"><label>Meehl et al.(2000)</label><mixed-citation>Meehl, G. A., Boer, G. J., Covey, C., Latif, M., and Stouffer, R. J.: The
Coupled Model Intercomparison Project (CMIP), B. Am. Meteorol. Soc., 81,
313–318, <ext-link xlink:href="https://doi.org/10.1175/1520-0477(2000)081&lt;0313:TCMIPC&gt;2.3.CO;2" ext-link-type="DOI">10.1175/1520-0477(2000)081&lt;0313:TCMIPC&gt;2.3.CO;2</ext-link>, 2000.</mixed-citation></ref>
      <?pagebreak page1290?><ref id="bib1.bibx105"><label>Meier et al.(2013)</label><mixed-citation>Meier, W., Fetterrer, F., Savoie, M., Mallory, S., Duerr, R., and Stroeve,
J.: NOAA/NSIDC Climate Data Record of Passive Microwave Sea Ice
Concentration, Version 2, <ext-link xlink:href="https://doi.org/10.7265/N55M63M1" ext-link-type="DOI">10.7265/N55M63M1</ext-link>, 2013.</mixed-citation></ref>
      <ref id="bib1.bibx106"><label>Menemenlis et al.(2008)</label><mixed-citation>
Menemenlis, D., Campin, J.-M., Heimbach, P., Hill, C., Lee, T., Nguyen,
A. T., Schodlok, M. P., and Zhang, H.: ECCO2: High resolution global ocean
and sea ice data synthesis, Mercator Ocean Quarterly Newsletter, 31, 13–21,
2008.</mixed-citation></ref>
      <ref id="bib1.bibx107"><label>Mengel and Levermann(2014)</label><mixed-citation>Mengel, M. and Levermann, A.: Ice plug prevents irreversible discharge from
East Antarctica, Nat. Clim. Change, 4, 451–455, <ext-link xlink:href="https://doi.org/10.1038/nclimate2226" ext-link-type="DOI">10.1038/nclimate2226</ext-link>,
2014.</mixed-citation></ref>
      <ref id="bib1.bibx108"><label>Miller et al.(2012)</label><mixed-citation>Miller, K. G., Wright, J. D., Browning, J. V., Kulpecz, A., Kominz, M.,
Naish, T. R., Cramer, B. S., Rosenthal, Y., Peltier, W. R., and Sosdian, S.:
High tide of the warm Pliocene: Implications of global sea level for
Antarctic deglaciation, Geology, 40, 407–410, <ext-link xlink:href="https://doi.org/10.1130/G32869.1" ext-link-type="DOI">10.1130/G32869.1</ext-link>, 2012.</mixed-citation></ref>
      <ref id="bib1.bibx109"><label>Moffat et al.(2009)</label><mixed-citation>Moffat, C., Owens, B., and Beardsley, R. C.: On the characteristics of
Circumpolar Deep Water intrusions to the west Antarctic Peninsula Continental
Shelf, J. Geophys. Res.-Oceans, 114, C05017, <ext-link xlink:href="https://doi.org/10.1029/2008JC004955" ext-link-type="DOI">10.1029/2008JC004955</ext-link>,
2009.</mixed-citation></ref>
      <ref id="bib1.bibx110"><label>Mueller et al.(2012)</label><mixed-citation>Mueller, R. D., Padman, L., Dinniman, M. S., Erofeeva, S. Y., Fricker, H. A.,
and King, M. A.: Impact of tide-topography interactions on basal melting of
Larsen C Ice Shelf, Antarctica, J. Geophys. Res.-Oceans, 117, C05005,
<ext-link xlink:href="https://doi.org/10.1029/2011JC007263" ext-link-type="DOI">10.1029/2011JC007263</ext-link>, 2012.</mixed-citation></ref>
      <ref id="bib1.bibx111"><label>Mueller et al.(2018)</label><mixed-citation>
Mueller, R. D., Hattermann, T., Howard, S. L., and Padman, L.: Tidal
influences on a future evolution of the Filchner–Ronne Ice Shelf cavity in
the Weddell Sea, Antarctica, The Cryosphere, 12, 453–476,
https://doi.org/10.5194/tc-12-453-2018, 2018.</mixed-citation></ref>
      <ref id="bib1.bibx112"><?xmltex \def\ref@label{{M{\"{u}}ller et~al.(2010)}}?><label>Müller et al.(2010)</label><mixed-citation>Müller, M., Haak, H., Jungclaus, J. H., Sündermann, J., and
Thomas, M.: The effect of ocean tides on a climate model simulation, Ocean
Model., 35, 304–313, <ext-link xlink:href="https://doi.org/10.1016/j.ocemod.2010.09.001" ext-link-type="DOI">10.1016/j.ocemod.2010.09.001</ext-link>, 2010.</mixed-citation></ref>
      <ref id="bib1.bibx113"><label>Nakayama et al.(2014)</label><mixed-citation>Nakayama, Y., Timmermann, R., Schröder, M., and Hellmer, H. H.: On the
difficulty of modeling Circumpolar Deep Water intrusions onto the Amundsen
Sea continental shelf, Ocean Model., 84, 26–34,
<ext-link xlink:href="https://doi.org/10.1016/j.ocemod.2014.09.007" ext-link-type="DOI">10.1016/j.ocemod.2014.09.007</ext-link>, 2014.</mixed-citation></ref>
      <ref id="bib1.bibx114"><label>Naud et al.(2014)</label><mixed-citation>Naud, C. M., Booth, J. F., and Del Genio, A. D.: Evaluation of ERA-Interim
and MERRA cloudiness in the Southern Ocean, J. Climate, 27, 2109–2124,
<ext-link xlink:href="https://doi.org/10.1175/JCLI-D-13-00432.1" ext-link-type="DOI">10.1175/JCLI-D-13-00432.1</ext-link>, 2014.</mixed-citation></ref>
      <ref id="bib1.bibx115"><label>Naughten et al.(2017)</label><mixed-citation>Naughten, K. A., Galton-Fenzi, B. K., Meissner, K. J., England, M. H.,
Brassington, G. B., Colberg, F., Hattermann, T., and Debernard, J. B.:
Spurious sea ice formation caused by oscillatory ocean tracer advection
schemes, Ocean Model., 116, 108–117, <ext-link xlink:href="https://doi.org/10.1016/j.ocemod.2017.06.010" ext-link-type="DOI">10.1016/j.ocemod.2017.06.010</ext-link>,
2017.</mixed-citation></ref>
      <ref id="bib1.bib1"><label>1</label><mixed-citation>Naughten, K. A., Meissner, K. J., Galton-Fenzi, B. K., England, M. H., Timmermann, R., and Hellmer, H. H.: MetROMS-iceshelf, <ext-link xlink:href="https://doi.org/10.5281/zenodo.1157230" ext-link-type="DOI">10.5281/zenodo.1157230</ext-link>, 2018a.</mixed-citation></ref>
      <ref id="bib1.bib2"><label>2</label><mixed-citation>Naughten, K. A., Meissner, K. J., Galton-Fenzi, B. K., England, M. H., Timmermann, R., and Hellmer, H. H.: Finite Element Sea-ice/ice-shelf Ocean Model (FESOM) 1.4, <ext-link xlink:href="https://doi.org/10.5281/zenodo.1157228" ext-link-type="DOI">10.5281/zenodo.1157228</ext-link>, 2018b.</mixed-citation></ref>
      <ref id="bib1.bibx116"><label>Nicholls and Johnson(2001)</label><mixed-citation>Nicholls, K. W. and Johnson, M. R.: Oceanographic conditions south of
Berkner Island, beneath Filchner-Ronne Ice Shelf, Antarctica, J. Geophys.
Res., 106, 11481–11492, <ext-link xlink:href="https://doi.org/10.1029/2000JC000350" ext-link-type="DOI">10.1029/2000JC000350</ext-link>, 2001.</mixed-citation></ref>
      <ref id="bib1.bibx117"><?xmltex \def\ref@label{{Nicholls and {\O}sterhus(2004)}}?><label>Nicholls and Østerhus(2004)</label><mixed-citation>Nicholls, K. W. and Østerhus, S.: Interannual variability and ventilation
timescales in the ocean cavity beneath Filchner-Ronne Ice Shelf, Antarctica,
J. Geophys. Res.-Oceans, 109, 1–9, <ext-link xlink:href="https://doi.org/10.1029/2003JC002149" ext-link-type="DOI">10.1029/2003JC002149</ext-link>, 2004.</mixed-citation></ref>
      <ref id="bib1.bibx118"><label>Nicholls et al.(2004)</label><mixed-citation>Nicholls, K. W., Makinson, K., and Østerhus, S.: Circulation and water
masses beneath the northern Ronne Ice Shelf, Antarctica, J. Geophys.
Res.-Oceans, 109, 1–11, <ext-link xlink:href="https://doi.org/10.1029/2004JC002302" ext-link-type="DOI">10.1029/2004JC002302</ext-link>, 2004.</mixed-citation></ref>
      <ref id="bib1.bibx119"><label>Nicholls et al.(2006)</label><mixed-citation>Nicholls, K. W., Abrahamsen, E. P., Buck, J. J. H., Dodd, P. A., Goldblatt,
C., Griffiths, G., Heywood, K. J., Hughes, N. E., Kaletzky, A., Lane-Serff,
G. F., McPhail, S. D., Millard, N. W., Oliver, K. I. C., Perrett, J., Price,
M. R., Pudsey, C. J., Saw, K., Stansfield, K., Stott, M. J., Wadhams, P.,
Webb, A. T., and Wilkinson, J. P.: Measurements beneath an Antarctic ice
shelf using an autonomous underwater vehicle, Geophys. Res. Lett., 33, 2–5,
<ext-link xlink:href="https://doi.org/10.1029/2006GL025998" ext-link-type="DOI">10.1029/2006GL025998</ext-link>, 2006.</mixed-citation></ref>
      <ref id="bib1.bibx120"><label>Nicholls et al.(2009)</label><mixed-citation>Nicholls, K. W., Østerhus, S., Makinson, K., Gammelsrød, T., and
Fahrbach, E.: Ice-ocean processes over the continental shelf of the southern
Weddell Sea, Antarctica: a review, Rev. Geophys., 47, 1–23,
<ext-link xlink:href="https://doi.org/10.1029/2007RG000250" ext-link-type="DOI">10.1029/2007RG000250</ext-link>, 2009.</mixed-citation></ref>
      <ref id="bib1.bibx121"><label>O'Leary et al.(2013)</label><mixed-citation>O'Leary, M. J., Hearty, P. J., Thompson, W. G., Raymo, M. E., Mitrovica,
J. X., Webster, J. M., and O'Leary, M. J.: Ice sheet collapse following a
prolonged period of stable sea level during the last interglacial, Nat.
Geosci., 6, 796–800, <ext-link xlink:href="https://doi.org/10.1038/ngeo1890" ext-link-type="DOI">10.1038/ngeo1890</ext-link>, 2013.</mixed-citation></ref>
      <ref id="bib1.bibx122"><label>Pacanowski and Philander(1981)</label><mixed-citation>Pacanowski, R. C. and Philander, S. G. H.: Parameterization of Vertical
Mixing in Numerical Models of Tropical Oceans, J. Phys. Oceanogr., 11,
1443–1451, <ext-link xlink:href="https://doi.org/10.1175/1520-0485(1981)011&lt;1443:POVMIN&gt;2.0.CO;2" ext-link-type="DOI">10.1175/1520-0485(1981)011&lt;1443:POVMIN&gt;2.0.CO;2</ext-link>, 1981.</mixed-citation></ref>
      <ref id="bib1.bibx123"><label>Padman et al.(2009)Padman, Howard, Orsi, and Muench</label><mixed-citation>Padman, L., Howard, S. L., Orsi, A. H., and Muench, R. D.: Tides of the
northwestern Ross Sea and their impact on dense outflows of Antarctic Bottom
Water, Deep-Sea Res. Pt. II, 56, 818–834, <ext-link xlink:href="https://doi.org/10.1016/j.dsr2.2008.10.026" ext-link-type="DOI">10.1016/j.dsr2.2008.10.026</ext-link>,
2009.</mixed-citation></ref>
      <ref id="bib1.bibx124"><label>Paolo et al.(2015)</label><mixed-citation>Paolo, F. S., Fricker, H. A., and Padman, L.: Volume loss from Antarctic ice
shelves is accelerating, Science, 348, 327–331,
<ext-link xlink:href="https://doi.org/10.1126/science.aaa0940" ext-link-type="DOI">10.1126/science.aaa0940</ext-link>, 2015.</mixed-citation></ref>
      <ref id="bib1.bibx125"><label>Parkinson and Washington(1979)</label><mixed-citation>Parkinson, C. L. and Washington, W. M.: A large-scale numerical model of sea
ice, J. Geophys. Res., 84, 311–337, <ext-link xlink:href="https://doi.org/10.1029/JC084iC01p00311" ext-link-type="DOI">10.1029/JC084iC01p00311</ext-link>, 1979.</mixed-citation></ref>
      <ref id="bib1.bibx126"><label>Pellichero et al.(2017)</label><mixed-citation>Pellichero, V., Sallée, J.-B., Schmidtko, S., Roquet, F., and
Charrassin, J.-B.: The ocean mixed layer under Southern Ocean sea-ice:
Seasonal cycle and forcing, J. Geophys. Res.-Oceans, 122, 1608–1633,
<ext-link xlink:href="https://doi.org/10.1002/2016JC011970" ext-link-type="DOI">10.1002/2016JC011970</ext-link>, 2017.</mixed-citation></ref>
      <ref id="bib1.bibx127"><label>Petty et al.(2013)</label><mixed-citation>Petty, A. A., Feltham, D. L., and Holland, P. R.: Impact of Atmospheric
Forcing on Antarctic Continental Shelf Water Masses, J. Phys. Oceanogr., 43,
920–940, <ext-link xlink:href="https://doi.org/10.1175/JPO-D-12-0172.1" ext-link-type="DOI">10.1175/JPO-D-12-0172.1</ext-link>, 2013.</mixed-citation></ref>
      <ref id="bib1.bibx128"><label>Petty et al.(2014)Petty, Holland, and Feltham</label><mixed-citation>
Petty, A. A., Holland, P. R., and Feltham, D. L.: Sea ice and the ocean mixed
layer over the Antarctic shelf seas, The Cryosphere, 8, 761–783,
https://doi.org/10.5194/tc-8-761-2014, 2014.</mixed-citation></ref>
      <ref id="bib1.bibx129"><label>Pritchard et al.(2012)</label><mixed-citation>Pritchard, H. D., Ligtenberg, S. R. M., Fricker, H. A., Vaughan, D. G.,
van den Broeke, M. R., and Padman, L.: Antarctic ice-sheet loss driven by
basal melting of ice shelves, Nature, 484, 502–505,
<ext-link xlink:href="https://doi.org/10.1038/nature10968" ext-link-type="DOI">10.1038/nature10968</ext-link>, 2012.</mixed-citation></ref>
      <ref id="bib1.bibx130"><label>Rack and Rott(2004)</label><mixed-citation>Rack, W. and Rott, H.: Pattern of retreat and disintegration of the Larsen B
Ice Shelf, Ann. Glaciol., 39, 505–510, <ext-link xlink:href="https://doi.org/10.3189/172756404781814005" ext-link-type="DOI">10.3189/172756404781814005</ext-link>,
2004.</mixed-citation></ref>
      <ref id="bib1.bibx131"><label>Raymo and Mitrovica(2012)</label><mixed-citation>Raymo, M. E. and Mitrovica, J. X.: Collapse of polar ice sheets during the
stage 11 interglacial, Nature, 483, 453–456, <ext-link xlink:href="https://doi.org/10.1038/nature10891" ext-link-type="DOI">10.1038/nature10891</ext-link>,
2012.</mixed-citation></ref>
      <?pagebreak page1291?><ref id="bib1.bibx132"><label>Reddy et al.(2010)Reddy, Holland, and Arrigo</label><mixed-citation>Reddy, T. E., Holland, D. M., and Arrigo, K. R.: Ross ice shelf cavity
circulation, residence time, and melting: Results from a model of oceanic
chlorofluorocarbons, Cont. Shelf Res., 30, 733–742,
<ext-link xlink:href="https://doi.org/10.1016/j.csr.2010.01.007" ext-link-type="DOI">10.1016/j.csr.2010.01.007</ext-link>, 2010.</mixed-citation></ref>
      <ref id="bib1.bibx133"><label>Rignot et al.(2011)</label><mixed-citation>Rignot, E. J., Velicogna, I., van den Broeke, M. R., Monaghan, A., and
Lenaerts, J. T. M.: Acceleration of the contribution of the Greenland and
Antarctic ice sheets to sea level rise, Geophys. Res. Lett., 38, 1–5,
<ext-link xlink:href="https://doi.org/10.1029/2011GL046583" ext-link-type="DOI">10.1029/2011GL046583</ext-link>, 2011.</mixed-citation></ref>
      <ref id="bib1.bibx134"><label>Rignot et al.(2013)</label><mixed-citation>Rignot, E. J., Jacobs, S. S., Mouginot, J., and Scheuchl, B.: Ice-Shelf
Melting Around Antarctica, Science, 341, 266–270,
<ext-link xlink:href="https://doi.org/10.1126/science.1235798" ext-link-type="DOI">10.1126/science.1235798</ext-link>, 2013.</mixed-citation></ref>
      <ref id="bib1.bibx135"><label>Rignot et al.(2014)</label><mixed-citation>Rignot, E. J., Mouginot, J., Morlighem, M., Seroussi, H., and Scheuchl, B.:
Widespread, rapid grounding line retreat of Pine Island, Thwaites, Smith,
and Kohler glaciers, West Antarctica, from 1992 to 2011, Geophys. Res.
Lett., 41, 3502–3509, <ext-link xlink:href="https://doi.org/10.1002/2014GL060140" ext-link-type="DOI">10.1002/2014GL060140</ext-link>, 2014.</mixed-citation></ref>
      <ref id="bib1.bibx136"><label>Rintoul et al.(2010)</label><mixed-citation>Rintoul, S. R., Speer, K., Sparrow, M., Meredith, M., Hofmann, E., Fahrbach,
E., Summerhayes, C., Worby, A., England, M., Bellerby, R. G., Speich, S.,
Costa, D., Hall, J., Hindell, M., Hosie, G., Stansfield, K., Fukamachi, Y.,
de Bruin, T., Naveira Garabato, A., Alverson, K., Ryabinin, V., Shin,
H. C., and Gladyshev, S.: Southern Ocean Observing System (SOOS): Rationale
and strategy for sustained observations of the Southern Ocean, in:
Proceedings of OceanObs'09: Sustained Ocean Observations and Information for
Society, edited by: Hall, J., Harrison, D. E., and Stammer, D., vol. 2, ESA
Publication WPP-306, Venice, <ext-link xlink:href="https://doi.org/10.5270/OceanObs09.cwp.74" ext-link-type="DOI">10.5270/OceanObs09.cwp.74</ext-link>, 2010.</mixed-citation></ref>
      <ref id="bib1.bibx137"><label>Rintoul et al.(2016)</label><mixed-citation>Rintoul, S. R., Silvano, A., Pena-Molino, B., van Wijk, E., Rosenberg, M.,
Greenbaum, J. S., and Blankenship, D. D.: Ocean heat drives rapid basal melt
of the Totten Ice Shelf, Sci. Adv., 2, 1–6, <ext-link xlink:href="https://doi.org/10.1126/sciadv.1601610" ext-link-type="DOI">10.1126/sciadv.1601610</ext-link>,
2016.</mixed-citation></ref>
      <ref id="bib1.bibx138"><label>Robertson et al.(1998)</label><mixed-citation>Robertson, R., Padman, L., and Egbert, G. D.: Tides in the Weddell Sea, in:
Ocean, Ice, and Atmosphere: Interactions at the Antarctic Continental Margin,
edited by: Jacobs, S. S. and Weiss, R. F., American Geophysical Union,
Washington, DC, 341–369, <ext-link xlink:href="https://doi.org/10.1029/AR075p0341" ext-link-type="DOI">10.1029/AR075p0341</ext-link>, 1998.</mixed-citation></ref>
      <ref id="bib1.bibx139"><label>Rothrock(1975)</label><mixed-citation>
Rothrock, D. A.: The Energetics of the Plastic Deformation of Pack Ice by
Ridging, J. Geophys. Res., 80, 4514–4519, 1975.</mixed-citation></ref>
      <ref id="bib1.bibx140"><label>Rott et al.(1996)Rott, Skvarca, and Nagler</label><mixed-citation>Rott, H., Skvarca, P., and Nagler, T.: Rapid Collapse of Northern Larsen Ice
Shelf, Antarctica, Science, 271, 788–792,
<ext-link xlink:href="https://doi.org/10.1126/science.271.5250.788" ext-link-type="DOI">10.1126/science.271.5250.788</ext-link>, 1996.</mixed-citation></ref>
      <ref id="bib1.bibx141"><?xmltex \def\ref@label{{Sall{\'{e}}e et~al.(2013)}}?><label>Sallée et al.(2013)</label><mixed-citation>Sallée, J.-B., Shuckburgh, E., Bruneau, N., Meijers, A. J. S.,
Bracegirdle, T. J., and Wang, Z.: Assessment of Southern Ocean mixed-layer
depths in CMIP5 models: Historical bias and forcing response, J. Geophys.
Res.-Oceans, 118, 1845–1862, <ext-link xlink:href="https://doi.org/10.1002/jgrc.20157" ext-link-type="DOI">10.1002/jgrc.20157</ext-link>, 2013.</mixed-citation></ref>
      <ref id="bib1.bibx142"><label>Schmidtko et al.(2014)</label><mixed-citation>Schmidtko, S., Heywood, K. J., Thompson, A. F., and Aoki, S.: Multidecadal
warming of Antarctic waters, Science, 346, 1227–1232,
<ext-link xlink:href="https://doi.org/10.1126/science.1256117" ext-link-type="DOI">10.1126/science.1256117</ext-link>, 2014.</mixed-citation></ref>
      <ref id="bib1.bibx143"><label>Semtner(1976)</label><mixed-citation>Semtner, A. J.: A Model for the Thermodynamic Growth of Sea Ice in Numerical
Investigations of Climate, J. Phys. Oceanogr., 6, 379–389,
<ext-link xlink:href="https://doi.org/10.1175/1520-0485(1976)006&lt;0379:AMFTTG&gt;2.0.CO;2" ext-link-type="DOI">10.1175/1520-0485(1976)006&lt;0379:AMFTTG&gt;2.0.CO;2</ext-link>, 1976.</mixed-citation></ref>
      <ref id="bib1.bibx144"><label>Shchepetkin(2003)</label><mixed-citation>Shchepetkin, A. F.: A method for computing horizontal pressure-gradient
force in an oceanic model with a nonaligned vertical coordinate, J. Geophys.
Res., 108, 3090, <ext-link xlink:href="https://doi.org/10.1029/2001JC001047" ext-link-type="DOI">10.1029/2001JC001047</ext-link>, 2003.</mixed-citation></ref>
      <ref id="bib1.bibx145"><label>Shchepetkin and McWilliams(2005)</label><mixed-citation>Shchepetkin, A. F. and McWilliams, J. C.: The regional oceanic modeling
system (ROMS): A split-explicit, free-surface,
topography-following-coordinate oceanic model, Ocean Model., 9, 347–404,
<ext-link xlink:href="https://doi.org/10.1016/j.ocemod.2004.08.002" ext-link-type="DOI">10.1016/j.ocemod.2004.08.002</ext-link>, 2005.</mixed-citation></ref>
      <ref id="bib1.bibx146"><label>Shepherd et al.(2012)</label><mixed-citation>Shepherd, A. P., Ivins, E. R., Geruo, A., Barletta, V. R., Bentley, M. J.,
Bettadpur, S., Briggs, K. H., Bromwich, D. H., Forsberg, R., Galin, N.,
Horwath, M., Jacobs, S. S., Joughin, I., King, M. A., Lenaerts, J. T. M., Li,
J., Ligtenberg, S. R. M., Luckman, A., Luthcke, S. B., McMillan, M., Meister,
R., Milne, G. A., Mouginot, J., Muir, A., Nicolas, J. P., Paden, J., Payne,
A. J., Pritchard, H. D., Rignot, E. J., Rott, H., Sandberg Sørensen, L.,
Scambos, T. A., Scheuchl, B., Schrama, E. J. O., Smith, B., Sundal, A. V.,
van Angelen, J. H., van de Berg, W. J., van den Broeke, M. R., Vaughan,
D. G., Velicogna, I., Wahr, J., Whitehouse, P. L., Wingham, D. J., Yi, D.,
Young, D. A., and Zwally, H. J.: A Reconciled Estimate of Ice-Sheet Mass
Balance, Science, 338, 1183–1189, <ext-link xlink:href="https://doi.org/10.1126/science.1228102" ext-link-type="DOI">10.1126/science.1228102</ext-link>, 2012.</mixed-citation></ref>
      <ref id="bib1.bibx147"><label>Silvano et al.(2017)</label><mixed-citation>Silvano, A., Rintoul, S. R., Pena-Molino, B., and Williams, G. D.:
Distribution of water masses and meltwater on the continental shelf near the
Totten and Moscow University ice shelves, J. Geophys. Res.-Oceans, 122,
2050–2068, <ext-link xlink:href="https://doi.org/10.1002/2016JC012115" ext-link-type="DOI">10.1002/2016JC012115</ext-link>, 2017.</mixed-citation></ref>
      <ref id="bib1.bibx148"><label>Smagorinsky(1963)</label><mixed-citation>Smagorinsky, J.: General Circulation Experiments With the Primitive
Equations I. The Basic Experiment, Mon. Weather Rev., 91, 99–164,
<ext-link xlink:href="https://doi.org/10.1175/1520-0493(1963)091&lt;0099:GCEWTP&gt;2.3.CO;2" ext-link-type="DOI">10.1175/1520-0493(1963)091&lt;0099:GCEWTP&gt;2.3.CO;2</ext-link>, 1963.</mixed-citation></ref>
      <ref id="bib1.bibx149"><label>Smagorinsky(1993)</label><mixed-citation>
Smagorinsky, J.: Some historical remarks on the use of non-linear
viscosities, in: Large Eddy Simulation of Complex Engineering and
Geophysical Flows, edited by: Galperin, B. and Orszag, S., Cambridge
University Press, Cambridge, 3–36, 1993.</mixed-citation></ref>
      <ref id="bib1.bibx150"><label>Smedsrud and Jenkins(2004)</label><mixed-citation>Smedsrud, L. H. and Jenkins, A.: Frazil ice formation in an ice shelf water
plume, J. Geophys. Res., 109, C03025, <ext-link xlink:href="https://doi.org/10.1029/2003JC001851" ext-link-type="DOI">10.1029/2003JC001851</ext-link>, 2004.</mixed-citation></ref>
      <ref id="bib1.bibx151"><label>Spence et al.(2014)</label><mixed-citation>Spence, P., Griffies, S. M., England, M. H., Hogg, A. M., Saenko, O. A., and
Jourdain, N. C.: Rapid subsurface warming and circulation changes of
Antarctic coastal waters by poleward shifting winds, Geophys. Res. Lett.,
41, 4601–4610, <ext-link xlink:href="https://doi.org/10.1002/2014GL060613" ext-link-type="DOI">10.1002/2014GL060613</ext-link>, 2014.</mixed-citation></ref>
      <ref id="bib1.bibx152"><label>Spence et al.(2017)</label><mixed-citation>Spence, P., Holmes, R. M., Hogg, A. M., Griffies, S. M., Stewart, K. D., and
England, M. H.: Localized rapid warming of West Antarctic subsurface waters
by remote winds, Nat. Clim. Change, 7, 595–603, <ext-link xlink:href="https://doi.org/10.1038/nclimate3335" ext-link-type="DOI">10.1038/nclimate3335</ext-link>,
2017.</mixed-citation></ref>
      <ref id="bib1.bibx153"><label>St-Laurent et al.(2013)</label><mixed-citation>St-Laurent, P., Klinck, J. M., and Dinniman, M. S.: On the Role of Coastal
Troughs in the Circulation of Warm Circumpolar Deep Water on Antarctic
Shelves, J. Phys. Oceanogr., 43, 51–64, <ext-link xlink:href="https://doi.org/10.1175/JPO-D-11-0237.1" ext-link-type="DOI">10.1175/JPO-D-11-0237.1</ext-link>,
2013.</mixed-citation></ref>
      <ref id="bib1.bibx154"><label>Stewart and Thompson(2015)</label><mixed-citation>Stewart, A. L. and Thompson, A. F.: Eddy-mediated transport of warm
Circumpolar Deep Water across the Antarctic Shelf Break, Geophys. Res.
Lett., 42, 432–440, <ext-link xlink:href="https://doi.org/10.1002/2014GL062281" ext-link-type="DOI">10.1002/2014GL062281</ext-link>, 2015.</mixed-citation></ref>
      <ref id="bib1.bibx155"><label>Talley(2007)</label><mixed-citation>
Talley, L. D.: Hydrographic Atlas of the World Ocean Circulation Experiment
(WOCE), Vol. 2, in: Pacific Ocean, edited by: Sparrow, M., Chapman, P., and
Gould, J., Tech. rep., International WOCE Project Office, Southampton, UK,
2007.</mixed-citation></ref>
      <ref id="bib1.bibx156"><label>Tamura and Ohshima(2008)</label><mixed-citation>Tamura, T. and Ohshima, K. I.: Mapping of sea ice production for Antarctic
coastal polynyas, J. Geophys. Res.-Oceans, 116, 1–5,
<ext-link xlink:href="https://doi.org/10.1029/2010JC006586" ext-link-type="DOI">10.1029/2010JC006586</ext-link>, 2008.</mixed-citation></ref>
      <?pagebreak page1292?><ref id="bib1.bibx157"><label>Tamura et al.(2016)</label><mixed-citation>Tamura, T., Ohshima, K. I., Fraser, A. D., and Williams, G. D.: Sea ice
production variability in Antarctic coastal polynyas, J. Geophys.
Res.-Oceans, 121, 2967–2979, <ext-link xlink:href="https://doi.org/10.1002/2015JC011537" ext-link-type="DOI">10.1002/2015JC011537</ext-link>, 2016.</mixed-citation></ref>
      <ref id="bib1.bibx158"><label>Thomas et al.(2001)</label><mixed-citation>Thomas, M., Sündermann, J., and Maier-Reimer, E.: Consideration of
ocean tides in an OGCM and impacts on subseasonal to decadal polar motion
excitation, Geophys. Res. Lett., 28, 2457–2460, <ext-link xlink:href="https://doi.org/10.1029/2000GL012234" ext-link-type="DOI">10.1029/2000GL012234</ext-link>,
2001.</mixed-citation></ref>
      <ref id="bib1.bibx159"><label>Timmermann and Beckmann(2004)</label><mixed-citation>Timmermann, R. and Beckmann, A.: Parameterization of vertical mixing in the
Weddell Sea, Ocean Model., 6, 83–100, <ext-link xlink:href="https://doi.org/10.1016/S1463-5003(02)00061-6" ext-link-type="DOI">10.1016/S1463-5003(02)00061-6</ext-link>,
2004.</mixed-citation></ref>
      <ref id="bib1.bibx160"><label>Timmermann and Goeller(2017)</label><mixed-citation>
Timmermann, R. and Goeller, S.: Response to Filchner–Ronne Ice Shelf cavity
warming in a coupled ocean–ice sheet model – Part 1: The ocean perspective,
Ocean Sci., 13, 765–776, https://doi.org/10.5194/os-13-765-2017, 2017.</mixed-citation></ref>
      <ref id="bib1.bibx161"><label>Timmermann and Hellmer(2013)</label><mixed-citation>Timmermann, R. and Hellmer, H. H.: Southern Ocean warming and increased ice
shelf basal melting in the twenty-first and twenty-second centuries based on
coupled ice-ocean finite-element modelling, Ocean Dynam., 63, 1011–1026,
<ext-link xlink:href="https://doi.org/10.1007/s10236-013-0642-0" ext-link-type="DOI">10.1007/s10236-013-0642-0</ext-link>, 2013.</mixed-citation></ref>
      <ref id="bib1.bibx162"><label>Timmermann et al.(2009)</label><mixed-citation>Timmermann, R., Danilov, S., Schröter, J., Böning, C., Sidorenko,
D., and Rollenhagen, K.: Ocean circulation and sea ice distribution in a
finite element global sea ice-ocean model, Ocean Model., 27, 114–129,
<ext-link xlink:href="https://doi.org/10.1016/j.ocemod.2008.10.009" ext-link-type="DOI">10.1016/j.ocemod.2008.10.009</ext-link>, 2009.</mixed-citation></ref>
      <ref id="bib1.bibx163"><label>Timmermann et al.(2010)</label><mixed-citation>
Timmermann, R., Le Brocq, A., Deen, T., Domack, E., Dutrieux, P.,
Galton-Fenzi, B., Hellmer, H., Humbert, A., Jansen, D., Jenkins, A.,
Lambrecht, A., Makinson, K., Niederjasper, F., Nitsche, F., Nøst, O. A.,
Smedsrud, L. H., and Smith, W. H. F.: A consistent data set of Antarctic ice
sheet topography, cavity geometry, and global bathymetry, Earth Syst. Sci.
Data, 2, 261–273, https://doi.org/10.5194/essd-2-261-2010, 2010.</mixed-citation></ref>
      <ref id="bib1.bibx164"><label>Timmermann et al.(2012)</label><mixed-citation>Timmermann, R., Wang, Q., and Hellmer, H. H.: Ice-shelf basal melting in a
global finite-element sea-ice/ice-shelf/ocean model, Ann. Glaciol., 53,
303–314, <ext-link xlink:href="https://doi.org/10.3189/2012AoG60A156" ext-link-type="DOI">10.3189/2012AoG60A156</ext-link>, 2012.</mixed-citation></ref>
      <ref id="bib1.bibx165"><label>Trenberth and Fasullo(2010)</label><mixed-citation>Trenberth, K. E. and Fasullo, J. T.: Simulation of present-day and
twenty-first-century energy budgets of the southern oceans, J. Climate, 23,
440–454, <ext-link xlink:href="https://doi.org/10.1175/2009JCLI3152.1" ext-link-type="DOI">10.1175/2009JCLI3152.1</ext-link>, 2010.</mixed-citation></ref>
      <ref id="bib1.bibx166"><label>Tsamados et al.(2013)</label><mixed-citation>Tsamados, M., Feltham, D. L., and Wilchinsky, A. V.: Impact of a new
anisotropic rheology on simulations of Arctic Sea ice, J. Geophys.
Res.-Oceans, 118, 91–107, <ext-link xlink:href="https://doi.org/10.1029/2012JC007990" ext-link-type="DOI">10.1029/2012JC007990</ext-link>, 2013.</mixed-citation></ref>
      <ref id="bib1.bibx167"><label>Turner et al.(2013a)</label><mixed-citation>Turner, A. K., Hunke, E. C., and Bitz, C. M.: Two modes of sea-ice gravity
drainage: A parameterization for large-scale modeling, J. Geophys.
Res.-Oceans, 118, 2279–2294, <ext-link xlink:href="https://doi.org/10.1002/jgrc.20171" ext-link-type="DOI">10.1002/jgrc.20171</ext-link>, 2013a.</mixed-citation></ref>
      <ref id="bib1.bibx168"><label>Turner et al.(2013b)</label><mixed-citation>Turner, J., Bracegirdle, T. J., Phillips, T., Marshall, G. J., and Hosking,
J. S.: An initial assessment of Antarctic sea ice extent in the CMIP5
models, J. Climate, 26, 1473–1484, <ext-link xlink:href="https://doi.org/10.1175/JCLI-D-12-00068.1" ext-link-type="DOI">10.1175/JCLI-D-12-00068.1</ext-link>,
2013b.
</mixed-citation></ref><?xmltex \hack{\newpage}?>
      <ref id="bib1.bibx169"><label>Venables and Meredith(2014)</label><mixed-citation>Venables, H. J. and Meredith, M. P.: Feedbacks between ice cover, ocean
stratification, and heat content in Ryder Bay, western Antarctic Peninsula,
J. Geophys. Res.-Oceans, 119, 5323–5336, <ext-link xlink:href="https://doi.org/10.1002/2013JC009669" ext-link-type="DOI">10.1002/2013JC009669</ext-link>, 2014.</mixed-citation></ref>
      <ref id="bib1.bibx170"><?xmltex \def\ref@label{{W{\aa}hlin et~al.(2010)}}?><label>Wåhlin et al.(2010)</label><mixed-citation>Wåhlin, A. K., Yuan, X., Björk, G., and Nohr, C.: Inflow of Warm
Circumpolar Deep Water in the Central Amundsen Shelf, J. Phys. Oceanogr.,
40, 1427–1434, <ext-link xlink:href="https://doi.org/10.1175/2010JPO4431.1" ext-link-type="DOI">10.1175/2010JPO4431.1</ext-link>, 2010.</mixed-citation></ref>
      <ref id="bib1.bibx171"><label>Wang et al.(2008)</label><mixed-citation>Wang, Q., Danilov, S., and Schröter, J.: Finite element ocean
circulation model based on triangular prismatic elements, with application in
studying the effect of topography representation, J. Geophys. Res., 113,
C05015, <ext-link xlink:href="https://doi.org/10.1029/2007JC004482" ext-link-type="DOI">10.1029/2007JC004482</ext-link>, 2008.</mixed-citation></ref>
      <ref id="bib1.bibx172"><label>Wang et al.(2013)</label><mixed-citation>Wang, Q., Danilov, S., Hellmer, H. H., Sidorenko, D., Schröter, J., and
Jung, T.: Enhanced cross-shelf exchange by tides in the western Ross Sea,
Geophys. Res. Lett., 40, 5735–5739, <ext-link xlink:href="https://doi.org/10.1002/2013GL058207" ext-link-type="DOI">10.1002/2013GL058207</ext-link>, 2013.</mixed-citation></ref>
      <ref id="bib1.bibx173"><label>Wang et al.(2014)</label><mixed-citation>
Wang, Q., Danilov, S., Sidorenko, D., Timmermann, R., Wekerle, C., Wang, X.,
Jung, T., and Schröter, J.: The Finite Element Sea Ice-Ocean Model
(FESOM) v.1.4: formulation of an ocean general circulation model, Geosci.
Model Dev., 7, 663–693, https://doi.org/10.5194/gmd-7-663-2014, 2014.</mixed-citation></ref>
      <ref id="bib1.bibx174"><label>Wen et al.(2010)</label><mixed-citation>Wen, J., Wang, Y., Wang, W., Jezek, K. C., Liu, H., and Allison, I.: Basal
melting and freezing under the Amery Ice Shelf, East Antarctica, J.
Glaciol., 56, 81–90, <ext-link xlink:href="https://doi.org/10.3189/002214310791190820" ext-link-type="DOI">10.3189/002214310791190820</ext-link>, 2010.</mixed-citation></ref>
      <ref id="bib1.bibx175"><label>Williams et al.(2013)</label><mixed-citation>Williams, K. D., Bodas-Salcedo, A., Déqué, M., Fermepin, S.,
Medeiros, B., Watanabe, M., Jakob, C., Klein, S. A., Senior, C. A., and
Williamson, D. L.: The transpose-AMIP II experiment and its application to
the understanding of southern ocean cloud biases in climate models, J.
Climate, 26, 3258–3274, <ext-link xlink:href="https://doi.org/10.1175/JCLI-D-12-00429.1" ext-link-type="DOI">10.1175/JCLI-D-12-00429.1</ext-link>, 2013.</mixed-citation></ref>
      <ref id="bib1.bibx176"><label>Worby et al.(2008)</label><mixed-citation>Worby, A. P., Geiger, C. A., Paget, M. J., van Woert, M. L., Ackley, S. F.,
and DeLiberty, T. L.: Thickness distribution of Antarctic sea ice, J.
Geophys. Res.-Oceans, 113, 1–14, <ext-link xlink:href="https://doi.org/10.1029/2007JC004254" ext-link-type="DOI">10.1029/2007JC004254</ext-link>, 2008.</mixed-citation></ref>
      <ref id="bib1.bibx177"><label>Wunsch et al.(2009)</label><mixed-citation>Wunsch, C., Heimbach, P., Ponte, R., and Fukumori, I.: The Global General
Circulation of the Ocean Estimated by the ECCO-Consortium, Oceanography, 22,
88–103, <ext-link xlink:href="https://doi.org/10.5670/oceanog.2009.41" ext-link-type="DOI">10.5670/oceanog.2009.41</ext-link>, 2009.</mixed-citation></ref>
      <ref id="bib1.bibx178"><label>Zwally and Giovinetto(2011)</label><mixed-citation>Zwally, H. J. and Giovinetto, M. B.: Overview and Assessment of Antarctic
Ice-Sheet Mass Balance Estimates: 1992-2009, Surv. Geophys., 32, 351–376,
<ext-link xlink:href="https://doi.org/10.1007/s10712-011-9123-5" ext-link-type="DOI">10.1007/s10712-011-9123-5</ext-link>, 2011.</mixed-citation></ref>
      <ref id="bib1.bibx179"><label>Zweng et al.(2013)</label><mixed-citation>
Zweng, M. M., Reagan, J. R., Antonov, J. I., Mishonov, A. V., Boyer, T. P.,
Garcia, H. E., Baranova, O. K., Johnson, D. R., Seidov, D., and Bidlle,
M. M.: World Ocean Atlas 2013, vol. 2, Salinity, NOAA Atlas NESDIS, 2,
39 pp., 2013.</mixed-citation></ref>

  </ref-list></back>
    <!--<article-title-html>Intercomparison of Antarctic ice-shelf, ocean, and sea-ice interactions simulated by MetROMS-iceshelf and FESOM 1.4</article-title-html>
<abstract-html><p>An increasing number of Southern Ocean models now include Antarctic ice-shelf
cavities, and simulate thermodynamics at the ice-shelf/ocean interface. This
adds another level of complexity to Southern Ocean simulations, as ice
shelves interact directly with the ocean and indirectly with sea ice. Here,
we present the first model intercomparison and evaluation of present-day
ocean/sea-ice/ice-shelf interactions, as simulated by two models: a
circumpolar Antarctic configuration of MetROMS (ROMS: Regional Ocean
Modelling System coupled to CICE: Community Ice CodE) and the global model
FESOM (Finite Element Sea-ice Ocean Model), where the latter is run at two
different levels of horizontal resolution. From a circumpolar Antarctic
perspective, we compare and evaluate simulated ice-shelf basal melting and
sub-ice-shelf circulation, as well as sea-ice properties and Southern Ocean
water mass characteristics as they influence the sub-ice-shelf processes.
Despite their differing numerical methods, the two models produce broadly
similar results and share similar biases in many cases. Both models reproduce
many key features of observations but struggle to reproduce others, such as
the high melt rates observed in the small warm-cavity ice shelves of the
Amundsen and Bellingshausen seas. Several
differences in model design show a particular influence on the simulations.
For example, FESOM's greater topographic smoothing can alter the geometry of
some ice-shelf cavities enough to affect their melt rates; this improves at
higher resolution, since less smoothing is required. In the interior Southern
Ocean, the vertical coordinate system affects the degree of water mass
erosion due to spurious diapycnal mixing, with MetROMS' terrain-following
coordinate leading to more erosion than FESOM's <i>z</i> coordinate. Finally,
increased horizontal resolution in FESOM leads to higher basal melt rates for
small ice shelves, through a combination of stronger circulation and
small-scale intrusions of warm water from offshore.</p></abstract-html>
<ref-html id="bib1.bib1"><label>Arakawa and Lamb(1977)</label><mixed-citation>
Arakawa, A. and Lamb, V. R.: Computational design of the basic dynamical
processes of the UCLA general circulation model, Methods in Computational
Physics: Advances in Research and Applications, 17, 173–265, 1977.
</mixed-citation></ref-html>
<ref-html id="bib1.bib2"><label>Asay-Davis et al.(2016)</label><mixed-citation>
Asay-Davis, X. S., Cornford, S. L., Durand, G., Galton-Fenzi, B. K.,
Gladstone, R. M., Gudmundsson, G. H., Hattermann, T., Holland, D. M.,
Holland, D., Holland, P. R., Martin, D. F., Mathiot, P., Pattyn, F., and
Seroussi, H.: Experimental design for three interrelated marine ice sheet and
ocean model intercomparison projects: MISMIP v. 3 (MISMIP +), ISOMIP v. 2
(ISOMIP +) and MISOMIP v. 1 (MISOMIP1), Geosci. Model Dev., 9, 2471–2497,
https://doi.org/10.5194/gmd-9-2471-2016, 2016.
</mixed-citation></ref-html>
<ref-html id="bib1.bib3"><label>AVISO(2011)</label><mixed-citation>
AVISO: AVISO Level 4 Absolute Dynamic Topography for Climate Model
Comparison, Version 1, <a href="https://doi.org/10.5067/DYNTO-1D1M1" target="_blank">https://doi.org/10.5067/DYNTO-1D1M1</a>, 2011.
</mixed-citation></ref-html>
<ref-html id="bib1.bib4"><label>Bodas-Salcedo et al.(2016)</label><mixed-citation>
Bodas-Salcedo, A., Andrews, T., Karmalkar, A. V., and Ringer, M. A.: Cloud
liquid water path and radiative feedbacks over the Southern Ocean, Geophys.
Res. Lett., 43, 10938–10946, <a href="https://doi.org/10.1002/2016GL070770" target="_blank">https://doi.org/10.1002/2016GL070770</a>, 2016.
</mixed-citation></ref-html>
<ref-html id="bib1.bib5"><label>Bouillon et al.(2013)</label><mixed-citation>
Bouillon, S., Fichefet, T., Legat, V., and Madec, G.: The
elastic-viscous-plastic method revisited, Ocean Model., 71, 2–12,
<a href="https://doi.org/10.1016/j.ocemod.2013.05.013" target="_blank">https://doi.org/10.1016/j.ocemod.2013.05.013</a>, 2013.
</mixed-citation></ref-html>
<ref-html id="bib1.bib6"><label>Briegleb and Light(2007)</label><mixed-citation>
Briegleb, B. P. and Light, B.: A Delta-Eddington multiple scattering
parameterization for solar radiation in the sea ice component of the
Community Climate System Model: NCAR Tech. Note NCAR/TN-472+STR, Tech. rep.,
National Center for Atmospheric Research, 2007.
</mixed-citation></ref-html>
<ref-html id="bib1.bib7"><label>Chapman(1985)</label><mixed-citation>
Chapman, D. C.: Numerical Treatment of Cross-Shelf Open Boundaries in a
Barotropic Coastal Ocean Model, J. Phys. Oceanogr., 15, 1060–1075,
<a href="https://doi.org/10.1175/1520-0485(1985)015&lt;1060:NTOCSO&gt;2.0.CO;2" target="_blank">https://doi.org/10.1175/1520-0485(1985)015&lt;1060:NTOCSO&gt;2.0.CO;2</a>, 1985.
</mixed-citation></ref-html>
<ref-html id="bib1.bib8"><label>Cook et al.(2013)</label><mixed-citation>
Cook, C. P., van de Flierdt, T., Williams, T., Hemming, S. R., Iwai, M.,
Kobayashi, M., Jimenez-Espejo, F. J., Escutia, C., González, J. J.,
Khim, B.-K., McKay, R. M., Passchier, S., Bohaty, S. M., Riesselman, C. R.,
Tauxe, L., Sugisaki, S., Galindo, A. L., Patterson, M. O., Sangiorgi, F.,
Pierce, E. L., Brinkhuis, H., Klaus, A., Fehr, A., Bendle, J. A. P., Bijl,
P. K., Carr, S. A., Dunbar, R. B., Flores, J. A., Hayden, T. G., Katsuki, K.,
Kong, G. S., Nakai, M., Olney, M. P., Pekar, S. F., Pross, J., Röhl,
U., Sakai, T., Shrivastava, P. K., Stickley, C. E., Tuo, S., Welsh, K., and
Yamane, M.: Dynamic behaviour of the East Antarctic ice sheet during
Pliocene warmth, Nat. Geosci., 6, 765–769, <a href="https://doi.org/10.1038/ngeo1889" target="_blank">https://doi.org/10.1038/ngeo1889</a>, 2013.
</mixed-citation></ref-html>
<ref-html id="bib1.bib9"><label>Cougnon et al.(2013)</label><mixed-citation>
Cougnon, E. A., Galton-Fenzi, B. K., Meijers, A. J. S., and Legrésy,
B.: Modeling interannual dense shelf water export in the region of the Mertz
Glacier Tongue (1992–2007), J. Geophys. Res.-Oceans, 118, 5858–5872,
<a href="https://doi.org/10.1002/2013JC008790" target="_blank">https://doi.org/10.1002/2013JC008790</a>, 2013.
</mixed-citation></ref-html>
<ref-html id="bib1.bib10"><label>Cunningham et al.(2003)</label><mixed-citation>
Cunningham, S. A., Alderson, S. G., King, B. A., and Brandon, M. A.:
Transport and variability of the Antarctic Circumpolar Current in Drake
Passage, J. Geophys. Res., 108, 8084, <a href="https://doi.org/10.1029/2001JC001147" target="_blank">https://doi.org/10.1029/2001JC001147</a>, 2003.
</mixed-citation></ref-html>
<ref-html id="bib1.bib11"><label>Danilov et al.(2015)</label><mixed-citation>
Danilov, S., Wang, Q., Timmermann, R., Iakovlev, N., Sidorenko, D., Kimmritz,
M., Jung, T., and Schröter, J.: Finite-Element Sea Ice Model (FESIM),
version 2, Geosci. Model Dev., 8, 1747-1761,
https://doi.org/10.5194/gmd-8-1747-2015, 2015.
</mixed-citation></ref-html>
<ref-html id="bib1.bib12"><label>Debernard et al.(2017)</label><mixed-citation>
Debernard, J. B., Kristensen, N. M., Maartensson, S., Wang, K., and Waagbo,
G. A.: metno/metroms: Intermediate release, <a href="https://doi.org/10.5281/zenodo.290667" target="_blank">https://doi.org/10.5281/zenodo.290667</a>,
2017.
</mixed-citation></ref-html>
<ref-html id="bib1.bib13"><label>Deconto and Pollard(2016)</label><mixed-citation>
Deconto, R. M. and Pollard, D.: Contribution of Antarctica to past and
future sea-level rise, Nature, 531, 591–597, <a href="https://doi.org/10.1038/nature17145" target="_blank">https://doi.org/10.1038/nature17145</a>,
2016.
</mixed-citation></ref-html>
<ref-html id="bib1.bib14"><label>Dee et al.(2011)</label><mixed-citation>
Dee, D. P., Uppala, S. M., Simmons, A. J., Berrisford, P., Poli, P.,
Kobayashi, S., Andrae, U., Balmaseda, M. A., Balsamo, G., Bauer, P.,
Bechtold, P., Beljaars, A. C. M., van de Berg, I., Biblot, J., Bormann, N.,
Delsol, C., Dragani, R., Fuentes, M., Greer, A. J., Haimberger, L., Healy, S.
B., Hersbach, H., Holm, E. V., Isaksen, L., Kallberg, P., Kohler, M.,
Matricardi, M., McNally, A. P., Mong-Sanz, B. M., Morcette, J.-J., Park,
B.-K., Peubey, C., de Rosnay, P., Tavolato, C., Thepaut, J. N., and Vitart,
F.: The ERA-Interim reanalysis: Configuration and performance of the data
assimilation system, Q. J. Roy. Meteorol. Soc., 137, 553–597,
doi:10.1002/qj.828, 2011.
</mixed-citation></ref-html>
<ref-html id="bib1.bib15"><label>Depoorter et al.(2013)</label><mixed-citation>
Depoorter, M. A., Bamber, J. L., Griggs, J. A., Lenaerts, J. T. M.,
Ligtenberg, S. R. M., van den Broeke, M. R., and Moholdt, G.: Calving fluxes
and basal melt rates of Antarctic ice shelves, Nature, 502, 89–92,
<a href="https://doi.org/10.1038/nature12567" target="_blank">https://doi.org/10.1038/nature12567</a>, 2013.
</mixed-citation></ref-html>
<ref-html id="bib1.bib16"><label>Dinniman et al.(2011)</label><mixed-citation>
Dinniman, M. S., Klinck, J. M., and Smith, W. O.: A model study of
Circumpolar Deep Water on the West Antarctic Peninsula and Ross Sea
continental shelves, Deep-Sea Res. Pt. II, 58, 1508–1523,
<a href="https://doi.org/10.1016/j.dsr2.2010.11.013" target="_blank">https://doi.org/10.1016/j.dsr2.2010.11.013</a>, 2011.
</mixed-citation></ref-html>
<ref-html id="bib1.bib17"><label>Dinniman et al.(2016)</label><mixed-citation>
Dinniman, M. S., Asay-Davis, X. S., Galton-Fenzi, B. K., Holland, P. R., and
Jenkins, A.: Modeling Ice Shelf/Ocean Interaction in Antarctica: a review,
Oceanography, 29, 144–153, <a href="https://doi.org/10.5670/oceanog.2016.106" target="_blank">https://doi.org/10.5670/oceanog.2016.106</a>, 2016.
</mixed-citation></ref-html>
<ref-html id="bib1.bib18"><label>Donohue et al.(2016)</label><mixed-citation>
Donohue, K. A., Tracey, K. L., Watts, D. R., Chidichimo, M. P., and
Chereskin, T. K.: Mean Antarctic Circumpolar Current transport measured in
Drake Passage, Geophys. Res. Lett., 43, 11,760–11,767,
<a href="https://doi.org/10.1002/2016GL070319" target="_blank">https://doi.org/10.1002/2016GL070319</a>, 2016.
</mixed-citation></ref-html>
<ref-html id="bib1.bib19"><label>Dupont(2005)</label><mixed-citation>
Dupont, T. K.: Assessment of the importance of ice-shelf buttressing to
ice-sheet flow, Geophys. Res. Letters, 32, L04503,
<a href="https://doi.org/10.1029/2004GL022024" target="_blank">https://doi.org/10.1029/2004GL022024</a>, 2005.
</mixed-citation></ref-html>
<ref-html id="bib1.bib20"><label>Dutrieux et al.(2014)</label><mixed-citation>
Dutrieux, P., De Rydt, J., Jenkins, A., Holland, P. R., Ha, H. K., Lee,
S. H., Steig, E. J., Ding, Q., Abrahamsen, E. P., and Schröder, M.:
Strong sensitivity of Pine Island ice-shelf melting to climatic
variability, Science, 343, 174–178, <a href="https://doi.org/10.1126/science.1244341" target="_blank">https://doi.org/10.1126/science.1244341</a>, 2014.
</mixed-citation></ref-html>
<ref-html id="bib1.bib21"><label>Dutton et al.(2015)</label><mixed-citation>
Dutton, A., Carlson, A. E., Long, A. J., Milne, G. A., Clark, P. U., Deconto,
R. M., Horton, B. P., Rahmstorf, S., and Raymo, M. E.: Sea-level rise due to
polar ice-sheet mass loss during past warm periods, Science, 349, aaa4019,
<a href="https://doi.org/10.1126/science.aaa4019" target="_blank">https://doi.org/10.1126/science.aaa4019</a>, 2015.
</mixed-citation></ref-html>
<ref-html id="bib1.bib22"><label>England(1993)</label><mixed-citation>
England, M. H.: Representing the Global-Scale Water Masses in Ocean General
Circulation Models, J. Phys. Oceanogr., 23, 1523–1552,
<a href="https://doi.org/10.1175/1520-0485(1993)023&lt;1523:RTGSWM&gt;2.0.CO;2" target="_blank">https://doi.org/10.1175/1520-0485(1993)023&lt;1523:RTGSWM&gt;2.0.CO;2</a>, 1993.
</mixed-citation></ref-html>
<ref-html id="bib1.bib23"><label>England et al.(1993)</label><mixed-citation>
England, M. H., Godfrey, J. S., Hirst, A. C., and Tomczak, M.: The Mechanism
for Antarctic Intermediate Water Renewal in a World Ocean Model, J. Phys.
Oceanogr., 23, 1553–1560,
<a href="https://doi.org/10.1175/1520-0485(1993)023&lt;1553:TMFAIW&gt;2.0.CO;2" target="_blank">https://doi.org/10.1175/1520-0485(1993)023&lt;1553:TMFAIW&gt;2.0.CO;2</a>, 1993.
</mixed-citation></ref-html>
<ref-html id="bib1.bib24"><label>Engwirda et al.(2017)</label><mixed-citation>
Engwirda, D., Kelley, M., and Marshall, J.: High-order accurate
finite-volume formulations for the pressure gradient force in layered ocean
models, Ocean Model., 116, 1–15, <a href="https://doi.org/10.1016/j.ocemod.2017.05.003" target="_blank">https://doi.org/10.1016/j.ocemod.2017.05.003</a>, 2017.
</mixed-citation></ref-html>
<ref-html id="bib1.bib25"><label>Fairall et al.(1996)</label><mixed-citation>
Fairall, C. W., Bradley, E. F., Rogers, D. P., Edson, J. B., and Young,
G. S.: Bulk parameterization of air-sea fluxes for Tropical Ocean-Global
Atmosphere Coupled-Ocean Atmosphere Response Experiment, J. Geophys.
Res.-Oceans, 101, 3747–3764, <a href="https://doi.org/10.1029/95JC03205" target="_blank">https://doi.org/10.1029/95JC03205</a>, 1996.
</mixed-citation></ref-html>
<ref-html id="bib1.bib26"><label>Fetterrer et al.(2016)</label><mixed-citation>
Fetterrer, F., Knowles, K., Meier, W., and Savoie, M.: Sea Ice Index,
Version 2, National Snow &amp; Ice Data Center, <a href="https://doi.org/10.7265/N5736NV7" target="_blank">https://doi.org/10.7265/N5736NV7</a>, 2016.
</mixed-citation></ref-html>
<ref-html id="bib1.bib27"><label>Flather(1976)</label><mixed-citation>
Flather, R.: A tidal model of the northwest European continental shelf,
Memoires de la Societe Royale de Sciences de Liege, 6, 141–164, 1976.
</mixed-citation></ref-html>
<ref-html id="bib1.bib28"><label>Foldvik et al.(1990)</label><mixed-citation>
Foldvik, A., Middleton, J. H., and Foster, T. D.: The tides of the southern
Weddell Sea, Deep-Sea Res. Pt. I, 37, 1345–1362,
<a href="https://doi.org/10.1016/0198-0149(90)90047-Y" target="_blank">https://doi.org/10.1016/0198-0149(90)90047-Y</a>, 1990.
</mixed-citation></ref-html>
<ref-html id="bib1.bib29"><label>Foldvik et al.(2004)</label><mixed-citation>
Foldvik, A., Gammelsrød, T., Østerhus, S., Fahrbach, E., Rohart, G.,
Schröder, M., Nicholls, K. W., Padman, L., and Woodgate, R. A.: Ice
shelf water overflow and bottom water formation in the southern Weddell Sea,
J. Geophys. Res., 109, C02015, <a href="https://doi.org/10.1029/2003JC002008" target="_blank">https://doi.org/10.1029/2003JC002008</a>, 2004.
</mixed-citation></ref-html>
<ref-html id="bib1.bib30"><label>Fraser et al.(2012)</label><mixed-citation>
Fraser, A. D., Massom, R. A., Michael, K. J., Galton-Fenzi, B. K., and
Lieser, J. L.: East Antarctic Landfast Sea Ice Distribution and Variability,
2000-08, J. Climate, 25, 1137–1156, <a href="https://doi.org/10.1175/JCLI-D-10-05032.1" target="_blank">https://doi.org/10.1175/JCLI-D-10-05032.1</a>, 2012.
</mixed-citation></ref-html>
<ref-html id="bib1.bib31"><label>Fretwell et al.(2013)</label><mixed-citation>
Fretwell, P., Pritchard, H. D., Vaughan, D. G., Bamber, J. L., Barrand, N.
E., Bell, R., Bianchi, C., Bingham, R. G., Blankenship, D. D., Casassa, G.,
Catania, G., Callens, D., Conway, H., Cook, A. J., Corr, H. F. J., Damaske,
D., Damm, V., Ferraccioli, F., Forsberg, R., Fujita, S., Gim, Y., Gogineni,
P., Griggs, J. A., Hindmarsh, R. C. A., Holmlund, P., Holt, J. W., Jacobel,
R. W., Jenkins, A., Jokat, W., Jordan, T., King, E. C., Kohler, J., Krabill,
W., Riger-Kusk, M., Langley, K. A., Leitchenkov, G., Leuschen, C., Luyendyk,
B. P., Matsuoka, K., Mouginot, J., Nitsche, F. O., Nogi, Y., Nost, O. A.,
Popov, S. V., Rignot, E., Rippin, D. M., Rivera, A., Roberts, J., Ross, N.,
Siegert, M. J., Smith, A. M., Steinhage, D., Studinger, M., Sun, B., Tinto,
B. K., Welch, B. C., Wilson, D., Young, D. A., Xiangbin, C., and Zirizzotti,
A.: Bedmap2: improved ice bed, surface and thickness datasets for Antarctica,
The Cryosphere, 7, 375–393, https://doi.org/10.5194/tc-7-375-2013, 2013.
</mixed-citation></ref-html>
<ref-html id="bib1.bib32"><label>Gade(1979)</label><mixed-citation>
Gade, H. G.: Melting of Ice in Sea Water: A Primitive Model with Application
to the Antarctic Ice Shelf and Icebergs, J. Phys. Oceanogr., 9, 189–198,
<a href="https://doi.org/10.1175/1520-0485(1979)009&lt;0189:MOIISW&gt;2.0.CO;2" target="_blank">https://doi.org/10.1175/1520-0485(1979)009&lt;0189:MOIISW&gt;2.0.CO;2</a>, 1979.
</mixed-citation></ref-html>
<ref-html id="bib1.bib33"><label>Galton-Fenzi(2009)</label><mixed-citation>
Galton-Fenzi, B. K.: Modelling Ice-Shelf/Ocean Interaction, PhD thesis,
University of Tasmania, 2009.
</mixed-citation></ref-html>
<ref-html id="bib1.bib34"><label>Galton-Fenzi et al.(2008)</label><mixed-citation>
Galton-Fenzi, B. K., Maraldi, C., Coleman, R., and Hunter, J. R.: The cavity
under the Amery Ice Shelf, East Antarctica, J. Glaciol., 54, 881–887,
<a href="https://doi.org/10.3189/002214308787779898" target="_blank">https://doi.org/10.3189/002214308787779898</a>, 2008.
</mixed-citation></ref-html>
<ref-html id="bib1.bib35"><label>Galton-Fenzi et al.(2012)</label><mixed-citation>
Galton-Fenzi, B. K., Hunter, J. R., Coleman, R., Marsland, S. J., and Warner,
R. C.: Modeling the basal melting and marine ice accretion of the Amery Ice
Shelf, J. Geophys. Res.-Oceans, 117, C09031, <a href="https://doi.org/10.1029/2012JC008214" target="_blank">https://doi.org/10.1029/2012JC008214</a>,
2012.
</mixed-citation></ref-html>
<ref-html id="bib1.bib36"><label>Gent and McWilliams(1990)</label><mixed-citation>
Gent, P. R. and McWilliams, J. C.: Isopycnal Mixing in Ocean Circulation
Models, J. Phys. Oceanogr., 20, 150–155,
<a href="https://doi.org/10.1175/1520-0485(1990)020&lt;0150:IMIOCM&gt;2.0.CO;2" target="_blank">https://doi.org/10.1175/1520-0485(1990)020&lt;0150:IMIOCM&gt;2.0.CO;2</a>, 1990.
</mixed-citation></ref-html>
<ref-html id="bib1.bib37"><label>Gent et al.(1995)</label><mixed-citation>
Gent, P. R., Willebrand, J., McDougall, T. J., and McWilliams, J. C.:
Parameterizing Eddy-Induced Tracer Transports in Ocean Circulation Models,
J. Phys. Oceanogr., 25, 463–474,
<a href="https://doi.org/10.1175/1520-0485(1995)025&lt;0463:PEITTI&gt;2.0.CO;2" target="_blank">https://doi.org/10.1175/1520-0485(1995)025&lt;0463:PEITTI&gt;2.0.CO;2</a>, 1995.
</mixed-citation></ref-html>
<ref-html id="bib1.bib38"><label>Gladstone et al.(2017)</label><mixed-citation>
Gladstone, R. M., Jong, L., Galton-Fenzi, B. K., Gwyther, D. E., and Moore,
J. C.: Simulating the interactions between marine ice sheets and their
surrounding oceans, in: Forum for Research into Ice Shelf Processes, Bergen,
Norway, 2017.
</mixed-citation></ref-html>
<ref-html id="bib1.bib39"><label>Golledge et al.(2015)</label><mixed-citation>
Golledge, N. R., Kowalewski, D. E., Naish, T. R., Levy, R. H., Fogwill,
C. J., and Gasson, E. G. W.: The multi-millennial Antarctic commitment to
future sea-level rise, Nature, 526, 421–425, <a href="https://doi.org/10.1038/nature15706" target="_blank">https://doi.org/10.1038/nature15706</a>,
2015.
</mixed-citation></ref-html>
<ref-html id="bib1.bib40"><label>Golledge et al.(2017)</label><mixed-citation>
Golledge, N. R., Levy, R. H., McKay, R. M., and Naish, T. R.: East Antarctic
ice sheet most vulnerable to Weddell Sea warming, Geophys. Res. Lett., 44,
1–9, <a href="https://doi.org/10.1002/2016GL072422" target="_blank">https://doi.org/10.1002/2016GL072422</a>, 2017.
</mixed-citation></ref-html>
<ref-html id="bib1.bib41"><label>Goosse and Fichefet(2001)</label><mixed-citation>
Goosse, H. and Fichefet, T.: Open-ocean convection and polynya formation in
a large-scale ice-ocean model, Tellus A, 53, 94–111,
<a href="https://doi.org/10.1034/j.1600-0870.2001.01061.x" target="_blank">https://doi.org/10.1034/j.1600-0870.2001.01061.x</a>, 2001.
</mixed-citation></ref-html>
<ref-html id="bib1.bib42"><label>Gordon et al.(2015)</label><mixed-citation>
Gordon, A. L., Huber, B. A., and Busecke, J.: Bottom water export from the
western Ross Sea, 2007 through 2010, Geophys. Res. Lett., 42, 5387–5394,
<a href="https://doi.org/10.1002/2015GL064457" target="_blank">https://doi.org/10.1002/2015GL064457</a>, 2015.
</mixed-citation></ref-html>
<ref-html id="bib1.bib43"><label>Graham et al.(2016)</label><mixed-citation>
Graham, J. A., Dinniman, M. S., and Klinck, J. M.: Impact of model
resolution for on-shelf heat transport along the West Antarctic Peninsula,
J. Geophys. Res.-Oceans, 121, 7880–7897, <a href="https://doi.org/10.1002/2016JC011875" target="_blank">https://doi.org/10.1002/2016JC011875</a>, 2016.
</mixed-citation></ref-html>
<ref-html id="bib1.bib44"><label>Greenbaum et al.(2015)</label><mixed-citation>
Greenbaum, J. S., Blankenship, D. D., Young, D. A., Richter, T. G., Roberts,
J. L., Aitken, A. R. A., Legrésy, B., Schroeder, D. M., Warner, R. C.,
van Ommen, T. D., and Siegert, M. J.: Ocean access to a cavity beneath
Totten Glacier in East Antarctica, Nat. Geosci., 8, 294–298,
<a href="https://doi.org/10.1038/NGEO2388" target="_blank">https://doi.org/10.1038/NGEO2388</a>, 2015.
</mixed-citation></ref-html>
<ref-html id="bib1.bib45"><label>Griffies et al.(2000)</label><mixed-citation>
Griffies, S. M., Böning, C., Bryan, F. O., Chassignet, E. P., Gerdes,
R., Hasumi, H., Hirst, A. C., Treguier, A.-M., and Webb, D.: Developments in
ocean climate modelling, J. Comput. Phys., 2, 123–192,
<a href="https://doi.org/10.1016/S1463-5003(00)00014-7" target="_blank">https://doi.org/10.1016/S1463-5003(00)00014-7</a>, 2000.
</mixed-citation></ref-html>
<ref-html id="bib1.bib46"><label>Griffies et al.(2009)</label><mixed-citation>
Griffies, S. M., Biastoch, A., Böning, C., Bryan, F. O., Danabasoglu,
G., Chassignet, E. P., England, M. H., Gerdes, R., Haak, H., Hallberg, R. W.,
Hazeleger, W., Jungclaus, J. H., Large, W. G., Madec, G., Pirani, A.,
Samuels, B. L., Scheinert, M., Gupta, A. S., Severijns, C. A., Simmons,
H. L., Treguier, A.-M., Winton, M., Yeager, S. G., and Yin, J.: Coordinated
Ocean-ice Reference Experiments (COREs), Ocean Model., 26, 1–46,
<a href="https://doi.org/10.1016/j.ocemod.2008.08.007" target="_blank">https://doi.org/10.1016/j.ocemod.2008.08.007</a>, 2009.
</mixed-citation></ref-html>
<ref-html id="bib1.bib47"><label>Gwyther(2016)</label><mixed-citation>
Gwyther, D. E.: Modelling regional ice shelf/ocean interaction with ROMS,
in: ROMS Asia-Pacific Workshop, Hobart, Australia, 2016.
</mixed-citation></ref-html>
<ref-html id="bib1.bib48"><label>Gwyther et al.(2014)</label><mixed-citation>
Gwyther, D. E., Galton-Fenzi, B. K., Hunter, J. R., and Roberts, J. L.:
Simulated melt rates for the Totten and Dalton ice shelves, Ocean Sci., 10,
267–279, https://doi.org/10.5194/os-10-267-2014, 2014.
</mixed-citation></ref-html>
<ref-html id="bib1.bib49"><label>Gwyther et al.(2015)</label><mixed-citation>
Gwyther, D. E., Galton-Fenzi, B. K., Dinniman, M. S., Roberts, J. L., and
Hunter, J. R.: The effect of basal friction on melting and freezing in ice
shelf-ocean models, Ocean Model., 95, 38–52,
<a href="https://doi.org/10.1016/j.ocemod.2015.09.004" target="_blank">https://doi.org/10.1016/j.ocemod.2015.09.004</a>, 2015.
</mixed-citation></ref-html>
<ref-html id="bib1.bib50"><label>Gwyther et al.(2016)</label><mixed-citation>
Gwyther, D. E., Cougnon, E. A., Galton-Fenzi, B. K., Roberts, J. L., Hunter,
J. R., and Dinniman, M. S.: Modelling the response of ice shelf basal
melting to different ocean cavity environmental regimes, Ann. Glaciol., 57,
131–141, <a href="https://doi.org/10.1017/aog.2016.31" target="_blank">https://doi.org/10.1017/aog.2016.31</a>, 2016.
</mixed-citation></ref-html>
<ref-html id="bib1.bib51"><label>Haas et al.(2008)</label><mixed-citation>
Haas, C., Nicolaus, M., Willmes, S., Worby, A., and Flinspach, D.: Sea ice
and snow thickness and physical properties of an ice floe in the western
Weddell Sea and their changes during spring warming, Deep-Sea Res. Pt. II,
55, 963–974, <a href="https://doi.org/10.1016/j.dsr2.2007.12.020" target="_blank">https://doi.org/10.1016/j.dsr2.2007.12.020</a>, 2008.
</mixed-citation></ref-html>
<ref-html id="bib1.bib52"><label>Haidvogel and Beckmann(1999)</label><mixed-citation>
Haidvogel, D. B. and Beckmann, A.: Numerical Ocean Circulation Modeling,
Imperial College Press, London, 1999.
</mixed-citation></ref-html>
<ref-html id="bib1.bib53"><label>Haney(1991)</label><mixed-citation>
Haney, R. L.: On the Pressure Gradient Force over Steep Topography in Sigma
Coordinate Ocean Models, J. Phys. Oceanogr., 21, 610–619,
<a href="https://doi.org/10.1175/1520-0485(1991)021&lt;0610:OTPGFO&gt;2.0.CO;2" target="_blank">https://doi.org/10.1175/1520-0485(1991)021&lt;0610:OTPGFO&gt;2.0.CO;2</a>, 1991.
</mixed-citation></ref-html>
<ref-html id="bib1.bib54"><label>Hattermann et al.(2012)</label><mixed-citation>
Hattermann, T., Nøst, O. A., Lilly, J. M., and Smedsrud, L. H.: Two years
of oceanic observations below the Fimbul Ice Shelf, Antarctica, Geophys.
Res. Lett., 39, L12605, <a href="https://doi.org/10.1029/2012GL051012" target="_blank">https://doi.org/10.1029/2012GL051012</a>, 2012.
</mixed-citation></ref-html>
<ref-html id="bib1.bib55"><label>Hellmer and Olbers(1989)</label><mixed-citation>
Hellmer, H. H. and Olbers, D. J.: A two-dimensional model for the
thermohaline circulation under an ice shelf, Antarct. Sci., 1, 325–336,
<a href="https://doi.org/10.1017/S0954102089000490" target="_blank">https://doi.org/10.1017/S0954102089000490</a>, 1989.
</mixed-citation></ref-html>
<ref-html id="bib1.bib56"><label>Hellmer et al.(1998)</label><mixed-citation>
Hellmer, H. H., Jacobs, S. S., and Jenkins, A.: Oceanic erosion of a
floating Antarctic glacier in the Amundsen Sea, in: Ocean, Ice, and
Atmosphere: Interactions at the Antarctic Continental Margin, edited by:
Jacobs, S. S. and Weiss, R. F., American Geophysical Union, Washington, DC,
319–339, <a href="https://doi.org/10.1029/AR075p0341" target="_blank">https://doi.org/10.1029/AR075p0341</a>, 1998.
</mixed-citation></ref-html>
<ref-html id="bib1.bib57"><label>Hellmer et al.(2012)</label><mixed-citation>
Hellmer, H. H., Kauker, F., Timmermann, R., Determann, J., and Rae, J.:
Twenty-first-century warming of a large Antarctic ice-shelf cavity by a
redirected coastal current, Nature, 485, 225–228,
<a href="https://doi.org/10.1038/nature11064" target="_blank">https://doi.org/10.1038/nature11064</a>, 2012.
</mixed-citation></ref-html>
<ref-html id="bib1.bib58"><label>Hellmer et al.(2017)</label><mixed-citation>
Hellmer, H. H., Kauker, F., Timmermann, R., and Hattermann, T.: The fate of
the southern Weddell Sea continental shelf in a warming climate, J. Climate,
30, 4337–4350, <a href="https://doi.org/10.1175/JCLI-D-16-0420.1" target="_blank">https://doi.org/10.1175/JCLI-D-16-0420.1</a>, 2017.
</mixed-citation></ref-html>
<ref-html id="bib1.bib59"><label>Herraiz-Borreguero et al.(2016)</label><mixed-citation>
Herraiz-Borreguero, L., Church, J. A., Allison, I., Pena-Molino, B., Coleman,
R., Tomczak, M., and Craven, M.: Basal melt, seasonal water mass
transformation, ocean current variability, and deep convection processes
along the Amery Ice Shelf, J. Geophys. Res.-Oceans, 121, 4946–4965,
<a href="https://doi.org/10.1002/2013JC008846" target="_blank">https://doi.org/10.1002/2013JC008846</a>, 2016.
</mixed-citation></ref-html>
<ref-html id="bib1.bib60"><label>Heuzé et al.(2015)</label><mixed-citation>
Heuzé, C., Ridley, J. K., Calvert, D., Stevens, D. P., and Heywood, K.
J.: Increasing vertical mixing to reduce Southern Ocean deep convection in
NEMO3.4, Geosci. Model Dev., 8, 3119–3130,
https://doi.org/10.5194/gmd-8-3119-2015, 2015.
</mixed-citation></ref-html>
<ref-html id="bib1.bib61"><label>Hogg and Gudmundsson(2017)</label><mixed-citation>
Hogg, A. E. and Gudmundsson, G. H.: Impacts of the Larsen-C Ice Shelf
calving event, Nat. Clim. Change, 7, 540–542, <a href="https://doi.org/10.1038/nclimate3359" target="_blank">https://doi.org/10.1038/nclimate3359</a>,
2017.
</mixed-citation></ref-html>
<ref-html id="bib1.bib62"><label>Holland and Jenkins(1999)</label><mixed-citation>
Holland, D. M. and Jenkins, A.: Modeling Thermodynamic Ice-Ocean
Interactions at the Base of an Ice Shelf, J. Phys. Oceanogr., 29,
1787–1800, <a href="https://doi.org/10.1175/1520-0485(1999)029&lt;1787:MTIOIA&gt;2.0.CO;2" target="_blank">https://doi.org/10.1175/1520-0485(1999)029&lt;1787:MTIOIA&gt;2.0.CO;2</a>, 1999.
</mixed-citation></ref-html>
<ref-html id="bib1.bib63"><label>Holland et al.(2009)</label><mixed-citation>
Holland, P. R., Corr, H. F. J., Vaughan, D. G., Jenkins, A., and Skvarca, P.:
Marine ice in Larsen Ice Shelf, Geophys. Res. Lett., 36, 1–6,
<a href="https://doi.org/10.1029/2009GL038162" target="_blank">https://doi.org/10.1029/2009GL038162</a>, 2009.
</mixed-citation></ref-html>
<ref-html id="bib1.bib64"><label>Holland et al.(2014)</label><mixed-citation>
Holland, P. R., Bruneau, N., Enright, C., Losch, M., Kurtz, N. T., and Kwok,
R.: Modeled trends in Antarctic Sea Ice Thickness, J. Climate, 27,
3784–3801, <a href="https://doi.org/10.1175/JCLI-D-13-00301.1" target="_blank">https://doi.org/10.1175/JCLI-D-13-00301.1</a>, 2014.
</mixed-citation></ref-html>
<ref-html id="bib1.bib65"><label>Hunke et al.(2013)</label><mixed-citation>
Hunke, E. C., Hebert, D. A., and Lecomte, O.: Level-ice melt ponds in the
Los Alamos sea ice model, CICE, Ocean Model., 71, 26–42,
<a href="https://doi.org/10.1016/j.ocemod.2012.11.008" target="_blank">https://doi.org/10.1016/j.ocemod.2012.11.008</a>, 2013.
</mixed-citation></ref-html>
<ref-html id="bib1.bib66"><label>Hunke et al.(2015)</label><mixed-citation>
Hunke, E. C., Lipscomb, W. H., Turner, A. K., Jeffery, N., and Elliott, S.:
CICE : the Los Alamos Sea Ice Model Documentation and Software User's Manual
LA-CC-06-012, Tech. rep., Los Alamos National Laboratory, 2015.
</mixed-citation></ref-html>
<ref-html id="bib1.bib67"><label>Hunter(2006)</label><mixed-citation>
Hunter, J. R.: Specification for Test Models of Ice Shelf Cavities, Tech.
rep., Antarctic Climate &amp; Ecosystems Cooperative Research Centre, 2006.
</mixed-citation></ref-html>
<ref-html id="bib1.bib68"><label>Jacob et al.(2005)</label><mixed-citation>
Jacob, R., Larson, J., and Ong, E.: M&thinsp; × &thinsp;N Communication and
Parallel Interpolation in Community Climate System Model Version 3 Using the
Model Coupling Toolkit, Int. J. High Perform. C, 19, 293–307,
<a href="https://doi.org/10.1177/1094342005056116" target="_blank">https://doi.org/10.1177/1094342005056116</a>, 2005.
</mixed-citation></ref-html>
<ref-html id="bib1.bib69"><label>Jacobs and Comiso(1989)</label><mixed-citation>
Jacobs, S. S. and Comiso, J. C.: Sea ice and oceanic processes on the Ross
Sea continental shelf, J. Geophys. Res., 94, 18195–18211,
<a href="https://doi.org/10.1029/JC094iC12p18195" target="_blank">https://doi.org/10.1029/JC094iC12p18195</a>, 1989.
</mixed-citation></ref-html>
<ref-html id="bib1.bib70"><label>Jacobs et al.(1979)</label><mixed-citation>
Jacobs, S. S., Gordon, A. L., and Ardai, J. L.: Circulation and melting
beneath the Ross Ice Shelf, Science, 203, 439–443,
<a href="https://doi.org/10.1126/science.203.4379.439" target="_blank">https://doi.org/10.1126/science.203.4379.439</a>, 1979.
</mixed-citation></ref-html>
<ref-html id="bib1.bib71"><label>Jacobs et al.(1992)</label><mixed-citation>
Jacobs, S. S., Hellmer, H. H., Doake, C. S. M., Jenkins, A., and Frolich,
R. M.: Melting of ice shelves and the mass balance of Antarctica, J.
Glaciol., 38, 375–387, <a href="https://doi.org/10.1038/228047a0" target="_blank">https://doi.org/10.1038/228047a0</a>, 1992.
</mixed-citation></ref-html>
<ref-html id="bib1.bib72"><label>Jacobs et al.(2011)</label><mixed-citation>
Jacobs, S. S., Jenkins, A., Giulivi, C. F., and Dutrieux, P.: Stronger ocean
circulation and increased melting under Pine Island Glacier ice shelf, Nat.
Geosci., 4, 519–523, <a href="https://doi.org/10.1038/ngeo1188" target="_blank">https://doi.org/10.1038/ngeo1188</a>, 2011.
</mixed-citation></ref-html>
<ref-html id="bib1.bib73"><label>Jenkins(1991)</label><mixed-citation>
Jenkins, A.: A One-Dimensional Model of Ice Shelf-Ocean Interaction, J.
Geophys. Res.-Oceans, 96, 20671–20677, <a href="https://doi.org/10.1029/91JC01842" target="_blank">https://doi.org/10.1029/91JC01842</a>, 1991.
</mixed-citation></ref-html>
<ref-html id="bib1.bib74"><label>Jenkins(2011)</label><mixed-citation>
Jenkins, A.: Convection-Driven Melting near the Grounding Lines of Ice
Shelves and Tidewater Glaciers, J. Phys. Oceanogr., 41, 2279–2294,
<a href="https://doi.org/10.1175/JPO-D-11-03.1" target="_blank">https://doi.org/10.1175/JPO-D-11-03.1</a>, 2011.
</mixed-citation></ref-html>
<ref-html id="bib1.bib75"><label>Jenkins(2016)</label><mixed-citation>
Jenkins, A.: A Simple Model of the Ice Shelf–Ocean Boundary Layer and
Current, J. Phys. Oceanogr., 46, 1785–1803, <a href="https://doi.org/10.1175/JPO-D-15-0194.1" target="_blank">https://doi.org/10.1175/JPO-D-15-0194.1</a>,
2016.
</mixed-citation></ref-html>
<ref-html id="bib1.bib76"><label>Jenkins and Jacobs(2008)</label><mixed-citation>
Jenkins, A. and Jacobs, S. S.: Circulation and melting beneath George VI Ice
Shelf, Antarctica, J. Geophys. Res., 113, 1–18, <a href="https://doi.org/10.1029/2007JC004449" target="_blank">https://doi.org/10.1029/2007JC004449</a>,
2008.
</mixed-citation></ref-html>
<ref-html id="bib1.bib77"><label>Jenkins et al.(2010)</label><mixed-citation>
Jenkins, A., Dutrieux, P., Jacobs, S. S., McPhail, S. D., Perrett, J. R.,
Webb, A. T., and White, D.: Observations beneath Pine Island Glacier in West
Antarctica and implications for its retreat, Nat. Geosci., 3, 468–472,
<a href="https://doi.org/10.1038/ngeo890" target="_blank">https://doi.org/10.1038/ngeo890</a>, 2010.
</mixed-citation></ref-html>
<ref-html id="bib1.bib78"><label>Joughin and Alley(2011)</label><mixed-citation>
Joughin, I. and Alley, R. B.: Stability of the West Antarctic ice sheet in a
warming world, Nat. Geosci., 4, 506–513, <a href="https://doi.org/10.1038/ngeo1194" target="_blank">https://doi.org/10.1038/ngeo1194</a>, 2011.
</mixed-citation></ref-html>
<ref-html id="bib1.bib79"><label>Joughin and Padman(2003)</label><mixed-citation>
Joughin, I. and Padman, L.: Melting and freezing beneath Filchner-Ronne Ice
Shelf, Antarctica, Geophys. Res. Lett., 30, 1477,
<a href="https://doi.org/10.1029/2003GL016941" target="_blank">https://doi.org/10.1029/2003GL016941</a>, 2003.
</mixed-citation></ref-html>
<ref-html id="bib1.bib80"><label>Khazendar et al.(2013)</label><mixed-citation>
Khazendar, A., Schodlok, M. P., Fenty, I. G., Ligtenberg, S. R. M., Rignot,
E. J., and van den Broeke, M. R.: Observed thinning of Totten Glacier is
linked to coastal polynya variability, Nat. Commun., 4, 2857,
<a href="https://doi.org/10.1038/ncomms3857" target="_blank">https://doi.org/10.1038/ncomms3857</a>, 2013.
</mixed-citation></ref-html>
<ref-html id="bib1.bib81"><label>Kjellsson et al.(2015)</label><mixed-citation>
Kjellsson, J., Holland, P. R., Marshall, G. J., Mathiot, P., Aksenov, Y.,
Coward, A. C., Bacon, S., Megann, A. P., and Ridley, J. K.: Model
sensitivity of the Weddell and Ross seas, Antarctica, to vertical mixing and
freshwater forcing, Ocean Model., 94, 141–152,
<a href="https://doi.org/10.1016/j.ocemod.2015.08.003" target="_blank">https://doi.org/10.1016/j.ocemod.2015.08.003</a>, 2015.
</mixed-citation></ref-html>
<ref-html id="bib1.bib82"><label>Koltermann et al.(2011)</label><mixed-citation>
Koltermann, K. P., Gouretski, V. V., and Jancke, K.: Hydrographic Atlas of
the World Ocean Circulation Experiment (WOCE), vol. 3, in: Atlantic Ocean
edited by: Sparrow, M., Chapman, P., and Gould, J., Tech. rep., International
WOCE Project Office, Southampton, UK, 2011.
</mixed-citation></ref-html>
<ref-html id="bib1.bib83"><label>Kurtz and Markus(2012)</label><mixed-citation>
Kurtz, N. T. and Markus, T.: Satellite observations of Antarctic sea ice
thickness and volume, J. Geophys. Res.-Oceans, 117, 1–9,
<a href="https://doi.org/10.1029/2012JC008141" target="_blank">https://doi.org/10.1029/2012JC008141</a>, 2012.
</mixed-citation></ref-html>
<ref-html id="bib1.bib84"><label>Kusahara et al.(2010)</label><mixed-citation>
Kusahara, K., Hasumi, H., and Tamura, T.: Modeling sea ice production and
dense shelf water formation in coastal polynyas around East Antarctica, J.
Geophys. Res.-Oceans, 115, C10006, <a href="https://doi.org/10.1029/2010JC006133" target="_blank">https://doi.org/10.1029/2010JC006133</a>, 2010.
</mixed-citation></ref-html>
<ref-html id="bib1.bib85"><label>Kusahara et al.(2017)</label><mixed-citation>
Kusahara, K., Williams, G. D., Massom, R., Reid, P., and Hasumi, H.: Roles
of wind stress and thermodynamic forcing in recent trends in Antarctic sea
ice and Southern Ocean SST: An ocean-sea ice model study, Global Planet.
Change, 158, 103–118, <a href="https://doi.org/10.1016/j.gloplacha.2017.09.012" target="_blank">https://doi.org/10.1016/j.gloplacha.2017.09.012</a>, 2017.
</mixed-citation></ref-html>
<ref-html id="bib1.bib86"><label>Kwok and Cunningham(2008)</label><mixed-citation>
Kwok, R. and Cunningham, G. F.: ICESat over Arctic sea ice: Estimation of
snow depth and ice thickness, J. Geophys. Res.-Oceans, 113, 1–17,
<a href="https://doi.org/10.1029/2008JC004753" target="_blank">https://doi.org/10.1029/2008JC004753</a>, 2008.
</mixed-citation></ref-html>
<ref-html id="bib1.bib87"><label>Langhorne et al.(2015)</label><mixed-citation>
Langhorne, P. J., Hughes, K. G., Gough, A. J., Smith, I. J., Williams, M.
J. M., Robinson, N. J., Stevens, C. L., Rack, W., Price, D., Leonard, G. H.,
Mahoney, A. R., Haas, C., and Haskell, T. G.: Observed platelet ice
distributions in Antarctic sea ice: An index for ocean-ice shelf heat flux,
Geophys. Res. Lett., 42, 5442–5451, <a href="https://doi.org/10.1002/2015GL064508" target="_blank">https://doi.org/10.1002/2015GL064508</a>, 2015.
</mixed-citation></ref-html>
<ref-html id="bib1.bib88"><label>Langley et al.(2014)</label><mixed-citation>
Langley, K. A., Kohler, J., Sinisalo, A., Øyan, M. J., Hamran, S. E.,
Hattermann, T., Matsuoka, K., Nøst, O. A., and Isaksson, E.: Low melt
rates with seasonal variability at the base of Fimbul Ice Shelf, East
Antarctica, revealed by in situ interferometric radar measurements, Geophys.
Res. Lett., 41, 8138–8146, <a href="https://doi.org/10.1002/2014GL061782" target="_blank">https://doi.org/10.1002/2014GL061782</a>, 2014.
</mixed-citation></ref-html>
<ref-html id="bib1.bib89"><label>Large et al.(1994)</label><mixed-citation>
Large, W. G., McWilliams, J. C., and Doney, S. C.: Oceanic Vertical Mixing
– a Review and a Model with a Nonlocal Boundary-Layer Parameterization,
Rev. Geophys., 32, 363–403, <a href="https://doi.org/10.1029/94rg01872" target="_blank">https://doi.org/10.1029/94rg01872</a>, 1994.
</mixed-citation></ref-html>
<ref-html id="bib1.bib90"><label>Larson et al.(2005)</label><mixed-citation>
Larson, J., Jacob, R., and Ong, E.: The Model Coupling Toolkit: A New
Fortran90 Toolkit for Building Multiphysics Parallel Coupled Models, Int. J.
High Perform. C., 19, 277–292, <a href="https://doi.org/10.1177/1094342005056115" target="_blank">https://doi.org/10.1177/1094342005056115</a>, 2005.
</mixed-citation></ref-html>
<ref-html id="bib1.bib91"><label>Lemarié et al.(2012)</label><mixed-citation>
Lemarié, F., Kurian, J., Shchepetkin, A. F., Jeroen Molemaker, M.,
Colas, F., and McWilliams, J. C.: Are there inescapable issues prohibiting
the use of terrain-following coordinates in climate models?, Ocean Model.,
42, 57–79, <a href="https://doi.org/10.1016/j.ocemod.2011.11.007" target="_blank">https://doi.org/10.1016/j.ocemod.2011.11.007</a>, 2012.
</mixed-citation></ref-html>
<ref-html id="bib1.bib92"><label>Lenaerts et al.(2012)</label><mixed-citation>
Lenaerts, J. T. M., van den Broeke, M. R., van de Berg, W. J., van Meijgaard,
E., and Munneke, P. K.: A new, high-resolution surface mass balance map of
Antarctica (1979–2010) based on regional atmospheric climate modeling,
Geophys. Res. Lett., 39, 1–5, <a href="https://doi.org/10.1029/2011GL050713" target="_blank">https://doi.org/10.1029/2011GL050713</a>, 2012.
</mixed-citation></ref-html>
<ref-html id="bib1.bib93"><label>Lewis and Perkin(1986)</label><mixed-citation>
Lewis, E. L. and Perkin, R. G.: Ice pumps and their rates, J. Geophys.
Res., 91, 11756–11762 , <a href="https://doi.org/10.1029/JC091iC10p11756" target="_blank">https://doi.org/10.1029/JC091iC10p11756</a>, 1986.
</mixed-citation></ref-html>
<ref-html id="bib1.bib94"><label>Lipscomb and Hunke(2004)</label><mixed-citation>
Lipscomb, W. H. and Hunke, E. C.: Modeling Sea Ice Transport Using
Incremental Remapping, Mon. Weather Rev., 132, 1341–1354,
<a href="https://doi.org/10.1175/1520-0493(2004)132&lt;1341:MSITUI&gt;2.0.CO;2" target="_blank">https://doi.org/10.1175/1520-0493(2004)132&lt;1341:MSITUI&gt;2.0.CO;2</a>, 2004.
</mixed-citation></ref-html>
<ref-html id="bib1.bib95"><label>Lipscomb et al.(2007)</label><mixed-citation>
Lipscomb, W. H., Hunke, E. C., Maslowski, W., and Jakacki, J.: Ridging,
strength, and stability in high-resolution sea ice models, J. Geophys.
Res.-Oceans, 112, 1–18, <a href="https://doi.org/10.1029/2005JC003355" target="_blank">https://doi.org/10.1029/2005JC003355</a>, 2007.
</mixed-citation></ref-html>
<ref-html id="bib1.bib96"><label>Makinson et al.(2011)</label><mixed-citation>
Makinson, K., Holland, P. R., Jenkins, A., Nicholls, K. W., and Holland,
D. M.: Influence of tides on melting and freezing beneath Filchner-Ronne Ice
Shelf, Antarctica, Geophys. Res. Lett., 38, L06601,
<a href="https://doi.org/10.1029/2010GL046462" target="_blank">https://doi.org/10.1029/2010GL046462</a>, 2011.
</mixed-citation></ref-html>
<ref-html id="bib1.bib97"><label>Marchesiello et al.(2001)</label><mixed-citation>
Marchesiello, P., McWilliams, J. C., and Shchepetkin, A. F.: Open boundary
conditions for long-term integration of regional oceanic models, Ocean
Model., 3, 1–20, <a href="https://doi.org/10.1016/S1463-5003(00)00013-5" target="_blank">https://doi.org/10.1016/S1463-5003(00)00013-5</a>, 2001.
</mixed-citation></ref-html>
<ref-html id="bib1.bib98"><label>Marchesiello et al.(2009)</label><mixed-citation>
Marchesiello, P., Debreu, L., and Couvelard, X.: Spurious diapycnal mixing
in terrain-following coordinate models: The problem and a solution, Ocean
Model., 26, 156–169, <a href="https://doi.org/10.1016/j.ocemod.2008.09.004" target="_blank">https://doi.org/10.1016/j.ocemod.2008.09.004</a>, 2009.
</mixed-citation></ref-html>
<ref-html id="bib1.bib99"><label>Martin and Adcroft(2010)</label><mixed-citation>
Martin, T. and Adcroft, A.: Parameterizing the fresh-water flux from land
ice to ocean with interactive icebergs in a coupled climate model, Ocean
Model., 34, 111–124, <a href="https://doi.org/10.1016/j.ocemod.2010.05.001" target="_blank">https://doi.org/10.1016/j.ocemod.2010.05.001</a>, 2010.
</mixed-citation></ref-html>
<ref-html id="bib1.bib100"><label>Martinson and McKee(2012)</label><mixed-citation>
Martinson, D. G. and McKee, D. C.: Transport of warm Upper Circumpolar Deep
Water onto the western Antarctic Peninsula continental shelf, Ocean Sci., 8,
433–442, https://doi.org/10.5194/os-8-433-2012, 2012.
</mixed-citation></ref-html>
<ref-html id="bib1.bib101"><label>Mathiot et al.(2017)</label><mixed-citation>
Mathiot, P., Jenkins, A., Harris, C., and Madec, G.: Explicit representation
and parametrised impacts of under ice shelf seas in the <i>z</i>* coordinate
ocean model NEMO 3.6, Geosci. Model Dev., 10, 2849–2874,
https://doi.org/10.5194/gmd-10-2849-2017, 2017.
</mixed-citation></ref-html>
<ref-html id="bib1.bib102"><label>McPhail et al.(2009)</label><mixed-citation>
McPhail, S. D., Furlong, M. E., Pebody, M., Perrett, J. R., Stevenson, P.,
Webb, A., and White, D.: Exploring beneath the PIG Ice Shelf with the
Autosub3 AUV, in: OCEANS '09 IEEE Bremen: Balancing Technology with Future
Needs, Piscataway, USA, p. 6, <a href="https://doi.org/10.1109/OCEANSE.2009.5278170" target="_blank">https://doi.org/10.1109/OCEANSE.2009.5278170</a>, 2009.
</mixed-citation></ref-html>
<ref-html id="bib1.bib103"><label>McPhee et al.(1987)</label><mixed-citation>
McPhee, M. G., Maykut, G. A., and Morison, J. H.: Dynamics and
thermodynamics of the ice/upper ocean system in the marginal ice zone of the
Greenland Sea, J. Geophys. Res.-Oceans, 92, 7017–7031,
<a href="https://doi.org/10.1029/JC092iC07p07017" target="_blank">https://doi.org/10.1029/JC092iC07p07017</a>, 1987.
</mixed-citation></ref-html>
<ref-html id="bib1.bib104"><label>Meehl et al.(2000)</label><mixed-citation>
Meehl, G. A., Boer, G. J., Covey, C., Latif, M., and Stouffer, R. J.: The
Coupled Model Intercomparison Project (CMIP), B. Am. Meteorol. Soc., 81,
313–318, <a href="https://doi.org/10.1175/1520-0477(2000)081&lt;0313:TCMIPC&gt;2.3.CO;2" target="_blank">https://doi.org/10.1175/1520-0477(2000)081&lt;0313:TCMIPC&gt;2.3.CO;2</a>, 2000.
</mixed-citation></ref-html>
<ref-html id="bib1.bib105"><label>Meier et al.(2013)</label><mixed-citation>
Meier, W., Fetterrer, F., Savoie, M., Mallory, S., Duerr, R., and Stroeve,
J.: NOAA/NSIDC Climate Data Record of Passive Microwave Sea Ice
Concentration, Version 2, <a href="https://doi.org/10.7265/N55M63M1" target="_blank">https://doi.org/10.7265/N55M63M1</a>, 2013.
</mixed-citation></ref-html>
<ref-html id="bib1.bib106"><label>Menemenlis et al.(2008)</label><mixed-citation>
Menemenlis, D., Campin, J.-M., Heimbach, P., Hill, C., Lee, T., Nguyen,
A. T., Schodlok, M. P., and Zhang, H.: ECCO2: High resolution global ocean
and sea ice data synthesis, Mercator Ocean Quarterly Newsletter, 31, 13–21,
2008.
</mixed-citation></ref-html>
<ref-html id="bib1.bib107"><label>Mengel and Levermann(2014)</label><mixed-citation>
Mengel, M. and Levermann, A.: Ice plug prevents irreversible discharge from
East Antarctica, Nat. Clim. Change, 4, 451–455, <a href="https://doi.org/10.1038/nclimate2226" target="_blank">https://doi.org/10.1038/nclimate2226</a>,
2014.
</mixed-citation></ref-html>
<ref-html id="bib1.bib108"><label>Miller et al.(2012)</label><mixed-citation>
Miller, K. G., Wright, J. D., Browning, J. V., Kulpecz, A., Kominz, M.,
Naish, T. R., Cramer, B. S., Rosenthal, Y., Peltier, W. R., and Sosdian, S.:
High tide of the warm Pliocene: Implications of global sea level for
Antarctic deglaciation, Geology, 40, 407–410, <a href="https://doi.org/10.1130/G32869.1" target="_blank">https://doi.org/10.1130/G32869.1</a>, 2012.
</mixed-citation></ref-html>
<ref-html id="bib1.bib109"><label>Moffat et al.(2009)</label><mixed-citation>
Moffat, C., Owens, B., and Beardsley, R. C.: On the characteristics of
Circumpolar Deep Water intrusions to the west Antarctic Peninsula Continental
Shelf, J. Geophys. Res.-Oceans, 114, C05017, <a href="https://doi.org/10.1029/2008JC004955" target="_blank">https://doi.org/10.1029/2008JC004955</a>,
2009.
</mixed-citation></ref-html>
<ref-html id="bib1.bib110"><label>Mueller et al.(2012)</label><mixed-citation>
Mueller, R. D., Padman, L., Dinniman, M. S., Erofeeva, S. Y., Fricker, H. A.,
and King, M. A.: Impact of tide-topography interactions on basal melting of
Larsen C Ice Shelf, Antarctica, J. Geophys. Res.-Oceans, 117, C05005,
<a href="https://doi.org/10.1029/2011JC007263" target="_blank">https://doi.org/10.1029/2011JC007263</a>, 2012.
</mixed-citation></ref-html>
<ref-html id="bib1.bib111"><label>Mueller et al.(2018)</label><mixed-citation>
Mueller, R. D., Hattermann, T., Howard, S. L., and Padman, L.: Tidal
influences on a future evolution of the Filchner–Ronne Ice Shelf cavity in
the Weddell Sea, Antarctica, The Cryosphere, 12, 453–476,
https://doi.org/10.5194/tc-12-453-2018, 2018.
</mixed-citation></ref-html>
<ref-html id="bib1.bib112"><label>Müller et al.(2010)</label><mixed-citation>
Müller, M., Haak, H., Jungclaus, J. H., Sündermann, J., and
Thomas, M.: The effect of ocean tides on a climate model simulation, Ocean
Model., 35, 304–313, <a href="https://doi.org/10.1016/j.ocemod.2010.09.001" target="_blank">https://doi.org/10.1016/j.ocemod.2010.09.001</a>, 2010.
</mixed-citation></ref-html>
<ref-html id="bib1.bib113"><label>Nakayama et al.(2014)</label><mixed-citation>
Nakayama, Y., Timmermann, R., Schröder, M., and Hellmer, H. H.: On the
difficulty of modeling Circumpolar Deep Water intrusions onto the Amundsen
Sea continental shelf, Ocean Model., 84, 26–34,
<a href="https://doi.org/10.1016/j.ocemod.2014.09.007" target="_blank">https://doi.org/10.1016/j.ocemod.2014.09.007</a>, 2014.
</mixed-citation></ref-html>
<ref-html id="bib1.bib114"><label>Naud et al.(2014)</label><mixed-citation>
Naud, C. M., Booth, J. F., and Del Genio, A. D.: Evaluation of ERA-Interim
and MERRA cloudiness in the Southern Ocean, J. Climate, 27, 2109–2124,
<a href="https://doi.org/10.1175/JCLI-D-13-00432.1" target="_blank">https://doi.org/10.1175/JCLI-D-13-00432.1</a>, 2014.
</mixed-citation></ref-html>
<ref-html id="bib1.bib115"><label>Naughten et al.(2017)</label><mixed-citation>
Naughten, K. A., Galton-Fenzi, B. K., Meissner, K. J., England, M. H.,
Brassington, G. B., Colberg, F., Hattermann, T., and Debernard, J. B.:
Spurious sea ice formation caused by oscillatory ocean tracer advection
schemes, Ocean Model., 116, 108–117, <a href="https://doi.org/10.1016/j.ocemod.2017.06.010" target="_blank">https://doi.org/10.1016/j.ocemod.2017.06.010</a>,
2017.
</mixed-citation></ref-html>
<ref-html id="bib1.bib116"><label>1</label><mixed-citation>
Naughten, K. A., Meissner, K. J., Galton-Fenzi, B. K., England, M. H., Timmermann, R., and Hellmer, H. H.: MetROMS-iceshelf, <a href="https://doi.org/10.5281/zenodo.1157230" target="_blank">https://doi.org/10.5281/zenodo.1157230</a>, 2018a.
</mixed-citation></ref-html>
<ref-html id="bib1.bib117"><label>2</label><mixed-citation>
Naughten, K. A., Meissner, K. J., Galton-Fenzi, B. K., England, M. H., Timmermann, R., and Hellmer, H. H.: Finite Element Sea-ice/ice-shelf Ocean Model (FESOM) 1.4, <a href="https://doi.org/10.5281/zenodo.1157228" target="_blank">https://doi.org/10.5281/zenodo.1157228</a>, 2018b.
</mixed-citation></ref-html>
<ref-html id="bib1.bib118"><label>Nicholls and Johnson(2001)</label><mixed-citation>
Nicholls, K. W. and Johnson, M. R.: Oceanographic conditions south of
Berkner Island, beneath Filchner-Ronne Ice Shelf, Antarctica, J. Geophys.
Res., 106, 11481–11492, <a href="https://doi.org/10.1029/2000JC000350" target="_blank">https://doi.org/10.1029/2000JC000350</a>, 2001.
</mixed-citation></ref-html>
<ref-html id="bib1.bib119"><label>Nicholls and Østerhus(2004)</label><mixed-citation>
Nicholls, K. W. and Østerhus, S.: Interannual variability and ventilation
timescales in the ocean cavity beneath Filchner-Ronne Ice Shelf, Antarctica,
J. Geophys. Res.-Oceans, 109, 1–9, <a href="https://doi.org/10.1029/2003JC002149" target="_blank">https://doi.org/10.1029/2003JC002149</a>, 2004.
</mixed-citation></ref-html>
<ref-html id="bib1.bib120"><label>Nicholls et al.(2004)</label><mixed-citation>
Nicholls, K. W., Makinson, K., and Østerhus, S.: Circulation and water
masses beneath the northern Ronne Ice Shelf, Antarctica, J. Geophys.
Res.-Oceans, 109, 1–11, <a href="https://doi.org/10.1029/2004JC002302" target="_blank">https://doi.org/10.1029/2004JC002302</a>, 2004.
</mixed-citation></ref-html>
<ref-html id="bib1.bib121"><label>Nicholls et al.(2006)</label><mixed-citation>
Nicholls, K. W., Abrahamsen, E. P., Buck, J. J. H., Dodd, P. A., Goldblatt,
C., Griffiths, G., Heywood, K. J., Hughes, N. E., Kaletzky, A., Lane-Serff,
G. F., McPhail, S. D., Millard, N. W., Oliver, K. I. C., Perrett, J., Price,
M. R., Pudsey, C. J., Saw, K., Stansfield, K., Stott, M. J., Wadhams, P.,
Webb, A. T., and Wilkinson, J. P.: Measurements beneath an Antarctic ice
shelf using an autonomous underwater vehicle, Geophys. Res. Lett., 33, 2–5,
<a href="https://doi.org/10.1029/2006GL025998" target="_blank">https://doi.org/10.1029/2006GL025998</a>, 2006.
</mixed-citation></ref-html>
<ref-html id="bib1.bib122"><label>Nicholls et al.(2009)</label><mixed-citation>
Nicholls, K. W., Østerhus, S., Makinson, K., Gammelsrød, T., and
Fahrbach, E.: Ice-ocean processes over the continental shelf of the southern
Weddell Sea, Antarctica: a review, Rev. Geophys., 47, 1–23,
<a href="https://doi.org/10.1029/2007RG000250" target="_blank">https://doi.org/10.1029/2007RG000250</a>, 2009.
</mixed-citation></ref-html>
<ref-html id="bib1.bib123"><label>O'Leary et al.(2013)</label><mixed-citation>
O'Leary, M. J., Hearty, P. J., Thompson, W. G., Raymo, M. E., Mitrovica,
J. X., Webster, J. M., and O'Leary, M. J.: Ice sheet collapse following a
prolonged period of stable sea level during the last interglacial, Nat.
Geosci., 6, 796–800, <a href="https://doi.org/10.1038/ngeo1890" target="_blank">https://doi.org/10.1038/ngeo1890</a>, 2013.
</mixed-citation></ref-html>
<ref-html id="bib1.bib124"><label>Pacanowski and Philander(1981)</label><mixed-citation>
Pacanowski, R. C. and Philander, S. G. H.: Parameterization of Vertical
Mixing in Numerical Models of Tropical Oceans, J. Phys. Oceanogr., 11,
1443–1451, <a href="https://doi.org/10.1175/1520-0485(1981)011&lt;1443:POVMIN&gt;2.0.CO;2" target="_blank">https://doi.org/10.1175/1520-0485(1981)011&lt;1443:POVMIN&gt;2.0.CO;2</a>, 1981.
</mixed-citation></ref-html>
<ref-html id="bib1.bib125"><label>Padman et al.(2009)Padman, Howard, Orsi, and Muench</label><mixed-citation>
Padman, L., Howard, S. L., Orsi, A. H., and Muench, R. D.: Tides of the
northwestern Ross Sea and their impact on dense outflows of Antarctic Bottom
Water, Deep-Sea Res. Pt. II, 56, 818–834, <a href="https://doi.org/10.1016/j.dsr2.2008.10.026" target="_blank">https://doi.org/10.1016/j.dsr2.2008.10.026</a>,
2009.
</mixed-citation></ref-html>
<ref-html id="bib1.bib126"><label>Paolo et al.(2015)</label><mixed-citation>
Paolo, F. S., Fricker, H. A., and Padman, L.: Volume loss from Antarctic ice
shelves is accelerating, Science, 348, 327–331,
<a href="https://doi.org/10.1126/science.aaa0940" target="_blank">https://doi.org/10.1126/science.aaa0940</a>, 2015.
</mixed-citation></ref-html>
<ref-html id="bib1.bib127"><label>Parkinson and Washington(1979)</label><mixed-citation>
Parkinson, C. L. and Washington, W. M.: A large-scale numerical model of sea
ice, J. Geophys. Res., 84, 311–337, <a href="https://doi.org/10.1029/JC084iC01p00311" target="_blank">https://doi.org/10.1029/JC084iC01p00311</a>, 1979.
</mixed-citation></ref-html>
<ref-html id="bib1.bib128"><label>Pellichero et al.(2017)</label><mixed-citation>
Pellichero, V., Sallée, J.-B., Schmidtko, S., Roquet, F., and
Charrassin, J.-B.: The ocean mixed layer under Southern Ocean sea-ice:
Seasonal cycle and forcing, J. Geophys. Res.-Oceans, 122, 1608–1633,
<a href="https://doi.org/10.1002/2016JC011970" target="_blank">https://doi.org/10.1002/2016JC011970</a>, 2017.
</mixed-citation></ref-html>
<ref-html id="bib1.bib129"><label>Petty et al.(2013)</label><mixed-citation>
Petty, A. A., Feltham, D. L., and Holland, P. R.: Impact of Atmospheric
Forcing on Antarctic Continental Shelf Water Masses, J. Phys. Oceanogr., 43,
920–940, <a href="https://doi.org/10.1175/JPO-D-12-0172.1" target="_blank">https://doi.org/10.1175/JPO-D-12-0172.1</a>, 2013.
</mixed-citation></ref-html>
<ref-html id="bib1.bib130"><label>Petty et al.(2014)Petty, Holland, and Feltham</label><mixed-citation>
Petty, A. A., Holland, P. R., and Feltham, D. L.: Sea ice and the ocean mixed
layer over the Antarctic shelf seas, The Cryosphere, 8, 761–783,
https://doi.org/10.5194/tc-8-761-2014, 2014.
</mixed-citation></ref-html>
<ref-html id="bib1.bib131"><label>Pritchard et al.(2012)</label><mixed-citation>
Pritchard, H. D., Ligtenberg, S. R. M., Fricker, H. A., Vaughan, D. G.,
van den Broeke, M. R., and Padman, L.: Antarctic ice-sheet loss driven by
basal melting of ice shelves, Nature, 484, 502–505,
<a href="https://doi.org/10.1038/nature10968" target="_blank">https://doi.org/10.1038/nature10968</a>, 2012.
</mixed-citation></ref-html>
<ref-html id="bib1.bib132"><label>Rack and Rott(2004)</label><mixed-citation>
Rack, W. and Rott, H.: Pattern of retreat and disintegration of the Larsen B
Ice Shelf, Ann. Glaciol., 39, 505–510, <a href="https://doi.org/10.3189/172756404781814005" target="_blank">https://doi.org/10.3189/172756404781814005</a>,
2004.
</mixed-citation></ref-html>
<ref-html id="bib1.bib133"><label>Raymo and Mitrovica(2012)</label><mixed-citation>
Raymo, M. E. and Mitrovica, J. X.: Collapse of polar ice sheets during the
stage 11 interglacial, Nature, 483, 453–456, <a href="https://doi.org/10.1038/nature10891" target="_blank">https://doi.org/10.1038/nature10891</a>,
2012.
</mixed-citation></ref-html>
<ref-html id="bib1.bib134"><label>Reddy et al.(2010)Reddy, Holland, and Arrigo</label><mixed-citation>
Reddy, T. E., Holland, D. M., and Arrigo, K. R.: Ross ice shelf cavity
circulation, residence time, and melting: Results from a model of oceanic
chlorofluorocarbons, Cont. Shelf Res., 30, 733–742,
<a href="https://doi.org/10.1016/j.csr.2010.01.007" target="_blank">https://doi.org/10.1016/j.csr.2010.01.007</a>, 2010.
</mixed-citation></ref-html>
<ref-html id="bib1.bib135"><label>Rignot et al.(2011)</label><mixed-citation>
Rignot, E. J., Velicogna, I., van den Broeke, M. R., Monaghan, A., and
Lenaerts, J. T. M.: Acceleration of the contribution of the Greenland and
Antarctic ice sheets to sea level rise, Geophys. Res. Lett., 38, 1–5,
<a href="https://doi.org/10.1029/2011GL046583" target="_blank">https://doi.org/10.1029/2011GL046583</a>, 2011.
</mixed-citation></ref-html>
<ref-html id="bib1.bib136"><label>Rignot et al.(2013)</label><mixed-citation>
Rignot, E. J., Jacobs, S. S., Mouginot, J., and Scheuchl, B.: Ice-Shelf
Melting Around Antarctica, Science, 341, 266–270,
<a href="https://doi.org/10.1126/science.1235798" target="_blank">https://doi.org/10.1126/science.1235798</a>, 2013.
</mixed-citation></ref-html>
<ref-html id="bib1.bib137"><label>Rignot et al.(2014)</label><mixed-citation>
Rignot, E. J., Mouginot, J., Morlighem, M., Seroussi, H., and Scheuchl, B.:
Widespread, rapid grounding line retreat of Pine Island, Thwaites, Smith,
and Kohler glaciers, West Antarctica, from 1992 to 2011, Geophys. Res.
Lett., 41, 3502–3509, <a href="https://doi.org/10.1002/2014GL060140" target="_blank">https://doi.org/10.1002/2014GL060140</a>, 2014.
</mixed-citation></ref-html>
<ref-html id="bib1.bib138"><label>Rintoul et al.(2010)</label><mixed-citation>
Rintoul, S. R., Speer, K., Sparrow, M., Meredith, M., Hofmann, E., Fahrbach,
E., Summerhayes, C., Worby, A., England, M., Bellerby, R. G., Speich, S.,
Costa, D., Hall, J., Hindell, M., Hosie, G., Stansfield, K., Fukamachi, Y.,
de Bruin, T., Naveira Garabato, A., Alverson, K., Ryabinin, V., Shin,
H. C., and Gladyshev, S.: Southern Ocean Observing System (SOOS): Rationale
and strategy for sustained observations of the Southern Ocean, in:
Proceedings of OceanObs'09: Sustained Ocean Observations and Information for
Society, edited by: Hall, J., Harrison, D. E., and Stammer, D., vol. 2, ESA
Publication WPP-306, Venice, <a href="https://doi.org/10.5270/OceanObs09.cwp.74" target="_blank">https://doi.org/10.5270/OceanObs09.cwp.74</a>, 2010.
</mixed-citation></ref-html>
<ref-html id="bib1.bib139"><label>Rintoul et al.(2016)</label><mixed-citation>
Rintoul, S. R., Silvano, A., Pena-Molino, B., van Wijk, E., Rosenberg, M.,
Greenbaum, J. S., and Blankenship, D. D.: Ocean heat drives rapid basal melt
of the Totten Ice Shelf, Sci. Adv., 2, 1–6, <a href="https://doi.org/10.1126/sciadv.1601610" target="_blank">https://doi.org/10.1126/sciadv.1601610</a>,
2016.
</mixed-citation></ref-html>
<ref-html id="bib1.bib140"><label>Robertson et al.(1998)</label><mixed-citation>
Robertson, R., Padman, L., and Egbert, G. D.: Tides in the Weddell Sea, in:
Ocean, Ice, and Atmosphere: Interactions at the Antarctic Continental Margin,
edited by: Jacobs, S. S. and Weiss, R. F., American Geophysical Union,
Washington, DC, 341–369, <a href="https://doi.org/10.1029/AR075p0341" target="_blank">https://doi.org/10.1029/AR075p0341</a>, 1998.
</mixed-citation></ref-html>
<ref-html id="bib1.bib141"><label>Rothrock(1975)</label><mixed-citation>
Rothrock, D. A.: The Energetics of the Plastic Deformation of Pack Ice by
Ridging, J. Geophys. Res., 80, 4514–4519, 1975.
</mixed-citation></ref-html>
<ref-html id="bib1.bib142"><label>Rott et al.(1996)Rott, Skvarca, and Nagler</label><mixed-citation>
Rott, H., Skvarca, P., and Nagler, T.: Rapid Collapse of Northern Larsen Ice
Shelf, Antarctica, Science, 271, 788–792,
<a href="https://doi.org/10.1126/science.271.5250.788" target="_blank">https://doi.org/10.1126/science.271.5250.788</a>, 1996.
</mixed-citation></ref-html>
<ref-html id="bib1.bib143"><label>Sallée et al.(2013)</label><mixed-citation>
Sallée, J.-B., Shuckburgh, E., Bruneau, N., Meijers, A. J. S.,
Bracegirdle, T. J., and Wang, Z.: Assessment of Southern Ocean mixed-layer
depths in CMIP5 models: Historical bias and forcing response, J. Geophys.
Res.-Oceans, 118, 1845–1862, <a href="https://doi.org/10.1002/jgrc.20157" target="_blank">https://doi.org/10.1002/jgrc.20157</a>, 2013.
</mixed-citation></ref-html>
<ref-html id="bib1.bib144"><label>Schmidtko et al.(2014)</label><mixed-citation>
Schmidtko, S., Heywood, K. J., Thompson, A. F., and Aoki, S.: Multidecadal
warming of Antarctic waters, Science, 346, 1227–1232,
<a href="https://doi.org/10.1126/science.1256117" target="_blank">https://doi.org/10.1126/science.1256117</a>, 2014.
</mixed-citation></ref-html>
<ref-html id="bib1.bib145"><label>Semtner(1976)</label><mixed-citation>
Semtner, A. J.: A Model for the Thermodynamic Growth of Sea Ice in Numerical
Investigations of Climate, J. Phys. Oceanogr., 6, 379–389,
<a href="https://doi.org/10.1175/1520-0485(1976)006&lt;0379:AMFTTG&gt;2.0.CO;2" target="_blank">https://doi.org/10.1175/1520-0485(1976)006&lt;0379:AMFTTG&gt;2.0.CO;2</a>, 1976.
</mixed-citation></ref-html>
<ref-html id="bib1.bib146"><label>Shchepetkin(2003)</label><mixed-citation>
Shchepetkin, A. F.: A method for computing horizontal pressure-gradient
force in an oceanic model with a nonaligned vertical coordinate, J. Geophys.
Res., 108, 3090, <a href="https://doi.org/10.1029/2001JC001047" target="_blank">https://doi.org/10.1029/2001JC001047</a>, 2003.
</mixed-citation></ref-html>
<ref-html id="bib1.bib147"><label>Shchepetkin and McWilliams(2005)</label><mixed-citation>
Shchepetkin, A. F. and McWilliams, J. C.: The regional oceanic modeling
system (ROMS): A split-explicit, free-surface,
topography-following-coordinate oceanic model, Ocean Model., 9, 347–404,
<a href="https://doi.org/10.1016/j.ocemod.2004.08.002" target="_blank">https://doi.org/10.1016/j.ocemod.2004.08.002</a>, 2005.
</mixed-citation></ref-html>
<ref-html id="bib1.bib148"><label>Shepherd et al.(2012)</label><mixed-citation>
Shepherd, A. P., Ivins, E. R., Geruo, A., Barletta, V. R., Bentley, M. J.,
Bettadpur, S., Briggs, K. H., Bromwich, D. H., Forsberg, R., Galin, N.,
Horwath, M., Jacobs, S. S., Joughin, I., King, M. A., Lenaerts, J. T. M., Li,
J., Ligtenberg, S. R. M., Luckman, A., Luthcke, S. B., McMillan, M., Meister,
R., Milne, G. A., Mouginot, J., Muir, A., Nicolas, J. P., Paden, J., Payne,
A. J., Pritchard, H. D., Rignot, E. J., Rott, H., Sandberg Sørensen, L.,
Scambos, T. A., Scheuchl, B., Schrama, E. J. O., Smith, B., Sundal, A. V.,
van Angelen, J. H., van de Berg, W. J., van den Broeke, M. R., Vaughan,
D. G., Velicogna, I., Wahr, J., Whitehouse, P. L., Wingham, D. J., Yi, D.,
Young, D. A., and Zwally, H. J.: A Reconciled Estimate of Ice-Sheet Mass
Balance, Science, 338, 1183–1189, <a href="https://doi.org/10.1126/science.1228102" target="_blank">https://doi.org/10.1126/science.1228102</a>, 2012.
</mixed-citation></ref-html>
<ref-html id="bib1.bib149"><label>Silvano et al.(2017)</label><mixed-citation>
Silvano, A., Rintoul, S. R., Pena-Molino, B., and Williams, G. D.:
Distribution of water masses and meltwater on the continental shelf near the
Totten and Moscow University ice shelves, J. Geophys. Res.-Oceans, 122,
2050–2068, <a href="https://doi.org/10.1002/2016JC012115" target="_blank">https://doi.org/10.1002/2016JC012115</a>, 2017.
</mixed-citation></ref-html>
<ref-html id="bib1.bib150"><label>Smagorinsky(1963)</label><mixed-citation>
Smagorinsky, J.: General Circulation Experiments With the Primitive
Equations I. The Basic Experiment, Mon. Weather Rev., 91, 99–164,
<a href="https://doi.org/10.1175/1520-0493(1963)091&lt;0099:GCEWTP&gt;2.3.CO;2" target="_blank">https://doi.org/10.1175/1520-0493(1963)091&lt;0099:GCEWTP&gt;2.3.CO;2</a>, 1963.
</mixed-citation></ref-html>
<ref-html id="bib1.bib151"><label>Smagorinsky(1993)</label><mixed-citation>
Smagorinsky, J.: Some historical remarks on the use of non-linear
viscosities, in: Large Eddy Simulation of Complex Engineering and
Geophysical Flows, edited by: Galperin, B. and Orszag, S., Cambridge
University Press, Cambridge, 3–36, 1993.
</mixed-citation></ref-html>
<ref-html id="bib1.bib152"><label>Smedsrud and Jenkins(2004)</label><mixed-citation>
Smedsrud, L. H. and Jenkins, A.: Frazil ice formation in an ice shelf water
plume, J. Geophys. Res., 109, C03025, <a href="https://doi.org/10.1029/2003JC001851" target="_blank">https://doi.org/10.1029/2003JC001851</a>, 2004.
</mixed-citation></ref-html>
<ref-html id="bib1.bib153"><label>Spence et al.(2014)</label><mixed-citation>
Spence, P., Griffies, S. M., England, M. H., Hogg, A. M., Saenko, O. A., and
Jourdain, N. C.: Rapid subsurface warming and circulation changes of
Antarctic coastal waters by poleward shifting winds, Geophys. Res. Lett.,
41, 4601–4610, <a href="https://doi.org/10.1002/2014GL060613" target="_blank">https://doi.org/10.1002/2014GL060613</a>, 2014.
</mixed-citation></ref-html>
<ref-html id="bib1.bib154"><label>Spence et al.(2017)</label><mixed-citation>
Spence, P., Holmes, R. M., Hogg, A. M., Griffies, S. M., Stewart, K. D., and
England, M. H.: Localized rapid warming of West Antarctic subsurface waters
by remote winds, Nat. Clim. Change, 7, 595–603, <a href="https://doi.org/10.1038/nclimate3335" target="_blank">https://doi.org/10.1038/nclimate3335</a>,
2017.
</mixed-citation></ref-html>
<ref-html id="bib1.bib155"><label>St-Laurent et al.(2013)</label><mixed-citation>
St-Laurent, P., Klinck, J. M., and Dinniman, M. S.: On the Role of Coastal
Troughs in the Circulation of Warm Circumpolar Deep Water on Antarctic
Shelves, J. Phys. Oceanogr., 43, 51–64, <a href="https://doi.org/10.1175/JPO-D-11-0237.1" target="_blank">https://doi.org/10.1175/JPO-D-11-0237.1</a>,
2013.
</mixed-citation></ref-html>
<ref-html id="bib1.bib156"><label>Stewart and Thompson(2015)</label><mixed-citation>
Stewart, A. L. and Thompson, A. F.: Eddy-mediated transport of warm
Circumpolar Deep Water across the Antarctic Shelf Break, Geophys. Res.
Lett., 42, 432–440, <a href="https://doi.org/10.1002/2014GL062281" target="_blank">https://doi.org/10.1002/2014GL062281</a>, 2015.
</mixed-citation></ref-html>
<ref-html id="bib1.bib157"><label>Talley(2007)</label><mixed-citation>
Talley, L. D.: Hydrographic Atlas of the World Ocean Circulation Experiment
(WOCE), Vol. 2, in: Pacific Ocean, edited by: Sparrow, M., Chapman, P., and
Gould, J., Tech. rep., International WOCE Project Office, Southampton, UK,
2007.
</mixed-citation></ref-html>
<ref-html id="bib1.bib158"><label>Tamura and Ohshima(2008)</label><mixed-citation>
Tamura, T. and Ohshima, K. I.: Mapping of sea ice production for Antarctic
coastal polynyas, J. Geophys. Res.-Oceans, 116, 1–5,
<a href="https://doi.org/10.1029/2010JC006586" target="_blank">https://doi.org/10.1029/2010JC006586</a>, 2008.
</mixed-citation></ref-html>
<ref-html id="bib1.bib159"><label>Tamura et al.(2016)</label><mixed-citation>
Tamura, T., Ohshima, K. I., Fraser, A. D., and Williams, G. D.: Sea ice
production variability in Antarctic coastal polynyas, J. Geophys.
Res.-Oceans, 121, 2967–2979, <a href="https://doi.org/10.1002/2015JC011537" target="_blank">https://doi.org/10.1002/2015JC011537</a>, 2016.
</mixed-citation></ref-html>
<ref-html id="bib1.bib160"><label>Thomas et al.(2001)</label><mixed-citation>
Thomas, M., Sündermann, J., and Maier-Reimer, E.: Consideration of
ocean tides in an OGCM and impacts on subseasonal to decadal polar motion
excitation, Geophys. Res. Lett., 28, 2457–2460, <a href="https://doi.org/10.1029/2000GL012234" target="_blank">https://doi.org/10.1029/2000GL012234</a>,
2001.
</mixed-citation></ref-html>
<ref-html id="bib1.bib161"><label>Timmermann and Beckmann(2004)</label><mixed-citation>
Timmermann, R. and Beckmann, A.: Parameterization of vertical mixing in the
Weddell Sea, Ocean Model., 6, 83–100, <a href="https://doi.org/10.1016/S1463-5003(02)00061-6" target="_blank">https://doi.org/10.1016/S1463-5003(02)00061-6</a>,
2004.
</mixed-citation></ref-html>
<ref-html id="bib1.bib162"><label>Timmermann and Goeller(2017)</label><mixed-citation>
Timmermann, R. and Goeller, S.: Response to Filchner–Ronne Ice Shelf cavity
warming in a coupled ocean–ice sheet model – Part 1: The ocean perspective,
Ocean Sci., 13, 765–776, https://doi.org/10.5194/os-13-765-2017, 2017.
</mixed-citation></ref-html>
<ref-html id="bib1.bib163"><label>Timmermann and Hellmer(2013)</label><mixed-citation>
Timmermann, R. and Hellmer, H. H.: Southern Ocean warming and increased ice
shelf basal melting in the twenty-first and twenty-second centuries based on
coupled ice-ocean finite-element modelling, Ocean Dynam., 63, 1011–1026,
<a href="https://doi.org/10.1007/s10236-013-0642-0" target="_blank">https://doi.org/10.1007/s10236-013-0642-0</a>, 2013.
</mixed-citation></ref-html>
<ref-html id="bib1.bib164"><label>Timmermann et al.(2009)</label><mixed-citation>
Timmermann, R., Danilov, S., Schröter, J., Böning, C., Sidorenko,
D., and Rollenhagen, K.: Ocean circulation and sea ice distribution in a
finite element global sea ice-ocean model, Ocean Model., 27, 114–129,
<a href="https://doi.org/10.1016/j.ocemod.2008.10.009" target="_blank">https://doi.org/10.1016/j.ocemod.2008.10.009</a>, 2009.
</mixed-citation></ref-html>
<ref-html id="bib1.bib165"><label>Timmermann et al.(2010)</label><mixed-citation>
Timmermann, R., Le Brocq, A., Deen, T., Domack, E., Dutrieux, P.,
Galton-Fenzi, B., Hellmer, H., Humbert, A., Jansen, D., Jenkins, A.,
Lambrecht, A., Makinson, K., Niederjasper, F., Nitsche, F., Nøst, O. A.,
Smedsrud, L. H., and Smith, W. H. F.: A consistent data set of Antarctic ice
sheet topography, cavity geometry, and global bathymetry, Earth Syst. Sci.
Data, 2, 261–273, https://doi.org/10.5194/essd-2-261-2010, 2010.
</mixed-citation></ref-html>
<ref-html id="bib1.bib166"><label>Timmermann et al.(2012)</label><mixed-citation>
Timmermann, R., Wang, Q., and Hellmer, H. H.: Ice-shelf basal melting in a
global finite-element sea-ice/ice-shelf/ocean model, Ann. Glaciol., 53,
303–314, <a href="https://doi.org/10.3189/2012AoG60A156" target="_blank">https://doi.org/10.3189/2012AoG60A156</a>, 2012.
</mixed-citation></ref-html>
<ref-html id="bib1.bib167"><label>Trenberth and Fasullo(2010)</label><mixed-citation>
Trenberth, K. E. and Fasullo, J. T.: Simulation of present-day and
twenty-first-century energy budgets of the southern oceans, J. Climate, 23,
440–454, <a href="https://doi.org/10.1175/2009JCLI3152.1" target="_blank">https://doi.org/10.1175/2009JCLI3152.1</a>, 2010.
</mixed-citation></ref-html>
<ref-html id="bib1.bib168"><label>Tsamados et al.(2013)</label><mixed-citation>
Tsamados, M., Feltham, D. L., and Wilchinsky, A. V.: Impact of a new
anisotropic rheology on simulations of Arctic Sea ice, J. Geophys.
Res.-Oceans, 118, 91–107, <a href="https://doi.org/10.1029/2012JC007990" target="_blank">https://doi.org/10.1029/2012JC007990</a>, 2013.
</mixed-citation></ref-html>
<ref-html id="bib1.bib169"><label>Turner et al.(2013a)</label><mixed-citation>
Turner, A. K., Hunke, E. C., and Bitz, C. M.: Two modes of sea-ice gravity
drainage: A parameterization for large-scale modeling, J. Geophys.
Res.-Oceans, 118, 2279–2294, <a href="https://doi.org/10.1002/jgrc.20171" target="_blank">https://doi.org/10.1002/jgrc.20171</a>, 2013a.
</mixed-citation></ref-html>
<ref-html id="bib1.bib170"><label>Turner et al.(2013b)</label><mixed-citation>
Turner, J., Bracegirdle, T. J., Phillips, T., Marshall, G. J., and Hosking,
J. S.: An initial assessment of Antarctic sea ice extent in the CMIP5
models, J. Climate, 26, 1473–1484, <a href="https://doi.org/10.1175/JCLI-D-12-00068.1" target="_blank">https://doi.org/10.1175/JCLI-D-12-00068.1</a>,
2013b.

</mixed-citation></ref-html>
<ref-html id="bib1.bib171"><label>Venables and Meredith(2014)</label><mixed-citation>
Venables, H. J. and Meredith, M. P.: Feedbacks between ice cover, ocean
stratification, and heat content in Ryder Bay, western Antarctic Peninsula,
J. Geophys. Res.-Oceans, 119, 5323–5336, <a href="https://doi.org/10.1002/2013JC009669" target="_blank">https://doi.org/10.1002/2013JC009669</a>, 2014.
</mixed-citation></ref-html>
<ref-html id="bib1.bib172"><label>Wåhlin et al.(2010)</label><mixed-citation>
Wåhlin, A. K., Yuan, X., Björk, G., and Nohr, C.: Inflow of Warm
Circumpolar Deep Water in the Central Amundsen Shelf, J. Phys. Oceanogr.,
40, 1427–1434, <a href="https://doi.org/10.1175/2010JPO4431.1" target="_blank">https://doi.org/10.1175/2010JPO4431.1</a>, 2010.
</mixed-citation></ref-html>
<ref-html id="bib1.bib173"><label>Wang et al.(2008)</label><mixed-citation>
Wang, Q., Danilov, S., and Schröter, J.: Finite element ocean
circulation model based on triangular prismatic elements, with application in
studying the effect of topography representation, J. Geophys. Res., 113,
C05015, <a href="https://doi.org/10.1029/2007JC004482" target="_blank">https://doi.org/10.1029/2007JC004482</a>, 2008.
</mixed-citation></ref-html>
<ref-html id="bib1.bib174"><label>Wang et al.(2013)</label><mixed-citation>
Wang, Q., Danilov, S., Hellmer, H. H., Sidorenko, D., Schröter, J., and
Jung, T.: Enhanced cross-shelf exchange by tides in the western Ross Sea,
Geophys. Res. Lett., 40, 5735–5739, <a href="https://doi.org/10.1002/2013GL058207" target="_blank">https://doi.org/10.1002/2013GL058207</a>, 2013.
</mixed-citation></ref-html>
<ref-html id="bib1.bib175"><label>Wang et al.(2014)</label><mixed-citation>
Wang, Q., Danilov, S., Sidorenko, D., Timmermann, R., Wekerle, C., Wang, X.,
Jung, T., and Schröter, J.: The Finite Element Sea Ice-Ocean Model
(FESOM) v.1.4: formulation of an ocean general circulation model, Geosci.
Model Dev., 7, 663–693, https://doi.org/10.5194/gmd-7-663-2014, 2014.
</mixed-citation></ref-html>
<ref-html id="bib1.bib176"><label>Wen et al.(2010)</label><mixed-citation>
Wen, J., Wang, Y., Wang, W., Jezek, K. C., Liu, H., and Allison, I.: Basal
melting and freezing under the Amery Ice Shelf, East Antarctica, J.
Glaciol., 56, 81–90, <a href="https://doi.org/10.3189/002214310791190820" target="_blank">https://doi.org/10.3189/002214310791190820</a>, 2010.
</mixed-citation></ref-html>
<ref-html id="bib1.bib177"><label>Williams et al.(2013)</label><mixed-citation>
Williams, K. D., Bodas-Salcedo, A., Déqué, M., Fermepin, S.,
Medeiros, B., Watanabe, M., Jakob, C., Klein, S. A., Senior, C. A., and
Williamson, D. L.: The transpose-AMIP II experiment and its application to
the understanding of southern ocean cloud biases in climate models, J.
Climate, 26, 3258–3274, <a href="https://doi.org/10.1175/JCLI-D-12-00429.1" target="_blank">https://doi.org/10.1175/JCLI-D-12-00429.1</a>, 2013.
</mixed-citation></ref-html>
<ref-html id="bib1.bib178"><label>Worby et al.(2008)</label><mixed-citation>
Worby, A. P., Geiger, C. A., Paget, M. J., van Woert, M. L., Ackley, S. F.,
and DeLiberty, T. L.: Thickness distribution of Antarctic sea ice, J.
Geophys. Res.-Oceans, 113, 1–14, <a href="https://doi.org/10.1029/2007JC004254" target="_blank">https://doi.org/10.1029/2007JC004254</a>, 2008.
</mixed-citation></ref-html>
<ref-html id="bib1.bib179"><label>Wunsch et al.(2009)</label><mixed-citation>
Wunsch, C., Heimbach, P., Ponte, R., and Fukumori, I.: The Global General
Circulation of the Ocean Estimated by the ECCO-Consortium, Oceanography, 22,
88–103, <a href="https://doi.org/10.5670/oceanog.2009.41" target="_blank">https://doi.org/10.5670/oceanog.2009.41</a>, 2009.
</mixed-citation></ref-html>
<ref-html id="bib1.bib180"><label>Zwally and Giovinetto(2011)</label><mixed-citation>
Zwally, H. J. and Giovinetto, M. B.: Overview and Assessment of Antarctic
Ice-Sheet Mass Balance Estimates: 1992-2009, Surv. Geophys., 32, 351–376,
<a href="https://doi.org/10.1007/s10712-011-9123-5" target="_blank">https://doi.org/10.1007/s10712-011-9123-5</a>, 2011.
</mixed-citation></ref-html>
<ref-html id="bib1.bib181"><label>Zweng et al.(2013)</label><mixed-citation>
Zweng, M. M., Reagan, J. R., Antonov, J. I., Mishonov, A. V., Boyer, T. P.,
Garcia, H. E., Baranova, O. K., Johnson, D. R., Seidov, D., and Bidlle,
M. M.: World Ocean Atlas 2013, vol. 2, Salinity, NOAA Atlas NESDIS, 2,
39 pp., 2013.
</mixed-citation></ref-html>--></article>
