<?xml version="1.0" encoding="UTF-8"?>
<!DOCTYPE article PUBLIC "-//NLM//DTD Journal Publishing with OASIS Tables v3.0 20080202//EN" "https://jats.nlm.nih.gov/nlm-dtd/publishing/3.0/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" article-type="research-article">
  <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-19-11041-2026</article-id><title-group><article-title>An antiplane shear model for evaluating shear-margin stability (Ortholine v1.0)</article-title><alt-title>Ortholine model</alt-title>
      </title-group>
      <contrib-group>
        <contrib contrib-type="author" corresp="yes" rid="aff1 aff2">
          <name><surname>Suckale</surname><given-names>Jenny</given-names></name>
          <email>jsuckale@stanford.edu</email>
        <ext-link>https://orcid.org/0000-0002-9787-8180</ext-link></contrib>
        <contrib contrib-type="author" corresp="no" rid="aff1">
          <name><surname>Elsworth</surname><given-names>Cooper W.</given-names></name>
          
        </contrib>
        <contrib contrib-type="author" corresp="no" rid="aff3">
          <name><surname>Akuli</surname><given-names>Neelanjan</given-names></name>
          
        </contrib>
        <aff id="aff1"><label>1</label><institution>Department of Geophysics, Stanford University, 397 Panama Mall, Stanford, CA 94305, USA</institution>
        </aff>
        <aff id="aff2"><label>2</label><institution>Institute for Computational and Mathematical Engineering, Stanford University, 397 Panama Mall, Stanford, CA 94305, USA</institution>
        </aff>
        <aff id="aff3"><label>3</label><institution>Department of Mechanical Engineering, Stanford University, 440 Escondido Mall, Bldg 530, Stanford, CA 94305, USA</institution>
        </aff>
      </contrib-group>
      <author-notes><corresp id="corr1">Jenny Suckale (jsuckale@stanford.edu)</corresp></author-notes><pub-date><day>19</day><month>November</month><year>2026</year></pub-date>
      
      <volume>19</volume>
      <issue>22</issue>
      <fpage>11041</fpage><lpage>11059</lpage>
      <history>
        <date date-type="received"><day>5</day><month>January</month><year>2026</year></date>
           <date date-type="rev-request"><day>23</day><month>February</month><year>2026</year></date>
           <date date-type="rev-recd"><day>17</day><month>August</month><year>2026</year></date>
           <date date-type="accepted"><day>19</day><month>August</month><year>2026</year></date>
      </history>
      <permissions>
        <copyright-statement>Copyright: © 2026 Jenny Suckale et al.</copyright-statement>
        <copyright-year>2026</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/19/11041/2026/gmd-19-11041-2026.html">This article is available from https://gmd.copernicus.org/articles/19/11041/2026/gmd-19-11041-2026.html</self-uri><self-uri xlink:href="https://gmd.copernicus.org/articles/19/11041/2026/gmd-19-11041-2026.pdf">The full text article is available as a PDF file from https://gmd.copernicus.org/articles/19/11041/2026/gmd-19-11041-2026.pdf</self-uri>
      <abstract><title>Abstract</title>

      <p id="d2e116">The lateral boundaries of ice streams and glaciers, commonly referred to as shear margins, play a key role in the mass balance of ice sheets. In glaciers, shear margins are often controlled by topography, whereas in ice streams they are inherently dynamic and can migrate. Margin migration alters the width of fast-moving ice and thus the force balance that controls ice speed and mass loss. Quantifying whether a shear margin is stable or prone to migration is hence important for estimating the future mass balance, but requires understanding the processes governing its current position. Numerous englacial and subglacial processes may contribute, including shear heating, fabric development, basal topography, lithology, and hydrology. Although the potential importance of these processes has been established, we currently lack models aimed at evaluating their relative importance at a specific field site and we aim to fill this gap. Here, we introduce “Ortholine”, an antiplane-shear free-boundary model developed to evaluate the relative importance of different englacial and subglacial processes in controlling a specific shear-margin location. Ortholine evaluates the relative importance of subglacial lithology, subglacial hydrology including channelized versus distributed drainage, basal topography and ice overburden, and englacial temperature through shear heating, while processes such as fabric development lie outside its current scope. By focusing on the cross-sectional force balance of ice streams, Ortholine complements commonly used flow-line models. After deriving and verifying Ortholine, we outline the four key steps for applying it to a field site, using Institute Ice Stream in the Weddell Sea Sector, Antarctica, as a proof-of-concept. We have selected Institute Ice Stream as an example, because of its intermediate complexity and the sufficient availability of field data to guide the model setup. Applying Ortholine at Institute, we find that spatial association alone is not a reliable indicator of the controlling process: the western shear margin is co-located with a topographic trough, yet our simulations indicate that a meltwater channel in that trough, rather than the topography itself, is the important control.</p>
  </abstract>
    
<funding-group>
<award-group id="gs1">
<funding-source>Directorate for Geosciences</funding-source>
<award-id>1341499</award-id>
<award-id>1744758</award-id>
<award-id>1739027</award-id>
</award-group>
</funding-group>
</article-meta>
  </front>
<body>
      

<sec id="Ch1.S1" sec-type="intro">
  <label>1</label><title>Introduction</title>
      <p id="d2e128">Ice streams and glaciers together account for the majority of mass loss from the Antarctic continent <xref ref-type="bibr" rid="bib1.bibx4" id="paren.1"/>. Their mass loss depends not only on the speed of the ice, but also on the width of the fast-flowing zone, which is set by the position of the shear margins <xref ref-type="bibr" rid="bib1.bibx53" id="paren.2"/>. Migration of a shear margin therefore changes the width of fast-moving ice and, with it, the discharge, so that shear-margin position directly affects the mass balance. Observations show that shear margins can move, either gradually <xref ref-type="bibr" rid="bib1.bibx25 bib1.bibx5 bib1.bibx11 bib1.bibx10" id="paren.3"/> or abruptly <xref ref-type="bibr" rid="bib1.bibx9" id="paren.4"/>. It is less clear how commonly shear margins move, partly because shear margins are difficult to pinpoint on satellite data due to heavy crevassing, and partly because migration rates on the order of meters per year are challenging to resolve. If shear margins were to move, the consequences for sea-level rise projections could be significant because the outward motion of a shear margin not only increases the width, but also the speed of fast-moving ice <xref ref-type="bibr" rid="bib1.bibx53" id="paren.5"/>.</p>
      <p id="d2e146">The puzzling variability of observed shear-margin behavior raises the question of what physical processes control shear-margin mobility and how these processes might be affected by evolving climatic conditions. Previous models have improved our understanding of how thermal <xref ref-type="bibr" rid="bib1.bibx62 bib1.bibx67 bib1.bibx72 bib1.bibx50 bib1.bibx26" id="paren.6"/> and hydrological <xref ref-type="bibr" rid="bib1.bibx49 bib1.bibx51 bib1.bibx13" id="paren.7"/> processes at the base of the ice contribute to the stability of shear margins. However, most of these models consider only a single shear margin assuming symmetric and idealized ice streams sliding over a homogeneous bed, inspired by the Siple Coast Ice Streams that are characterized by low topography, complex subglacial hydrology and widespread sediment deposits <xref ref-type="bibr" rid="bib1.bibx50 bib1.bibx72 bib1.bibx13" id="paren.8"><named-content content-type="pre">e.g.,</named-content></xref>.</p>
      <p id="d2e160">Figure <xref ref-type="fig" rid="F1"/> highlights the location of the Siple Coast Ice Streams superimposed on the shear-strain rate field of the Antarctic Ice Sheet as estimated from Landsat 8 satellite data by <xref ref-type="bibr" rid="bib1.bibx2" id="text.9"/>. It also shows several other locations where shear-margin properties are likely controlled by other physical processes than in the Siple Coast, ranging from topography at Lambert Glacier <xref ref-type="bibr" rid="bib1.bibx60" id="paren.10"/>, to spatial variations in subglacial conditions at Institute Ice Stream <xref ref-type="bibr" rid="bib1.bibx69" id="paren.11"/>, and possibly a combination of all of these at Thwaites Glacier. Thwaites Glacier is particularly relevant, because it is a case where one shear margin might be more prone to migration than the other <xref ref-type="bibr" rid="bib1.bibx42" id="paren.12"/> and because of its prominent contribution to current sea-level rise <xref ref-type="bibr" rid="bib1.bibx30" id="paren.13"><named-content content-type="pre">e.g.,</named-content></xref>.</p>

      <fig id="F1" specific-use="star"><label>Figure 1</label><caption><p id="d2e186">Map of the shear strain rate estimated from the satellite-derived velocity field of the Antarctic ice sheet based on Landsat 8 satellite data by <xref ref-type="bibr" rid="bib1.bibx2" id="text.14"/>. Superimposed are the locations where shear margins might be controlled by different physical processes.</p></caption>
        <graphic xlink:href="https://gmd.copernicus.org/articles/19/11041/2026/gmd-19-11041-2026-f01.png"/>

      </fig>

      <p id="d2e198">The goal of this study is to advance our ability to evaluate the relative importance of different englacial and subglacial processes in controlling a specific shear-margin location by developing an easily portable model, “Ortholine”. Methodologically, Ortholine attempts to strike a balance between the appealing simplicity of idealized shear-margin models that have contributed to our theoretical understanding <xref ref-type="bibr" rid="bib1.bibx33 bib1.bibx62 bib1.bibx72 bib1.bibx26 bib1.bibx49 bib1.bibx13 bib1.bibx27" id="paren.15"><named-content content-type="pre">e.g.,</named-content></xref> and the inevitable complexity of the field sites for which we aim to assess shear-margin stability. We build on the cross-sectional, free-boundary formulation of <xref ref-type="bibr" rid="bib1.bibx65 bib1.bibx64" id="text.16"/>, which allows us to determine the slipping regions at the ice-bed interface self-consistently from an integrated force balance.</p>
      <p id="d2e209">Our model design is inspired by the scientific successes of simple flow-line models pioneered by <xref ref-type="bibr" rid="bib1.bibx47" id="text.17"/> and widely adopted by the scientific community in the 1980s <xref ref-type="bibr" rid="bib1.bibx31 bib1.bibx44 bib1.bibx54" id="paren.18"><named-content content-type="pre">e.g.,</named-content></xref>. Flow-line models neglect shear margins and thereby reduce a three-dimensional problem to plane flow along the main flow direction. A complementary point of view that allows for an explicit focus on shear margins is to assume antiplane shear and instead neglect longitudinal stresses <xref ref-type="bibr" rid="bib1.bibx33" id="paren.19"><named-content content-type="pre">e.g.,</named-content></xref>. By antiplane shear we mean the configuration in which the ice velocity is directed along the flow direction <inline-formula><mml:math id="M1" display="inline"><mml:mi>x</mml:mi></mml:math></inline-formula>, out of the cross-sectional plane, while its gradients are resolved within that cross-section; this is the standard usage in continuum mechanics and is distinct from the plane-flow assumption of flow-line models. The name of our model, “Ortholine”, is motivated partly by the orthogonal orientation of our model domain to the main flow line and partly by its conceptual orthogonality to flow-line models. Ortholine allows researchers to focus on an aspect of ice dynamics, the shear margin, that is neglected in flow-line models.</p>
      <p id="d2e232">Previous antiplane shear models of shear margins simplify the domain as symmetric and assume a flat basal topography <xref ref-type="bibr" rid="bib1.bibx33 bib1.bibx62 bib1.bibx72 bib1.bibx49 bib1.bibx26 bib1.bibx13" id="paren.20"/>. Instead, we implement a body-fitted mesh using the Gmsh package <xref ref-type="bibr" rid="bib1.bibx19" id="paren.21"/> that can capture variable subglacial topography in the model domain. We also allow for heterogeneous basal conditions representative of perfectly plastic sediment behavior <xref ref-type="bibr" rid="bib1.bibx76 bib1.bibx36" id="paren.22"/>, velocity-strengthening hard rock sliding laws <xref ref-type="bibr" rid="bib1.bibx77" id="paren.23"/>, and regularized Coulomb sliding over hard rock <xref ref-type="bibr" rid="bib1.bibx63" id="paren.24"/> while maximizing computational efficiency. Including these heterogeneities implies that the antiplane shear plane might not be identical to the cross-section of the ice stream. We have developed an accompanying algorithm to compute the appropriate model plane from surface velocities.</p>
      <p id="d2e250">Although Ortholine is intended as a general framework that can be easily adapted to any given field site, we detail the steps to apply it for the case of Institute Ice Stream in the Weddell Sea Sector, Antarctica (see Fig. <xref ref-type="fig" rid="F1"/>). Institute Ice Stream represents a field setting of moderate complexity that shares similarities with the Siple Coast, such as low overall driving stresses and a widespread unconsolidated sediment layer <xref ref-type="bibr" rid="bib1.bibx6" id="paren.25"/>. Radio-echo sounding surveys map out a mildly varying topography and suggest the existence of a distinct transition between hardrock and saturated sediments beneath the ice stream <xref ref-type="bibr" rid="bib1.bibx69" id="paren.26"/>. From a dynamic point of view, Institute Ice Stream has attracted attention because it is located on a reverse-sloping bed with few pinning points <xref ref-type="bibr" rid="bib1.bibx6 bib1.bibx57" id="paren.27"/>, making it prone to the marine ice-sheet instability <xref ref-type="bibr" rid="bib1.bibx78 bib1.bibx74 bib1.bibx66 bib1.bibx34" id="paren.28"/>.</p>
</sec>
<sec id="Ch1.S2">
  <label>2</label><title>Methods</title>
      <p id="d2e275">We develop a thermomechanical model of an ice stream cross section oriented orthogonal to the flow line. Instead of prescribing the margin positions, we solve a free-boundary problem for the slipping portion of the bed based on an integrated force balance across the domain, following the derivation by <xref ref-type="bibr" rid="bib1.bibx65" id="text.29"/>. We formulate the governing equations of ice flow for three different subglacial boundary conditions representative of failing sediments, locked sediments, and hard-bed sliding. We then convert the free-boundary problem to a convex optimization problem and solve it through the Disciplined Convex Programming framework <xref ref-type="bibr" rid="bib1.bibx21" id="paren.30"/>.</p>
<sec id="Ch1.S2.SS1">
  <label>2.1</label><title>Governing Equations</title>
      <p id="d2e291">Our free-boundary model extends the mechanical model by <xref ref-type="bibr" rid="bib1.bibx65" id="text.31"/> to include thermomechanical effects. To reduce the complexity of the 3-D geometry, we neglect variability in the downstream direction (<inline-formula><mml:math id="M2" display="inline"><mml:mrow><mml:mi>d</mml:mi><mml:mo>/</mml:mo><mml:mi mathvariant="normal">d</mml:mi><mml:mi>x</mml:mi><mml:mo>≈</mml:mo><mml:mn mathvariant="normal">0</mml:mn></mml:mrow></mml:math></inline-formula>), with the ice moving in the positive <inline-formula><mml:math id="M3" display="inline"><mml:mi>x</mml:mi></mml:math></inline-formula> direction, shown in the right-hand panel of Fig. <xref ref-type="fig" rid="F2"/>. Neglecting downstream variability reduces the momentum balance to

            <disp-formula id="Ch1.E1" content-type="numbered"><label>1</label><mml:math id="M4" display="block"><mml:mrow><mml:mo>-</mml:mo><mml:mi mathvariant="normal">∇</mml:mi><mml:mo>⋅</mml:mo><mml:mfenced open="(" close=")"><mml:mrow><mml:mi mathvariant="italic">η</mml:mi><mml:mo>(</mml:mo><mml:mi>T</mml:mi><mml:mo>)</mml:mo><mml:mo>|</mml:mo><mml:mi mathvariant="normal">∇</mml:mi><mml:mi>u</mml:mi><mml:msup><mml:mo>|</mml:mo><mml:mrow><mml:mn mathvariant="normal">1</mml:mn><mml:mo>/</mml:mo><mml:mi>n</mml:mi><mml:mo>-</mml:mo><mml:mn mathvariant="normal">1</mml:mn></mml:mrow></mml:msup><mml:mi mathvariant="normal">∇</mml:mi><mml:mi>u</mml:mi></mml:mrow></mml:mfenced><mml:mo>=</mml:mo><mml:mi mathvariant="italic">ρ</mml:mi><mml:mi>g</mml:mi><mml:mi>sin⁡</mml:mi><mml:mi mathvariant="italic">α</mml:mi><mml:mspace width="0.125em" linebreak="nobreak"/><mml:mspace linebreak="nobreak" width="0.125em"/><mml:mspace width="0.125em" linebreak="nobreak"/><mml:mspace linebreak="nobreak" width="0.125em"/><mml:mtext>on</mml:mtext><mml:mspace width="0.125em" linebreak="nobreak"/><mml:mspace linebreak="nobreak" width="0.125em"/><mml:mspace linebreak="nobreak" width="0.125em"/><mml:mspace width="0.125em" linebreak="nobreak"/><mml:mi mathvariant="normal">Ω</mml:mi><mml:mo>,</mml:mo></mml:mrow></mml:math></disp-formula>

          with antiplane speed, <inline-formula><mml:math id="M5" display="inline"><mml:mi>u</mml:mi></mml:math></inline-formula>, temperature-dependent power-law coefficient, <inline-formula><mml:math id="M6" display="inline"><mml:mrow><mml:mi mathvariant="italic">η</mml:mi><mml:mo>(</mml:mo><mml:mi>T</mml:mi><mml:mo>)</mml:mo></mml:mrow></mml:math></inline-formula>, power-law exponent, <inline-formula><mml:math id="M7" display="inline"><mml:mi>n</mml:mi></mml:math></inline-formula>, ice density, <inline-formula><mml:math id="M8" display="inline"><mml:mi mathvariant="italic">ρ</mml:mi></mml:math></inline-formula>, gravitational acceleration, <inline-formula><mml:math id="M9" display="inline"><mml:mi>g</mml:mi></mml:math></inline-formula>, and surface slope, <inline-formula><mml:math id="M10" display="inline"><mml:mi mathvariant="italic">α</mml:mi></mml:math></inline-formula>. A typical non-Newtonian Glen's rheology would entail <inline-formula><mml:math id="M11" display="inline"><mml:mrow><mml:mi>n</mml:mi><mml:mo>=</mml:mo><mml:mn mathvariant="normal">3</mml:mn></mml:mrow></mml:math></inline-formula> <xref ref-type="bibr" rid="bib1.bibx20" id="paren.32"/>. We define <inline-formula><mml:math id="M12" display="inline"><mml:mrow><mml:mi mathvariant="italic">η</mml:mi><mml:mo>(</mml:mo><mml:mi>T</mml:mi><mml:mo>)</mml:mo><mml:mo>=</mml:mo><mml:mo>[</mml:mo><mml:mn mathvariant="normal">2</mml:mn><mml:mi>E</mml:mi><mml:mi>A</mml:mi><mml:mo>(</mml:mo><mml:mi>T</mml:mi><mml:mo>)</mml:mo><mml:msup><mml:mo>]</mml:mo><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">1</mml:mn><mml:mo>/</mml:mo><mml:mi>n</mml:mi></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula>, where <inline-formula><mml:math id="M13" display="inline"><mml:mrow><mml:mi>A</mml:mi><mml:mo>(</mml:mo><mml:mi>T</mml:mi><mml:mo>)</mml:mo></mml:mrow></mml:math></inline-formula> is the Glen rate factor and <inline-formula><mml:math id="M14" display="inline"><mml:mrow><mml:mi>E</mml:mi><mml:mo>=</mml:mo><mml:mn mathvariant="normal">4</mml:mn></mml:mrow></mml:math></inline-formula> is the enhancement factor. For <inline-formula><mml:math id="M15" display="inline"><mml:mrow><mml:mi>n</mml:mi><mml:mo>=</mml:mo><mml:mn mathvariant="normal">3</mml:mn></mml:mrow></mml:math></inline-formula>, <inline-formula><mml:math id="M16" display="inline"><mml:mrow><mml:mi mathvariant="italic">η</mml:mi><mml:mo>(</mml:mo><mml:mi>T</mml:mi><mml:mo>)</mml:mo><mml:mo>=</mml:mo><mml:mstyle displaystyle="false"><mml:mfrac style="text"><mml:mn mathvariant="normal">1</mml:mn><mml:mn mathvariant="normal">2</mml:mn></mml:mfrac></mml:mstyle><mml:mi>A</mml:mi><mml:mo>(</mml:mo><mml:mi>T</mml:mi><mml:msup><mml:mo>)</mml:mo><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">1</mml:mn><mml:mo>/</mml:mo><mml:mn mathvariant="normal">3</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula>.</p>

      <fig id="F2" specific-use="star"><label>Figure 2</label><caption><p id="d2e585">Model domain of previous models as compared to Ortholine. Both model approaches neglect variability in the along-flow direction, <inline-formula><mml:math id="M17" display="inline"><mml:mi>x</mml:mi></mml:math></inline-formula>. Advancing on previous idealized models (left), Ortholine captures the entire width of the zone of fast-moving ice including both shear margins and allows for spatially varying lithology, subglacial hydrology and basal topography (right).</p></caption>
          <graphic xlink:href="https://gmd.copernicus.org/articles/19/11041/2026/gmd-19-11041-2026-f02.png"/>

        </fig>

      <p id="d2e601">We neglect downstream variation in ice thickness and impose a stress-free surface,

            <disp-formula id="Ch1.E2" content-type="numbered"><label>2</label><mml:math id="M18" display="block"><mml:mrow><mml:mo>-</mml:mo><mml:mi mathvariant="italic">η</mml:mi><mml:mo>(</mml:mo><mml:mi>T</mml:mi><mml:mo>)</mml:mo><mml:mo>|</mml:mo><mml:mi mathvariant="normal">∇</mml:mi><mml:mi>u</mml:mi><mml:msup><mml:mo>|</mml:mo><mml:mrow><mml:mn mathvariant="normal">1</mml:mn><mml:mo>/</mml:mo><mml:mi>n</mml:mi><mml:mo>-</mml:mo><mml:mn mathvariant="normal">1</mml:mn></mml:mrow></mml:msup><mml:msub><mml:mi>u</mml:mi><mml:mi>n</mml:mi></mml:msub><mml:mo>=</mml:mo><mml:mn mathvariant="normal">0</mml:mn><mml:mspace width="0.125em" linebreak="nobreak"/><mml:mspace linebreak="nobreak" width="0.125em"/><mml:mspace linebreak="nobreak" width="0.125em"/><mml:mspace width="0.125em" linebreak="nobreak"/><mml:mtext>on</mml:mtext><mml:mspace linebreak="nobreak" width="0.125em"/><mml:mspace linebreak="nobreak" width="0.125em"/><mml:mspace width="0.125em" linebreak="nobreak"/><mml:mspace linebreak="nobreak" width="0.125em"/><mml:msub><mml:mi mathvariant="normal">Γ</mml:mi><mml:mtext>surf</mml:mtext></mml:msub><mml:mo>,</mml:mo></mml:mrow></mml:math></disp-formula>

          where <inline-formula><mml:math id="M19" display="inline"><mml:mrow><mml:msub><mml:mi>u</mml:mi><mml:mi>n</mml:mi></mml:msub><mml:mo>=</mml:mo><mml:mo>∂</mml:mo><mml:mi>u</mml:mi><mml:mo>/</mml:mo><mml:mo>∂</mml:mo><mml:mi>n</mml:mi></mml:mrow></mml:math></inline-formula> where <inline-formula><mml:math id="M20" display="inline"><mml:mi mathvariant="bold-italic">n</mml:mi></mml:math></inline-formula> is the outward unit normal to the interface. Additionally, the domain has stress-free lateral bounds,

            <disp-formula id="Ch1.E3" content-type="numbered"><label>3</label><mml:math id="M21" display="block"><mml:mrow><mml:mo>-</mml:mo><mml:mi mathvariant="italic">η</mml:mi><mml:mo>(</mml:mo><mml:mi>T</mml:mi><mml:mo>)</mml:mo><mml:mo>|</mml:mo><mml:mi mathvariant="normal">∇</mml:mi><mml:mi>u</mml:mi><mml:msup><mml:mo>|</mml:mo><mml:mrow><mml:mn mathvariant="normal">1</mml:mn><mml:mo>/</mml:mo><mml:mi>n</mml:mi><mml:mo>-</mml:mo><mml:mn mathvariant="normal">1</mml:mn></mml:mrow></mml:msup><mml:msub><mml:mi>u</mml:mi><mml:mi>n</mml:mi></mml:msub><mml:mo>=</mml:mo><mml:mn mathvariant="normal">0</mml:mn><mml:mspace linebreak="nobreak" width="0.125em"/><mml:mspace width="0.125em" linebreak="nobreak"/><mml:mspace linebreak="nobreak" width="0.125em"/><mml:mspace width="0.125em" linebreak="nobreak"/><mml:mtext>on</mml:mtext><mml:mspace linebreak="nobreak" width="0.125em"/><mml:mspace linebreak="nobreak" width="0.125em"/><mml:mspace width="0.125em" linebreak="nobreak"/><mml:mspace width="0.125em" linebreak="nobreak"/><mml:msub><mml:mi mathvariant="normal">Γ</mml:mi><mml:mtext>lat</mml:mtext></mml:msub><mml:mo>.</mml:mo></mml:mrow></mml:math></disp-formula>

          The basal traction at the bed,

            <disp-formula id="Ch1.E4" content-type="numbered"><label>4</label><mml:math id="M22" display="block"><mml:mrow><mml:mo>-</mml:mo><mml:mi mathvariant="italic">η</mml:mi><mml:mo>(</mml:mo><mml:mi>T</mml:mi><mml:mo>)</mml:mo><mml:mo>|</mml:mo><mml:mi mathvariant="normal">∇</mml:mi><mml:mi>u</mml:mi><mml:msup><mml:mo>|</mml:mo><mml:mrow><mml:mn mathvariant="normal">1</mml:mn><mml:mo>/</mml:mo><mml:mi>n</mml:mi><mml:mo>-</mml:mo><mml:mn mathvariant="normal">1</mml:mn></mml:mrow></mml:msup><mml:msub><mml:mi>u</mml:mi><mml:mi>n</mml:mi></mml:msub><mml:mo>=</mml:mo><mml:msub><mml:mi mathvariant="italic">τ</mml:mi><mml:mi mathvariant="normal">b</mml:mi></mml:msub><mml:mo>(</mml:mo><mml:mi>x</mml:mi><mml:mo>,</mml:mo><mml:mi>y</mml:mi><mml:mo>,</mml:mo><mml:mi>u</mml:mi><mml:mo>)</mml:mo><mml:mspace linebreak="nobreak" width="0.125em"/><mml:mspace width="0.125em" linebreak="nobreak"/><mml:mspace width="0.125em" linebreak="nobreak"/><mml:mspace width="0.125em" linebreak="nobreak"/><mml:mtext>on</mml:mtext><mml:mspace linebreak="nobreak" width="0.125em"/><mml:mspace linebreak="nobreak" width="0.125em"/><mml:mspace linebreak="nobreak" width="0.125em"/><mml:mspace width="0.125em" linebreak="nobreak"/><mml:msub><mml:mi mathvariant="normal">Γ</mml:mi><mml:mtext>bed</mml:mtext></mml:msub><mml:mo>,</mml:mo></mml:mrow></mml:math></disp-formula>

          takes one of two forms depending on whether the bed at a given location consists of saturated sediment or hard rock. We prescribe the partition of the bed into sediment and hard-rock regions from observations. The location of the rock–sediment transition and the degree to which this transition implies a significant difference in basal properties need to be assessed on a case-by-case basis. For example, we rely on the geophysical imaging by <xref ref-type="bibr" rid="bib1.bibx69" id="text.33"/> to assign boundary conditions for Institute Ice Stream, Antarctica, as detailed below. The sediment boundary condition captures regions of water-saturated sediments that abide by Coulomb plasticity as observed experimentally <xref ref-type="bibr" rid="bib1.bibx76" id="paren.34"/> with

                <disp-formula specific-use="align" content-type="numbered"><mml:math id="M23" display="block"><mml:mtable displaystyle="true"><mml:mlabeledtr id="Ch1.E5"><mml:mtd><mml:mtext>5</mml:mtext></mml:mtd><mml:mtd><mml:mrow><mml:mstyle class="stylechange" displaystyle="true"/><mml:msub><mml:mi mathvariant="italic">τ</mml:mi><mml:mi mathvariant="normal">b</mml:mi></mml:msub><mml:mo>=</mml:mo><mml:msub><mml:mi>C</mml:mi><mml:mi mathvariant="normal">f</mml:mi></mml:msub><mml:mi>N</mml:mi><mml:mo>+</mml:mo><mml:mi>c</mml:mi><mml:mo>,</mml:mo><mml:mspace width="2em" linebreak="nobreak"/><mml:mtext>for </mml:mtext><mml:mi>u</mml:mi><mml:mo>&gt;</mml:mo><mml:mn mathvariant="normal">0</mml:mn><mml:mspace linebreak="nobreak" width="1em"/><mml:mtext>(failing)</mml:mtext><mml:mo>.</mml:mo></mml:mrow></mml:mtd></mml:mlabeledtr><mml:mlabeledtr id="Ch1.E6"><mml:mtd><mml:mtext>6</mml:mtext></mml:mtd><mml:mtd><mml:mrow><mml:mstyle class="stylechange" displaystyle="true"/><mml:msub><mml:mi mathvariant="italic">τ</mml:mi><mml:mi mathvariant="normal">b</mml:mi></mml:msub><mml:mo>≤</mml:mo><mml:msub><mml:mi>C</mml:mi><mml:mi mathvariant="normal">f</mml:mi></mml:msub><mml:mi>N</mml:mi><mml:mo>+</mml:mo><mml:mi>c</mml:mi><mml:mo>,</mml:mo><mml:mspace width="2em" linebreak="nobreak"/><mml:mtext>for </mml:mtext><mml:mi>u</mml:mi><mml:mo>=</mml:mo><mml:mn mathvariant="normal">0</mml:mn><mml:mspace width="1em" linebreak="nobreak"/><mml:mtext>(locked)</mml:mtext></mml:mrow></mml:mtd></mml:mlabeledtr></mml:mtable></mml:math></disp-formula>

          dependent on effective pressure, <inline-formula><mml:math id="M24" display="inline"><mml:mrow><mml:mi>N</mml:mi><mml:mo>=</mml:mo><mml:msub><mml:mi mathvariant="italic">σ</mml:mi><mml:mi>n</mml:mi></mml:msub><mml:mo>-</mml:mo><mml:mi>p</mml:mi></mml:mrow></mml:math></inline-formula>, where <inline-formula><mml:math id="M25" display="inline"><mml:mrow><mml:msub><mml:mi mathvariant="italic">σ</mml:mi><mml:mi>n</mml:mi></mml:msub></mml:mrow></mml:math></inline-formula> is the overburden, <inline-formula><mml:math id="M26" display="inline"><mml:mi>p</mml:mi></mml:math></inline-formula> the pore pressure, <inline-formula><mml:math id="M27" display="inline"><mml:mrow><mml:msub><mml:mi>C</mml:mi><mml:mi mathvariant="normal">f</mml:mi></mml:msub></mml:mrow></mml:math></inline-formula> the friction coefficient and <inline-formula><mml:math id="M28" display="inline"><mml:mi>c</mml:mi></mml:math></inline-formula> the cohesion of the sediment.</p>
      <p id="d2e995">For the portion of the subglacial bed that consists of hard rock, we assume that the basal traction follows a sliding law,

            <disp-formula id="Ch1.E7" content-type="numbered"><label>7</label><mml:math id="M29" display="block"><mml:mrow><mml:msub><mml:mi mathvariant="italic">τ</mml:mi><mml:mi mathvariant="normal">b</mml:mi></mml:msub><mml:mo>=</mml:mo><mml:mo>(</mml:mo><mml:mi>m</mml:mi><mml:mo>+</mml:mo><mml:mn mathvariant="normal">1</mml:mn><mml:mo>)</mml:mo><mml:msup><mml:mi mathvariant="italic">β</mml:mi><mml:mn mathvariant="normal">2</mml:mn></mml:msup><mml:mo>|</mml:mo><mml:mi>u</mml:mi><mml:msup><mml:mo>|</mml:mo><mml:mi>m</mml:mi></mml:msup></mml:mrow></mml:math></disp-formula>

          dependent on the downstream ice speed, <inline-formula><mml:math id="M30" display="inline"><mml:mi>u</mml:mi></mml:math></inline-formula>, to the power <inline-formula><mml:math id="M31" display="inline"><mml:mi>m</mml:mi></mml:math></inline-formula>. For the velocity-strengthening hard-rock cases below, we use <inline-formula><mml:math id="M32" display="inline"><mml:mrow><mml:mi>m</mml:mi><mml:mo>=</mml:mo><mml:mn mathvariant="normal">3</mml:mn><mml:mo>/</mml:mo><mml:mn mathvariant="normal">2</mml:mn></mml:mrow></mml:math></inline-formula>; the <inline-formula><mml:math id="M33" display="inline"><mml:mrow><mml:mi>m</mml:mi><mml:mo>=</mml:mo><mml:mn mathvariant="normal">0</mml:mn></mml:mrow></mml:math></inline-formula> limit is retained for the Coulomb-sliding comparisons. The coefficient <inline-formula><mml:math id="M34" display="inline"><mml:mi mathvariant="italic">β</mml:mi></mml:math></inline-formula> is defined through the hard-rock basal potential <inline-formula><mml:math id="M35" display="inline"><mml:mrow><mml:msup><mml:mi mathvariant="italic">β</mml:mi><mml:mn mathvariant="normal">2</mml:mn></mml:msup><mml:mo>|</mml:mo><mml:mi>u</mml:mi><mml:msup><mml:mo>|</mml:mo><mml:mrow><mml:mi>m</mml:mi><mml:mo>+</mml:mo><mml:mn mathvariant="normal">1</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula>, whose derivative gives Eq. (7). This sliding law can represent three different regimes: velocity-strengthening behavior, <inline-formula><mml:math id="M36" display="inline"><mml:mrow><mml:mi>m</mml:mi><mml:mo>&gt;</mml:mo><mml:mn mathvariant="normal">0</mml:mn></mml:mrow></mml:math></inline-formula>, as might arise when sliding over bumps without cavitation <xref ref-type="bibr" rid="bib1.bibx77" id="paren.35"><named-content content-type="pre">e.g.,</named-content></xref>; velocity-weakening behavior, <inline-formula><mml:math id="M37" display="inline"><mml:mrow><mml:mi>m</mml:mi><mml:mo>&lt;</mml:mo><mml:mn mathvariant="normal">0</mml:mn></mml:mrow></mml:math></inline-formula>, potentially indicative of bed bumps with cavitation <xref ref-type="bibr" rid="bib1.bibx41" id="paren.36"><named-content content-type="pre">e.g.,</named-content></xref>; and regularized Coulomb sliding over hard rock, <inline-formula><mml:math id="M38" display="inline"><mml:mrow><mml:mi>m</mml:mi><mml:mo>=</mml:mo><mml:mn mathvariant="normal">0</mml:mn></mml:mrow></mml:math></inline-formula> <xref ref-type="bibr" rid="bib1.bibx63" id="paren.37"><named-content content-type="pre">e.g.,</named-content></xref>. Of these three possibilities, velocity-weakening behavior <xref ref-type="bibr" rid="bib1.bibx41 bib1.bibx79" id="paren.38"/> with <inline-formula><mml:math id="M39" display="inline"><mml:mrow><mml:mi>m</mml:mi><mml:mo>&lt;</mml:mo><mml:mn mathvariant="normal">0</mml:mn></mml:mrow></mml:math></inline-formula> is unstable when applied to the entire hard-rock bed or most of it. We therefore do not consider this limit in our simulations.</p>
      <p id="d2e1177">We assume that the temperature distribution in the ice column is driven by diffusion of shear heating,

            <disp-formula id="Ch1.E8" content-type="numbered"><label>8</label><mml:math id="M40" display="block"><mml:mrow><mml:mo>-</mml:mo><mml:mi mathvariant="normal">∇</mml:mi><mml:mo>⋅</mml:mo><mml:mo>(</mml:mo><mml:mi>k</mml:mi><mml:mi mathvariant="normal">∇</mml:mi><mml:mi>T</mml:mi><mml:mo>)</mml:mo><mml:mo>=</mml:mo><mml:mn mathvariant="normal">2</mml:mn><mml:msub><mml:mi mathvariant="italic">τ</mml:mi><mml:mi mathvariant="normal">E</mml:mi></mml:msub><mml:mo>(</mml:mo><mml:mi>u</mml:mi><mml:mo>)</mml:mo><mml:msub><mml:mover accent="true"><mml:mi mathvariant="italic">ϵ</mml:mi><mml:mo mathvariant="normal">˙</mml:mo></mml:mover><mml:mi mathvariant="normal">E</mml:mi></mml:msub><mml:mo>(</mml:mo><mml:mi>u</mml:mi><mml:mo>)</mml:mo><mml:mo>,</mml:mo><mml:mspace width="0.125em" linebreak="nobreak"/><mml:mspace width="0.125em" linebreak="nobreak"/><mml:mspace width="0.125em" linebreak="nobreak"/><mml:mspace width="0.125em" linebreak="nobreak"/><mml:mi>T</mml:mi><mml:mo>&lt;</mml:mo><mml:msub><mml:mi>T</mml:mi><mml:mtext>melt</mml:mtext></mml:msub><mml:mo>,</mml:mo></mml:mrow></mml:math></disp-formula>

          with thermal conductivity, <inline-formula><mml:math id="M41" display="inline"><mml:mi>k</mml:mi></mml:math></inline-formula>, effective stress, <inline-formula><mml:math id="M42" display="inline"><mml:mrow><mml:msub><mml:mi mathvariant="italic">τ</mml:mi><mml:mi mathvariant="normal">E</mml:mi></mml:msub><mml:mo>(</mml:mo><mml:mi>u</mml:mi><mml:mo>)</mml:mo></mml:mrow></mml:math></inline-formula>, effective strain rate, <inline-formula><mml:math id="M43" display="inline"><mml:mrow><mml:msub><mml:mover accent="true"><mml:mi mathvariant="italic">ϵ</mml:mi><mml:mo mathvariant="normal">˙</mml:mo></mml:mover><mml:mi mathvariant="normal">E</mml:mi></mml:msub><mml:mo>(</mml:mo><mml:mi>u</mml:mi><mml:mo>)</mml:mo></mml:mrow></mml:math></inline-formula>, and pressure melting point, <inline-formula><mml:math id="M44" display="inline"><mml:mrow><mml:msub><mml:mi>T</mml:mi><mml:mtext>melt</mml:mtext></mml:msub></mml:mrow></mml:math></inline-formula>. The effective strain rate is the second invariant of the strain-rate tensor, which for the antiplane field reduces to <inline-formula><mml:math id="M45" display="inline"><mml:mrow><mml:msub><mml:mover accent="true"><mml:mi mathvariant="italic">ϵ</mml:mi><mml:mo mathvariant="normal">˙</mml:mo></mml:mover><mml:mi mathvariant="normal">E</mml:mi></mml:msub><mml:mo>=</mml:mo><mml:mo>(</mml:mo><mml:mn mathvariant="normal">1</mml:mn><mml:mo>/</mml:mo><mml:mn mathvariant="normal">2</mml:mn><mml:mo>)</mml:mo><mml:mo>|</mml:mo><mml:mi mathvariant="normal">∇</mml:mi><mml:mi>u</mml:mi><mml:mo>|</mml:mo></mml:mrow></mml:math></inline-formula>, and the effective stress is the corresponding viscous stress, <inline-formula><mml:math id="M46" display="inline"><mml:mrow><mml:msub><mml:mi mathvariant="italic">τ</mml:mi><mml:mi mathvariant="normal">E</mml:mi></mml:msub><mml:mo>=</mml:mo><mml:msup><mml:mn mathvariant="normal">2</mml:mn><mml:mrow><mml:mn mathvariant="normal">1</mml:mn><mml:mo>/</mml:mo><mml:mi>n</mml:mi></mml:mrow></mml:msup><mml:mi mathvariant="italic">η</mml:mi><mml:mo>(</mml:mo><mml:mi>T</mml:mi><mml:mo>)</mml:mo><mml:msubsup><mml:mover accent="true"><mml:mi mathvariant="italic">ϵ</mml:mi><mml:mo mathvariant="normal">˙</mml:mo></mml:mover><mml:mi mathvariant="normal">E</mml:mi><mml:mrow><mml:mn mathvariant="normal">1</mml:mn><mml:mo>/</mml:mo><mml:mi>n</mml:mi></mml:mrow></mml:msubsup><mml:mo>.</mml:mo></mml:mrow></mml:math></inline-formula>, so that the product <inline-formula><mml:math id="M47" display="inline"><mml:mrow><mml:mn mathvariant="normal">2</mml:mn><mml:msub><mml:mi mathvariant="italic">τ</mml:mi><mml:mi mathvariant="normal">E</mml:mi></mml:msub><mml:msub><mml:mover accent="true"><mml:mi mathvariant="italic">ϵ</mml:mi><mml:mo mathvariant="normal">˙</mml:mo></mml:mover><mml:mi mathvariant="normal">E</mml:mi></mml:msub></mml:mrow></mml:math></inline-formula> is the local rate of viscous shear heating. We retain only the diffusion of this shear heating and neglect advection of heat. As a consequence, Ortholine cannot represent the narrowing of a margin by lateral advection of cold ice, a point we return to in Sect. <xref ref-type="sec" rid="Ch1.S3"/> when discussing the eastern margin of Institute. We emphasize that analytical approximations of lateral advection could be added to the model, as done in <xref ref-type="bibr" rid="bib1.bibx72" id="text.39"/>, but our current solver is not designed to solve for lateral advection numerically.</p>
      <p id="d2e1414">The boundary conditions are a fixed surface temperature, <inline-formula><mml:math id="M48" display="inline"><mml:mrow><mml:msub><mml:mi>T</mml:mi><mml:mi mathvariant="normal">s</mml:mi></mml:msub></mml:mrow></mml:math></inline-formula>,

            <disp-formula id="Ch1.E9" content-type="numbered"><label>9</label><mml:math id="M49" display="block"><mml:mrow><mml:mi>T</mml:mi><mml:mo>=</mml:mo><mml:msub><mml:mi>T</mml:mi><mml:mi mathvariant="normal">s</mml:mi></mml:msub><mml:mspace width="0.125em" linebreak="nobreak"/><mml:mspace width="0.125em" linebreak="nobreak"/><mml:mspace linebreak="nobreak" width="0.125em"/><mml:mspace linebreak="nobreak" width="0.125em"/><mml:mtext>on</mml:mtext><mml:mspace width="0.125em" linebreak="nobreak"/><mml:mspace width="0.125em" linebreak="nobreak"/><mml:mspace linebreak="nobreak" width="0.125em"/><mml:mspace width="0.125em" linebreak="nobreak"/><mml:msub><mml:mi mathvariant="normal">Γ</mml:mi><mml:mtext>surf</mml:mtext></mml:msub><mml:mo>,</mml:mo></mml:mrow></mml:math></disp-formula>

          no-flux lateral bounds,

            <disp-formula id="Ch1.E10" content-type="numbered"><label>10</label><mml:math id="M50" display="block"><mml:mrow><mml:mo>-</mml:mo><mml:mi>k</mml:mi><mml:mi mathvariant="normal">∇</mml:mi><mml:mi>T</mml:mi><mml:mo>⋅</mml:mo><mml:mi mathvariant="bold-italic">n</mml:mi><mml:mo>=</mml:mo><mml:mn mathvariant="normal">0</mml:mn><mml:mspace linebreak="nobreak" width="0.125em"/><mml:mspace width="0.125em" linebreak="nobreak"/><mml:mspace linebreak="nobreak" width="0.125em"/><mml:mspace linebreak="nobreak" width="0.125em"/><mml:mtext>on</mml:mtext><mml:mspace linebreak="nobreak" width="0.125em"/><mml:mspace linebreak="nobreak" width="0.125em"/><mml:mspace linebreak="nobreak" width="0.125em"/><mml:mspace linebreak="nobreak" width="0.125em"/><mml:msub><mml:mi mathvariant="normal">Γ</mml:mi><mml:mtext>lat</mml:mtext></mml:msub><mml:mo>,</mml:mo></mml:mrow></mml:math></disp-formula>

          and constant geothermal heat flux, <inline-formula><mml:math id="M51" display="inline"><mml:mrow><mml:msub><mml:mi>G</mml:mi><mml:mtext>geo</mml:mtext></mml:msub></mml:mrow></mml:math></inline-formula>, at the bed,

            <disp-formula id="Ch1.E11" content-type="numbered"><label>11</label><mml:math id="M52" display="block"><mml:mrow><mml:mi>k</mml:mi><mml:mi mathvariant="normal">∇</mml:mi><mml:mi>T</mml:mi><mml:mo>⋅</mml:mo><mml:mi mathvariant="bold-italic">n</mml:mi><mml:mo>=</mml:mo><mml:msub><mml:mi>G</mml:mi><mml:mi mathvariant="normal">geo</mml:mi></mml:msub><mml:mspace width="1em" linebreak="nobreak"/><mml:mtext>on </mml:mtext><mml:msub><mml:mi mathvariant="normal">Γ</mml:mi><mml:mi mathvariant="normal">bed</mml:mi></mml:msub><mml:mspace linebreak="nobreak" width="1em"/><mml:mtext>while </mml:mtext><mml:mi>T</mml:mi><mml:mo>&lt;</mml:mo><mml:msub><mml:mi>T</mml:mi><mml:mi mathvariant="normal">melt</mml:mi></mml:msub><mml:mo>.</mml:mo></mml:mrow></mml:math></disp-formula>

          The primary difficulty in solving these governing equations arises from the unknown location and extent of the slip zone. When sliding occurs over perfectly plastic sediment, the slip zone ends at the point where the sediments transition from failing to locking, but the location of these transition points depends on the integrated force balance of the ice stream. Whether the sediment fails at a given point is set by the local basal shear stress, which in turn reflects how the depth-integrated driving stress is redistributed across the whole cross-section. For example, if part of the bed is very strong, the same change in driving stress could lead to a minimal change in width, because the weak portions of the bed play a relatively less important role in the force balance than in a scenario where the basal strength is relatively even.</p>
      <p id="d2e1558">Since the redistribution depends on where the bed is already slipping, the transition points and the velocity field must be determined together rather than independently. Without knowing the locations of the transition points a priori, the solution of the governing equations includes both the depth-dependent ice speed and the location of the boundary conditions. Similarly, for the thermal model shear heating occurs until the melting point is reached, without regions of temperate ice known a priori. These types of free-boundary partial differential equations can be solved as a variational inequality <xref ref-type="bibr" rid="bib1.bibx65 bib1.bibx17" id="paren.40"/>.</p>
      <p id="d2e1564">The variational form of the mechanical free-boundary problem is derived in <xref ref-type="bibr" rid="bib1.bibx65" id="text.41"/>. In the variational formulation, solving the antiplane momentum balance, Eq. (<xref ref-type="disp-formula" rid="Ch1.E1"/>), becomes equivalent to minimizing the energy functional, <inline-formula><mml:math id="M53" display="inline"><mml:mrow><mml:mi>J</mml:mi><mml:mo>(</mml:mo><mml:mi>u</mml:mi><mml:mo>)</mml:mo></mml:mrow></mml:math></inline-formula>, subject to the constraint that flow occurs only in the downstream direction:

                <disp-formula specific-use="align" content-type="numbered"><mml:math id="M54" display="block"><mml:mtable displaystyle="true"><mml:mtr><mml:mtd><mml:mrow><mml:mstyle displaystyle="true" class="stylechange"/><mml:mtext>minimize</mml:mtext><mml:mspace linebreak="nobreak" width="1em"/><mml:mi>J</mml:mi><mml:mo>(</mml:mo><mml:mi>u</mml:mi><mml:mo>)</mml:mo></mml:mrow></mml:mtd><mml:mtd><mml:mrow><mml:mstyle displaystyle="true" class="stylechange"/><mml:mo>=</mml:mo><mml:mstyle displaystyle="true"><mml:mfrac style="display"><mml:mn mathvariant="normal">1</mml:mn><mml:mi>p</mml:mi></mml:mfrac></mml:mstyle><mml:munder><mml:mo movablelimits="false">∫</mml:mo><mml:mi mathvariant="normal">Ω</mml:mi></mml:munder><mml:mi mathvariant="italic">η</mml:mi><mml:mo>(</mml:mo><mml:mi>T</mml:mi><mml:mo>)</mml:mo><mml:mo>|</mml:mo><mml:mi mathvariant="normal">∇</mml:mi><mml:mi>u</mml:mi><mml:msup><mml:mo>|</mml:mo><mml:mi>p</mml:mi></mml:msup><mml:mo>+</mml:mo><mml:munder><mml:mo movablelimits="false">∫</mml:mo><mml:mrow><mml:msub><mml:mi mathvariant="normal">Γ</mml:mi><mml:mi mathvariant="normal">rock</mml:mi></mml:msub></mml:mrow></mml:munder><mml:msup><mml:mi mathvariant="italic">β</mml:mi><mml:mn mathvariant="normal">2</mml:mn></mml:msup><mml:mo>|</mml:mo><mml:mi>u</mml:mi><mml:msup><mml:mo>|</mml:mo><mml:mrow><mml:mi>m</mml:mi><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:mstyle class="stylechange" displaystyle="true"/></mml:mtd><mml:mtd><mml:mrow><mml:mstyle class="stylechange" displaystyle="true"/><mml:mo>+</mml:mo><mml:munder><mml:mo movablelimits="false">∫</mml:mo><mml:mrow><mml:msub><mml:mi mathvariant="normal">Γ</mml:mi><mml:mi mathvariant="normal">sed</mml:mi></mml:msub></mml:mrow></mml:munder><mml:msub><mml:mi mathvariant="italic">τ</mml:mi><mml:mi>c</mml:mi></mml:msub><mml:mo>|</mml:mo><mml:mi>u</mml:mi><mml:mo>|</mml:mo><mml:mo>-</mml:mo><mml:munder><mml:mo movablelimits="false">∫</mml:mo><mml:mi mathvariant="normal">Ω</mml:mi></mml:munder><mml:mi mathvariant="italic">ρ</mml:mi><mml:mi>g</mml:mi><mml:mi>sin⁡</mml:mi><mml:mi mathvariant="italic">α</mml:mi><mml:mspace linebreak="nobreak" width="0.125em"/><mml:mi>u</mml:mi><mml:mo>,</mml:mo></mml:mrow></mml:mtd></mml:mtr><mml:mlabeledtr id="Ch1.E12"><mml:mtd><mml:mtext>12</mml:mtext></mml:mtd><mml:mtd><mml:mstyle class="stylechange" displaystyle="true"/></mml:mtd><mml:mtd><mml:mrow><mml:mstyle class="stylechange" displaystyle="true"/><mml:mtext>subject to </mml:mtext><mml:mi>u</mml:mi><mml:mo>≥</mml:mo><mml:mn mathvariant="normal">0</mml:mn><mml:mo>.</mml:mo></mml:mrow></mml:mtd></mml:mlabeledtr></mml:mtable></mml:math></disp-formula>

          Here <inline-formula><mml:math id="M55" display="inline"><mml:mrow><mml:msub><mml:mi mathvariant="normal">Γ</mml:mi><mml:mi mathvariant="normal">rock</mml:mi></mml:msub></mml:mrow></mml:math></inline-formula> and <inline-formula><mml:math id="M56" display="inline"><mml:mrow><mml:msub><mml:mi mathvariant="normal">Γ</mml:mi><mml:mi mathvariant="normal">sed</mml:mi></mml:msub></mml:mrow></mml:math></inline-formula> are the hard-rock and sediment portions of the bed, and <inline-formula><mml:math id="M57" display="inline"><mml:mrow><mml:msub><mml:mi mathvariant="italic">τ</mml:mi><mml:mi>c</mml:mi></mml:msub><mml:mo>=</mml:mo><mml:msub><mml:mi>C</mml:mi><mml:mi mathvariant="normal">f</mml:mi></mml:msub><mml:mi>N</mml:mi><mml:mo>+</mml:mo><mml:mi>c</mml:mi></mml:mrow></mml:math></inline-formula> on sediment. For <inline-formula><mml:math id="M58" display="inline"><mml:mrow><mml:mi>s</mml:mi><mml:mo>≥</mml:mo><mml:mn mathvariant="normal">0</mml:mn></mml:mrow></mml:math></inline-formula>, define <inline-formula><mml:math id="M59" display="inline"><mml:mrow><mml:msub><mml:mi>q</mml:mi><mml:mi mathvariant="normal">b</mml:mi></mml:msub><mml:mo>(</mml:mo><mml:mi>s</mml:mi><mml:mo>)</mml:mo><mml:mo>=</mml:mo><mml:msup><mml:mi mathvariant="italic">β</mml:mi><mml:mn mathvariant="normal">2</mml:mn></mml:msup><mml:msup><mml:mi>s</mml:mi><mml:mi>m</mml:mi></mml:msup></mml:mrow></mml:math></inline-formula> on hard rock and <inline-formula><mml:math id="M60" display="inline"><mml:mrow><mml:msub><mml:mi>q</mml:mi><mml:mi mathvariant="normal">b</mml:mi></mml:msub><mml:mo>(</mml:mo><mml:mi>s</mml:mi><mml:mo>)</mml:mo><mml:mo>=</mml:mo><mml:msub><mml:mi mathvariant="italic">τ</mml:mi><mml:mi>c</mml:mi></mml:msub></mml:mrow></mml:math></inline-formula> on sediment. The basal potential is <inline-formula><mml:math id="M61" display="inline"><mml:mrow><mml:msub><mml:mi>q</mml:mi><mml:mi mathvariant="normal">b</mml:mi></mml:msub><mml:mo>(</mml:mo><mml:mo>|</mml:mo><mml:mi>u</mml:mi><mml:mo>|</mml:mo><mml:mo>)</mml:mo><mml:mo>|</mml:mo><mml:mi>u</mml:mi><mml:mo>|</mml:mo></mml:mrow></mml:math></inline-formula>. This energy coefficient differs from physical traction: <inline-formula><mml:math id="M62" display="inline"><mml:mrow><mml:msub><mml:mi mathvariant="italic">τ</mml:mi><mml:mi mathvariant="normal">b</mml:mi></mml:msub><mml:mo>=</mml:mo><mml:mo>(</mml:mo><mml:mi>m</mml:mi><mml:mo>+</mml:mo><mml:mn mathvariant="normal">1</mml:mn><mml:mo>)</mml:mo><mml:msub><mml:mi>q</mml:mi><mml:mi mathvariant="normal">b</mml:mi></mml:msub></mml:mrow></mml:math></inline-formula> on hard rock and <inline-formula><mml:math id="M63" display="inline"><mml:mrow><mml:msub><mml:mi mathvariant="italic">τ</mml:mi><mml:mi mathvariant="normal">b</mml:mi></mml:msub><mml:mo>=</mml:mo><mml:msub><mml:mi>q</mml:mi><mml:mi mathvariant="normal">b</mml:mi></mml:msub></mml:mrow></mml:math></inline-formula> on sliding sediment. In the vector equations, <inline-formula><mml:math id="M64" display="inline"><mml:mrow><mml:msub><mml:mi>q</mml:mi><mml:mi mathvariant="normal">b</mml:mi></mml:msub></mml:mrow></mml:math></inline-formula> acts componentwise. To simplify the notation, we set <inline-formula><mml:math id="M65" display="inline"><mml:mrow><mml:mi>p</mml:mi><mml:mo>=</mml:mo><mml:mn mathvariant="normal">1</mml:mn><mml:mo>+</mml:mo><mml:mn mathvariant="normal">1</mml:mn><mml:mo>/</mml:mo><mml:mi>n</mml:mi></mml:mrow></mml:math></inline-formula>. The energy functional, <inline-formula><mml:math id="M66" display="inline"><mml:mrow><mml:mi>J</mml:mi><mml:mo>(</mml:mo><mml:mi>u</mml:mi><mml:mo>)</mml:mo></mml:mrow></mml:math></inline-formula>, is convex in <inline-formula><mml:math id="M67" display="inline"><mml:mrow><mml:mi>u</mml:mi><mml:mo>≥</mml:mo><mml:mn mathvariant="normal">0</mml:mn></mml:mrow></mml:math></inline-formula>, and its minimizer is unique when a solution exists <xref ref-type="bibr" rid="bib1.bibx65" id="paren.42"/>. Throughout, we use <inline-formula><mml:math id="M68" display="inline"><mml:mi>u</mml:mi></mml:math></inline-formula> for the antiplane velocity field that solves the minimization problem; where a trial field appears in the discretized expressions below we denote it <inline-formula><mml:math id="M69" display="inline"><mml:mi mathvariant="bold">v</mml:mi></mml:math></inline-formula>, and the minimizer satisfies <inline-formula><mml:math id="M70" display="inline"><mml:mrow><mml:mi mathvariant="bold">v</mml:mi><mml:mo>=</mml:mo><mml:mi mathvariant="bold">u</mml:mi></mml:mrow></mml:math></inline-formula>.</p>
      <p id="d2e2014">We use the same method to develop a variational inequality for englacial temperature in the domain. We define the thermal energy functional, <inline-formula><mml:math id="M71" display="inline"><mml:mrow><mml:mi>L</mml:mi><mml:mo>(</mml:mo><mml:mi>T</mml:mi><mml:mo>)</mml:mo></mml:mrow></mml:math></inline-formula>, which we minimize subject to the constraint that the temperature cannot exceed the pressure melting point:

            <disp-formula id="Ch1.E13" content-type="numbered"><label>13</label><mml:math id="M72" display="block"><mml:mtable class="split" rowspacing="0.2ex" displaystyle="true" columnalign="right left"><mml:mtr><mml:mtd><mml:mrow><mml:mtext>minimize</mml:mtext><mml:mspace linebreak="nobreak" width="0.125em"/><mml:mspace linebreak="nobreak" width="0.125em"/><mml:mspace width="0.125em" linebreak="nobreak"/><mml:mspace linebreak="nobreak" width="0.125em"/><mml:mi>L</mml:mi><mml:mo>(</mml:mo><mml:mi>T</mml:mi><mml:mo>)</mml:mo></mml:mrow></mml:mtd><mml:mtd><mml:mrow><mml:mo>=</mml:mo><mml:mstyle displaystyle="true"><mml:mfrac style="display"><mml:mn mathvariant="normal">1</mml:mn><mml:mn mathvariant="normal">2</mml:mn></mml:mfrac></mml:mstyle><mml:munder><mml:mo movablelimits="false">∫</mml:mo><mml:mi mathvariant="normal">Ω</mml:mi></mml:munder><mml:mi>k</mml:mi><mml:mo>|</mml:mo><mml:mi mathvariant="normal">∇</mml:mi><mml:mi>T</mml:mi><mml:msup><mml:mo>|</mml:mo><mml:mn mathvariant="normal">2</mml:mn></mml:msup><mml:mo>-</mml:mo><mml:munder><mml:mo movablelimits="false">∫</mml:mo><mml:mrow><mml:msub><mml:mi mathvariant="normal">Γ</mml:mi><mml:mtext>bed</mml:mtext></mml:msub></mml:mrow></mml:munder><mml:msub><mml:mi>G</mml:mi><mml:mtext>geo</mml:mtext></mml:msub><mml:mi>T</mml:mi></mml:mrow></mml:mtd></mml:mtr><mml:mtr><mml:mtd/><mml:mtd><mml:mrow><mml:mo>-</mml:mo><mml:munder><mml:mo movablelimits="false">∫</mml:mo><mml:mi mathvariant="normal">Ω</mml:mi></mml:munder><mml:mn mathvariant="normal">2</mml:mn><mml:msub><mml:mi mathvariant="italic">τ</mml:mi><mml:mi mathvariant="normal">E</mml:mi></mml:msub><mml:mo>(</mml:mo><mml:mi>u</mml:mi><mml:mo>)</mml:mo><mml:msub><mml:mover accent="true"><mml:mi mathvariant="italic">ϵ</mml:mi><mml:mo mathvariant="normal">˙</mml:mo></mml:mover><mml:mi mathvariant="normal">E</mml:mi></mml:msub><mml:mo>(</mml:mo><mml:mi>u</mml:mi><mml:mo>)</mml:mo><mml:mi>T</mml:mi></mml:mrow></mml:mtd></mml:mtr><mml:mtr><mml:mtd/><mml:mtd><mml:mrow><mml:mtext>subject to</mml:mtext><mml:mspace linebreak="nobreak" width="0.125em"/><mml:mspace width="0.125em" linebreak="nobreak"/><mml:mspace width="0.125em" linebreak="nobreak"/><mml:mspace linebreak="nobreak" width="0.125em"/><mml:mi>T</mml:mi><mml:mo>≤</mml:mo><mml:msub><mml:mi>T</mml:mi><mml:mtext>melt</mml:mtext></mml:msub><mml:mo>.</mml:mo></mml:mrow></mml:mtd></mml:mtr></mml:mtable></mml:math></disp-formula>

          Our model hence extends the approach of <xref ref-type="bibr" rid="bib1.bibx43" id="text.43"/> to two dimensions.</p>
      <p id="d2e2167">The constraint <inline-formula><mml:math id="M73" display="inline"><mml:mrow><mml:mi>T</mml:mi><mml:mo>≤</mml:mo><mml:msub><mml:mi>T</mml:mi><mml:mtext>melt</mml:mtext></mml:msub></mml:mrow></mml:math></inline-formula> in Eq. (13) makes the thermal problem a free-boundary problem in which the extent of the temperate region is part of the solution rather than prescribed in advance. Where the constraint is inactive, the temperature satisfies the diffusion balance (Eq. 8) with the geothermal flux entering through the Neumann condition (Eq. 11). Where the constraint is active, the temperature remains at the pressure-melting point, <inline-formula><mml:math id="M74" display="inline"><mml:mrow><mml:mi>T</mml:mi><mml:mo>=</mml:mo><mml:msub><mml:mi>T</mml:mi><mml:mtext>melt</mml:mtext></mml:msub></mml:mrow></mml:math></inline-formula>, and the surplus energy that cannot raise the temperature further instead drives melting. This surplus is given by the Lagrange multiplier, <inline-formula><mml:math id="M75" display="inline"><mml:mi mathvariant="italic">λ</mml:mi></mml:math></inline-formula>, associated with the constraint: at the optimum, <inline-formula><mml:math id="M76" display="inline"><mml:mi mathvariant="italic">λ</mml:mi></mml:math></inline-formula> is nonzero only where <inline-formula><mml:math id="M77" display="inline"><mml:mrow><mml:mi>T</mml:mi><mml:mo>=</mml:mo><mml:msub><mml:mi>T</mml:mi><mml:mtext>melt</mml:mtext></mml:msub></mml:mrow></mml:math></inline-formula> and equals the local excess heat flux at those nodes. After solving the thermal minimization with CVX, we recover <inline-formula><mml:math id="M78" display="inline"><mml:mi mathvariant="italic">λ</mml:mi></mml:math></inline-formula> at the basal nodes, normalize by the boundary integration weights, <inline-formula><mml:math id="M79" display="inline"><mml:mrow><mml:msub><mml:mi>d</mml:mi><mml:mi mathvariant="italic">γ</mml:mi></mml:msub></mml:mrow></mml:math></inline-formula>, to obtain a basal heat flux, and convert it to a melt rate, <inline-formula><mml:math id="M80" display="inline"><mml:mover accent="true"><mml:mi>m</mml:mi><mml:mo mathvariant="normal">˙</mml:mo></mml:mover></mml:math></inline-formula>, by dividing by the latent heat of fusion, <inline-formula><mml:math id="M81" display="inline"><mml:mrow><mml:msub><mml:mi>L</mml:mi><mml:mi mathvariant="normal">f</mml:mi></mml:msub></mml:mrow></mml:math></inline-formula>,

            <disp-formula id="Ch1.E14" content-type="numbered"><label>14</label><mml:math id="M82" display="block"><mml:mrow><mml:mover accent="true"><mml:mi>m</mml:mi><mml:mo mathvariant="normal">˙</mml:mo></mml:mover><mml:mo>=</mml:mo><mml:mstyle displaystyle="true"><mml:mfrac style="display"><mml:mi mathvariant="italic">λ</mml:mi><mml:mrow><mml:mi mathvariant="italic">ρ</mml:mi><mml:msub><mml:mi>L</mml:mi><mml:mi mathvariant="normal">f</mml:mi></mml:msub><mml:mspace linebreak="nobreak" width="0.125em"/><mml:msub><mml:mi>d</mml:mi><mml:mi mathvariant="italic">γ</mml:mi></mml:msub></mml:mrow></mml:mfrac></mml:mstyle><mml:mo>.</mml:mo></mml:mrow></mml:math></disp-formula>

          The melt rate is therefore diagnosed directly from the optimality conditions of the free-boundary thermal problem rather than imposed through an additional constitutive relation. The constraint can become active either at the bed or within the ice column, so the thermal model identifies where melting occurs rather than assuming it a priori; <inline-formula><mml:math id="M83" display="inline"><mml:mover accent="true"><mml:mi>m</mml:mi><mml:mo mathvariant="normal">˙</mml:mo></mml:mover></mml:math></inline-formula> as defined in Eq. (<xref ref-type="disp-formula" rid="Ch1.E14"/>) is the basal melt rate recovered at the bed, and englacial melting, where it occurs, is computed from the same multiplier at the corresponding interior nodes.</p>
</sec>
<sec id="Ch1.S2.SS2">
  <label>2.2</label><title>Numerical Method</title>
      <p id="d2e2323">As described above, we reformulate the initial governing equations as a set of constrained minimization problems that are convex in the antiplane speed, <inline-formula><mml:math id="M84" display="inline"><mml:mi>u</mml:mi></mml:math></inline-formula>, and temperature, <inline-formula><mml:math id="M85" display="inline"><mml:mi>T</mml:mi></mml:math></inline-formula>. For a given domain, we solve these optimization problems by discretizing the domain and solving the convex minimization numerically. We begin by discretizing the mechanical-energy functional, described by Eq. (<xref ref-type="disp-formula" rid="Ch1.E12"/>), on a domain, <inline-formula><mml:math id="M86" display="inline"><mml:mi mathvariant="normal">Ω</mml:mi></mml:math></inline-formula>, with an arbitrary triangulation consisting of elements, <inline-formula><mml:math id="M87" display="inline"><mml:mi>K</mml:mi></mml:math></inline-formula>, and boundary edges, <inline-formula><mml:math id="M88" display="inline"><mml:mi mathvariant="italic">γ</mml:mi></mml:math></inline-formula>,

            <disp-formula id="Ch1.E15" content-type="numbered"><label>15</label><mml:math id="M89" display="block"><mml:mtable rowspacing="0.2ex" class="split" displaystyle="true" columnalign="right left"><mml:mtr><mml:mtd><mml:mrow><mml:mi>J</mml:mi><mml:mo>(</mml:mo><mml:mi>u</mml:mi><mml:mo>)</mml:mo></mml:mrow></mml:mtd><mml:mtd><mml:mrow><mml:mo>=</mml:mo><mml:mstyle displaystyle="true"><mml:mfrac style="display"><mml:mn mathvariant="normal">1</mml:mn><mml:mi>p</mml:mi></mml:mfrac></mml:mstyle><mml:munder><mml:mo movablelimits="false">∑</mml:mo><mml:mi>K</mml:mi></mml:munder><mml:munder><mml:mo movablelimits="false">∫</mml:mo><mml:mi>K</mml:mi></mml:munder><mml:mi mathvariant="italic">η</mml:mi><mml:mo>(</mml:mo><mml:mi>T</mml:mi><mml:mo>)</mml:mo><mml:mo>|</mml:mo><mml:mi mathvariant="normal">∇</mml:mi><mml:mi>u</mml:mi><mml:msup><mml:mo>|</mml:mo><mml:mi>p</mml:mi></mml:msup><mml:mo>+</mml:mo><mml:munder><mml:mo movablelimits="false">∑</mml:mo><mml:mi mathvariant="italic">γ</mml:mi></mml:munder><mml:munder><mml:mo movablelimits="false">∫</mml:mo><mml:mi mathvariant="italic">γ</mml:mi></mml:munder><mml:msub><mml:mi>q</mml:mi><mml:mi mathvariant="normal">b</mml:mi></mml:msub><mml:mo>(</mml:mo><mml:mo>|</mml:mo><mml:mi>u</mml:mi><mml:mo>|</mml:mo><mml:mo>)</mml:mo><mml:mo>|</mml:mo><mml:mi>u</mml:mi><mml:mo>|</mml:mo></mml:mrow></mml:mtd></mml:mtr><mml:mtr><mml:mtd/><mml:mtd><mml:mrow><mml:mo>-</mml:mo><mml:munder><mml:mo movablelimits="false">∑</mml:mo><mml:mi>K</mml:mi></mml:munder><mml:munder><mml:mo movablelimits="false">∫</mml:mo><mml:mi>K</mml:mi></mml:munder><mml:mi>f</mml:mi><mml:mi>u</mml:mi><mml:mo>.</mml:mo></mml:mrow></mml:mtd></mml:mtr></mml:mtable></mml:math></disp-formula>

          Choosing a piecewise-linear basis, <inline-formula><mml:math id="M90" display="inline"><mml:mrow><mml:msub><mml:mi mathvariant="italic">ϕ</mml:mi><mml:mi>i</mml:mi></mml:msub></mml:mrow></mml:math></inline-formula>, on each element with nodal weight function, <inline-formula><mml:math id="M91" display="inline"><mml:mrow><mml:msub><mml:mi>v</mml:mi><mml:mi>i</mml:mi></mml:msub></mml:mrow></mml:math></inline-formula>, the semi-discretized energy functional becomes

            <disp-formula id="Ch1.E16" content-type="numbered"><label>16</label><mml:math id="M92" display="block"><mml:mtable class="split" rowspacing="0.2ex" displaystyle="true" columnalign="right left"><mml:mtr><mml:mtd><mml:mrow><mml:mi>J</mml:mi><mml:mo>(</mml:mo><mml:msub><mml:mi>v</mml:mi><mml:mi>i</mml:mi></mml:msub><mml:mo>)</mml:mo></mml:mrow></mml:mtd><mml:mtd><mml:mrow><mml:mo>=</mml:mo><mml:mstyle displaystyle="true"><mml:mfrac style="display"><mml:mn mathvariant="normal">1</mml:mn><mml:mi>p</mml:mi></mml:mfrac></mml:mstyle><mml:munder><mml:mo movablelimits="false">∑</mml:mo><mml:mi>K</mml:mi></mml:munder><mml:msub><mml:mi mathvariant="italic">η</mml:mi><mml:mi>K</mml:mi></mml:msub><mml:mo>(</mml:mo><mml:mi>T</mml:mi><mml:mo>)</mml:mo><mml:munder><mml:mo movablelimits="false">∫</mml:mo><mml:mi>K</mml:mi></mml:munder><mml:msup><mml:mfenced close="|" open="|"><mml:mrow><mml:munderover><mml:mo movablelimits="false">∑</mml:mo><mml:mrow><mml:mi>i</mml:mi><mml:mo>=</mml:mo><mml:mn mathvariant="normal">1</mml:mn></mml:mrow><mml:mn mathvariant="normal">3</mml:mn></mml:munderover><mml:msub><mml:mi>v</mml:mi><mml:mi>i</mml:mi></mml:msub><mml:mi mathvariant="normal">∇</mml:mi><mml:msub><mml:mi mathvariant="italic">ϕ</mml:mi><mml:mi>i</mml:mi></mml:msub></mml:mrow></mml:mfenced><mml:mi>p</mml:mi></mml:msup></mml:mrow></mml:mtd></mml:mtr><mml:mtr><mml:mtd/><mml:mtd><mml:mrow><mml:mo>+</mml:mo><mml:munder><mml:mo movablelimits="false">∑</mml:mo><mml:mi mathvariant="italic">γ</mml:mi></mml:munder><mml:munderover><mml:mo movablelimits="false">∑</mml:mo><mml:mrow><mml:mi>i</mml:mi><mml:mo>=</mml:mo><mml:mn mathvariant="normal">1</mml:mn></mml:mrow><mml:mn mathvariant="normal">2</mml:mn></mml:munderover><mml:msub><mml:mi>v</mml:mi><mml:mi>i</mml:mi></mml:msub><mml:munder><mml:mo movablelimits="false">∫</mml:mo><mml:mi mathvariant="italic">γ</mml:mi></mml:munder><mml:msub><mml:mi>q</mml:mi><mml:mi mathvariant="normal">b</mml:mi></mml:msub><mml:mo>(</mml:mo><mml:msub><mml:mi>v</mml:mi><mml:mi>h</mml:mi></mml:msub><mml:mo>)</mml:mo><mml:msub><mml:mi mathvariant="italic">ϕ</mml:mi><mml:mi>i</mml:mi></mml:msub><mml:mo>-</mml:mo><mml:mi>f</mml:mi><mml:munder><mml:mo movablelimits="false">∑</mml:mo><mml:mi>K</mml:mi></mml:munder><mml:munderover><mml:mo movablelimits="false">∑</mml:mo><mml:mrow><mml:mi>i</mml:mi><mml:mo>=</mml:mo><mml:mn mathvariant="normal">1</mml:mn></mml:mrow><mml:mn mathvariant="normal">3</mml:mn></mml:munderover><mml:msub><mml:mi>v</mml:mi><mml:mi>i</mml:mi></mml:msub><mml:munder><mml:mo movablelimits="false">∫</mml:mo><mml:mi>K</mml:mi></mml:munder><mml:msub><mml:mi mathvariant="italic">ϕ</mml:mi><mml:mi>i</mml:mi></mml:msub><mml:mo>.</mml:mo></mml:mrow></mml:mtd></mml:mtr></mml:mtable></mml:math></disp-formula>

          On each element or edge, <inline-formula><mml:math id="M93" display="inline"><mml:mrow><mml:msub><mml:mi>v</mml:mi><mml:mi>h</mml:mi></mml:msub><mml:mo>=</mml:mo><mml:msub><mml:mo>∑</mml:mo><mml:mi>i</mml:mi></mml:msub><mml:msub><mml:mi>v</mml:mi><mml:mi>i</mml:mi></mml:msub><mml:msub><mml:mi mathvariant="italic">ϕ</mml:mi><mml:mi>i</mml:mi></mml:msub><mml:mo>≥</mml:mo><mml:mn mathvariant="normal">0</mml:mn></mml:mrow></mml:math></inline-formula> is the piecewise-linear trial field formed from its local nodal values.</p>
      <p id="d2e2702">The next step is to fully discretize each term of the energy functional, while ensuring that each term follows the convex composition rules of Disciplined Convex Programming described in <xref ref-type="bibr" rid="bib1.bibx22" id="text.44"/>. Disciplined Convex Programming is a framework for specifying convex minimization problems algorithmically, which can then be solved by generalized, efficient numerical methods <xref ref-type="bibr" rid="bib1.bibx21" id="paren.45"/>.</p>
      <p id="d2e2711">To demonstrate how the minimization problem in Eq. (<xref ref-type="disp-formula" rid="Ch1.E16"/>) can be reformulated to be amenable to the CVX algorithm, where CVX is short for “convex” <xref ref-type="bibr" rid="bib1.bibx21" id="paren.46"/>, each of the terms in the energy functional must be decomposed into the composition of convex operators. We define each term in the energy functional described by Eq. (<xref ref-type="disp-formula" rid="Ch1.E16"/>) as

                <disp-formula id="Ch1.E17" content-type="numbered"><label>17</label><mml:math id="M94" display="block"><mml:mrow><mml:mstyle class="stylechange" displaystyle="true"/><mml:mi>J</mml:mi><mml:mo>(</mml:mo><mml:msub><mml:mi>v</mml:mi><mml:mi>i</mml:mi></mml:msub><mml:mo>)</mml:mo></mml:mrow><mml:mrow><mml:mstyle class="stylechange" displaystyle="true"/><mml:mo>=</mml:mo><mml:msub><mml:mi>J</mml:mi><mml:mi>A</mml:mi></mml:msub><mml:mo>(</mml:mo><mml:msub><mml:mi>v</mml:mi><mml:mi>i</mml:mi></mml:msub><mml:mo>)</mml:mo><mml:mo>+</mml:mo><mml:msub><mml:mi>J</mml:mi><mml:mi>B</mml:mi></mml:msub><mml:mo>(</mml:mo><mml:msub><mml:mi>v</mml:mi><mml:mi>i</mml:mi></mml:msub><mml:mo>)</mml:mo><mml:mo>+</mml:mo><mml:msub><mml:mi>J</mml:mi><mml:mi>C</mml:mi></mml:msub><mml:mo>(</mml:mo><mml:msub><mml:mi>v</mml:mi><mml:mi>i</mml:mi></mml:msub><mml:mo>)</mml:mo><mml:mo>.</mml:mo></mml:mrow></mml:math></disp-formula>

          For example, <inline-formula><mml:math id="M95" display="inline"><mml:mrow><mml:msub><mml:mi>J</mml:mi><mml:mi>A</mml:mi></mml:msub><mml:mo>(</mml:mo><mml:msub><mml:mi>v</mml:mi><mml:mi>i</mml:mi></mml:msub><mml:mo>)</mml:mo></mml:mrow></mml:math></inline-formula> on a single element, <inline-formula><mml:math id="M96" display="inline"><mml:mi>K</mml:mi></mml:math></inline-formula>, becomes

            <disp-formula id="Ch1.E18" content-type="numbered"><label>18</label><mml:math id="M97" display="block"><mml:mtable class="split" rowspacing="0.2ex" displaystyle="true" columnalign="right left"><mml:mtr><mml:mtd><mml:mrow><mml:msub><mml:mi>J</mml:mi><mml:mi>A</mml:mi></mml:msub><mml:mo>(</mml:mo><mml:msub><mml:mi>v</mml:mi><mml:mi>i</mml:mi></mml:msub><mml:mo>)</mml:mo></mml:mrow></mml:mtd><mml:mtd><mml:mrow><mml:mo>=</mml:mo><mml:mstyle displaystyle="true"><mml:mfrac style="display"><mml:mn mathvariant="normal">1</mml:mn><mml:mi>p</mml:mi></mml:mfrac></mml:mstyle><mml:mi mathvariant="italic">η</mml:mi><mml:munder><mml:mo movablelimits="false">∫</mml:mo><mml:mi>K</mml:mi></mml:munder><mml:msup><mml:mfenced close="|" open="|"><mml:mrow><mml:munderover><mml:mo movablelimits="false">∑</mml:mo><mml:mrow><mml:mi>i</mml:mi><mml:mo>=</mml:mo><mml:mn mathvariant="normal">1</mml:mn></mml:mrow><mml:mn mathvariant="normal">3</mml:mn></mml:munderover><mml:msub><mml:mi>v</mml:mi><mml:mi>i</mml:mi></mml:msub><mml:mi mathvariant="normal">∇</mml:mi><mml:msub><mml:mi mathvariant="italic">ϕ</mml:mi><mml:mi>i</mml:mi></mml:msub></mml:mrow></mml:mfenced><mml:mi>p</mml:mi></mml:msup></mml:mrow></mml:mtd></mml:mtr><mml:mtr><mml:mtd/><mml:mtd><mml:mrow><mml:mo>≈</mml:mo><mml:mstyle displaystyle="true"><mml:mfrac style="display"><mml:mn mathvariant="normal">1</mml:mn><mml:mi>p</mml:mi></mml:mfrac></mml:mstyle><mml:mi mathvariant="italic">η</mml:mi><mml:mspace linebreak="nobreak" width="0.125em"/><mml:mi mathvariant="normal">d</mml:mi><mml:mi>K</mml:mi><mml:msup><mml:mfenced close="|" open="|"><mml:mfenced open="[" close="]"><mml:mtable class="matrix" columnalign="center" framespacing="0em"><mml:mtr><mml:mtd><mml:mrow><mml:msub><mml:mi mathvariant="bold">D</mml:mi><mml:mi>x</mml:mi></mml:msub><mml:msub><mml:mi mathvariant="bold-italic">v</mml:mi><mml:mi>K</mml:mi></mml:msub></mml:mrow></mml:mtd></mml:mtr><mml:mtr><mml:mtd><mml:mrow><mml:msub><mml:mi mathvariant="bold">D</mml:mi><mml:mi>y</mml:mi></mml:msub><mml:msub><mml:mi mathvariant="bold-italic">v</mml:mi><mml:mi>K</mml:mi></mml:msub></mml:mrow></mml:mtd></mml:mtr></mml:mtable></mml:mfenced></mml:mfenced><mml:mi>p</mml:mi></mml:msup><mml:mo>,</mml:mo></mml:mrow></mml:mtd></mml:mtr></mml:mtable></mml:math></disp-formula>

          with discretized gradient operators, <inline-formula><mml:math id="M98" display="inline"><mml:mrow><mml:msub><mml:mi mathvariant="bold">D</mml:mi><mml:mi>x</mml:mi></mml:msub></mml:mrow></mml:math></inline-formula> and <inline-formula><mml:math id="M99" display="inline"><mml:mrow><mml:msub><mml:mi mathvariant="bold">D</mml:mi><mml:mi>y</mml:mi></mml:msub></mml:mrow></mml:math></inline-formula>, and strictly positive element area, d<inline-formula><mml:math id="M100" display="inline"><mml:mi>K</mml:mi></mml:math></inline-formula>. The local computational coordinates <inline-formula><mml:math id="M101" display="inline"><mml:mrow><mml:mo>(</mml:mo><mml:mi>x</mml:mi><mml:mo>,</mml:mo><mml:mi>y</mml:mi><mml:mo>)</mml:mo></mml:mrow></mml:math></inline-formula> used in Eqs. (4), (18), and (25)–(27) correspond to the physical cross-sectional coordinates <inline-formula><mml:math id="M102" display="inline"><mml:mrow><mml:mo>(</mml:mo><mml:mi>y</mml:mi><mml:mo>,</mml:mo><mml:mi>z</mml:mi><mml:mo>)</mml:mo></mml:mrow></mml:math></inline-formula>, respectively; the physical <inline-formula><mml:math id="M103" display="inline"><mml:mi>x</mml:mi></mml:math></inline-formula> direction remains the omitted downstream direction. The discretized gradient operations are affine functions on element velocities, <inline-formula><mml:math id="M104" display="inline"><mml:mrow><mml:msub><mml:mi mathvariant="bold-italic">v</mml:mi><mml:mi>K</mml:mi></mml:msub></mml:mrow></mml:math></inline-formula>, which are convex. The norm operation is convex, followed by the power operation, which is convex for <inline-formula><mml:math id="M105" display="inline"><mml:mrow><mml:mi>p</mml:mi><mml:mo>&gt;</mml:mo><mml:mn mathvariant="normal">1</mml:mn></mml:mrow></mml:math></inline-formula>. Multiplication with strictly positive coefficients shows that <inline-formula><mml:math id="M106" display="inline"><mml:mrow><mml:msub><mml:mi>J</mml:mi><mml:mi>A</mml:mi></mml:msub><mml:mo>(</mml:mo><mml:msub><mml:mi>v</mml:mi><mml:mi>i</mml:mi></mml:msub><mml:mo>)</mml:mo></mml:mrow></mml:math></inline-formula> is convex.</p>
      <p id="d2e3054">Similarly, we decompose <inline-formula><mml:math id="M107" display="inline"><mml:mrow><mml:msub><mml:mi>J</mml:mi><mml:mi>B</mml:mi></mml:msub><mml:mo>(</mml:mo><mml:msub><mml:mi>v</mml:mi><mml:mi>i</mml:mi></mml:msub><mml:mo>)</mml:mo></mml:mrow></mml:math></inline-formula> on a single boundary edge, <inline-formula><mml:math id="M108" display="inline"><mml:mi mathvariant="italic">γ</mml:mi></mml:math></inline-formula>,

                <disp-formula specific-use="align" content-type="numbered"><mml:math id="M109" display="block"><mml:mtable displaystyle="true"><mml:mlabeledtr id="Ch1.E19"><mml:mtd><mml:mtext>19</mml:mtext></mml:mtd><mml:mtd><mml:mrow><mml:mstyle displaystyle="true" class="stylechange"/><mml:msub><mml:mi>J</mml:mi><mml:mi>B</mml:mi></mml:msub><mml:mo>(</mml:mo><mml:msub><mml:mi>v</mml:mi><mml:mi>h</mml:mi></mml:msub><mml:mo>)</mml:mo><mml:mo>=</mml:mo><mml:munder><mml:mo movablelimits="false">∫</mml:mo><mml:mi mathvariant="italic">γ</mml:mi></mml:munder><mml:msub><mml:mi>q</mml:mi><mml:mi mathvariant="normal">b</mml:mi></mml:msub><mml:mo>(</mml:mo><mml:msub><mml:mi>v</mml:mi><mml:mi>h</mml:mi></mml:msub><mml:mo>)</mml:mo><mml:msub><mml:mi>v</mml:mi><mml:mi>h</mml:mi></mml:msub><mml:mo>=</mml:mo><mml:munderover><mml:mo movablelimits="false">∑</mml:mo><mml:mrow><mml:mi>i</mml:mi><mml:mo>=</mml:mo><mml:mn mathvariant="normal">1</mml:mn></mml:mrow><mml:mn mathvariant="normal">2</mml:mn></mml:munderover><mml:msub><mml:mi>v</mml:mi><mml:mi>i</mml:mi></mml:msub><mml:munder><mml:mo movablelimits="false">∫</mml:mo><mml:mi mathvariant="italic">γ</mml:mi></mml:munder><mml:msub><mml:mi>q</mml:mi><mml:mi mathvariant="normal">b</mml:mi></mml:msub><mml:mo>(</mml:mo><mml:msub><mml:mi>v</mml:mi><mml:mi>h</mml:mi></mml:msub><mml:mo>)</mml:mo><mml:msub><mml:mi mathvariant="italic">ϕ</mml:mi><mml:mi>i</mml:mi></mml:msub><mml:mo>.</mml:mo></mml:mrow></mml:mtd></mml:mlabeledtr><mml:mlabeledtr id="Ch1.E20"><mml:mtd><mml:mtext>20</mml:mtext></mml:mtd><mml:mtd><mml:mrow><mml:mstyle displaystyle="true" class="stylechange"/><mml:msub><mml:mi>J</mml:mi><mml:mi>B</mml:mi></mml:msub><mml:mo>(</mml:mo><mml:msub><mml:mi>v</mml:mi><mml:mi>h</mml:mi></mml:msub><mml:mo>)</mml:mo><mml:mo>≈</mml:mo><mml:munderover><mml:mo movablelimits="false">∑</mml:mo><mml:mrow><mml:mi>i</mml:mi><mml:mo>=</mml:mo><mml:mn mathvariant="normal">1</mml:mn></mml:mrow><mml:mn mathvariant="normal">2</mml:mn></mml:munderover><mml:msub><mml:mi>v</mml:mi><mml:mi>i</mml:mi></mml:msub><mml:msub><mml:mi>q</mml:mi><mml:mi mathvariant="normal">b</mml:mi></mml:msub><mml:mo>(</mml:mo><mml:msub><mml:mi>v</mml:mi><mml:mi>i</mml:mi></mml:msub><mml:mo>)</mml:mo><mml:munder><mml:mo movablelimits="false">∫</mml:mo><mml:mi mathvariant="italic">γ</mml:mi></mml:munder><mml:msub><mml:mi mathvariant="italic">ϕ</mml:mi><mml:mi>i</mml:mi></mml:msub><mml:mo>.</mml:mo></mml:mrow></mml:mtd></mml:mlabeledtr><mml:mlabeledtr id="Ch1.E21"><mml:mtd><mml:mtext>21</mml:mtext></mml:mtd><mml:mtd><mml:mrow><mml:mstyle class="stylechange" displaystyle="true"/><mml:msub><mml:mi>J</mml:mi><mml:mi>B</mml:mi></mml:msub><mml:mo>(</mml:mo><mml:msub><mml:mi>v</mml:mi><mml:mi>h</mml:mi></mml:msub><mml:mo>)</mml:mo><mml:mo>≈</mml:mo><mml:mstyle displaystyle="true"><mml:mfrac style="display"><mml:mrow><mml:msub><mml:mi mathvariant="normal">ℓ</mml:mi><mml:mi mathvariant="italic">γ</mml:mi></mml:msub></mml:mrow><mml:mn mathvariant="normal">2</mml:mn></mml:mfrac></mml:mstyle><mml:munderover><mml:mo movablelimits="false">∑</mml:mo><mml:mrow><mml:mi>i</mml:mi><mml:mo>=</mml:mo><mml:mn mathvariant="normal">1</mml:mn></mml:mrow><mml:mn mathvariant="normal">2</mml:mn></mml:munderover><mml:msub><mml:mi>v</mml:mi><mml:mi>i</mml:mi></mml:msub><mml:msub><mml:mi>q</mml:mi><mml:mi mathvariant="normal">b</mml:mi></mml:msub><mml:mo>(</mml:mo><mml:msub><mml:mi>v</mml:mi><mml:mi>i</mml:mi></mml:msub><mml:mo>)</mml:mo><mml:mo>.</mml:mo></mml:mrow></mml:mtd></mml:mlabeledtr><mml:mlabeledtr id="Ch1.E22"><mml:mtd><mml:mtext>22</mml:mtext></mml:mtd><mml:mtd><mml:mrow><mml:mstyle class="stylechange" displaystyle="true"/><mml:msub><mml:mi>J</mml:mi><mml:mi>B</mml:mi></mml:msub><mml:mo>(</mml:mo><mml:msub><mml:mi>v</mml:mi><mml:mi>h</mml:mi></mml:msub><mml:mo>)</mml:mo><mml:mo>≈</mml:mo><mml:mstyle displaystyle="true"><mml:mfrac style="display"><mml:mrow><mml:msub><mml:mi mathvariant="normal">ℓ</mml:mi><mml:mi mathvariant="italic">γ</mml:mi></mml:msub></mml:mrow><mml:mn mathvariant="normal">2</mml:mn></mml:mfrac></mml:mstyle><mml:msub><mml:mi mathvariant="bold-italic">q</mml:mi><mml:mi mathvariant="normal">b</mml:mi></mml:msub><mml:mo>(</mml:mo><mml:msub><mml:mi mathvariant="bold-italic">v</mml:mi><mml:mi mathvariant="italic">γ</mml:mi></mml:msub><mml:msup><mml:mo>)</mml:mo><mml:mi mathvariant="normal">T</mml:mi></mml:msup><mml:msub><mml:mi mathvariant="bold-italic">v</mml:mi><mml:mi mathvariant="italic">γ</mml:mi></mml:msub><mml:mo>.</mml:mo></mml:mrow></mml:mtd></mml:mlabeledtr></mml:mtable></mml:math></disp-formula>

          Here <inline-formula><mml:math id="M110" display="inline"><mml:mrow><mml:msub><mml:mi>q</mml:mi><mml:mi mathvariant="normal">b</mml:mi></mml:msub><mml:mo>(</mml:mo><mml:msub><mml:mi>v</mml:mi><mml:mi>i</mml:mi></mml:msub><mml:mo>)</mml:mo></mml:mrow></mml:math></inline-formula> denotes the nodal basal-energy coefficient and <inline-formula><mml:math id="M111" display="inline"><mml:mrow><mml:msub><mml:mi mathvariant="normal">ℓ</mml:mi><mml:mi mathvariant="italic">γ</mml:mi></mml:msub></mml:mrow></mml:math></inline-formula> the boundary edge length. For the sliding laws considered here, <inline-formula><mml:math id="M112" display="inline"><mml:mrow><mml:mi>m</mml:mi><mml:mi mathvariant="normal">≥</mml:mi><mml:mn mathvariant="normal">0</mml:mn></mml:mrow></mml:math></inline-formula>, the hard-rock boundary integrand <inline-formula><mml:math id="M113" display="inline"><mml:mrow><mml:msup><mml:mi mathvariant="italic">β</mml:mi><mml:mn mathvariant="normal">2</mml:mn></mml:msup><mml:msup><mml:mi>v</mml:mi><mml:mrow><mml:mi>m</mml:mi><mml:mo>+</mml:mo><mml:mn mathvariant="normal">1</mml:mn></mml:mrow></mml:msup></mml:mrow></mml:math></inline-formula> is convex for <inline-formula><mml:math id="M114" display="inline"><mml:mrow><mml:mi>v</mml:mi><mml:mi mathvariant="normal">≥</mml:mi><mml:mn mathvariant="normal">0</mml:mn></mml:mrow></mml:math></inline-formula>, and the perfectly plastic sediment contribution <inline-formula><mml:math id="M115" display="inline"><mml:mrow><mml:msub><mml:mi mathvariant="italic">τ</mml:mi><mml:mi>c</mml:mi></mml:msub><mml:mi>v</mml:mi></mml:mrow></mml:math></inline-formula> is affine. We do not consider velocity-weakening sliding, <inline-formula><mml:math id="M116" display="inline"><mml:mrow><mml:mi>m</mml:mi><mml:mo>&lt;</mml:mo><mml:mn mathvariant="normal">0</mml:mn></mml:mrow></mml:math></inline-formula>, in the simulations.</p>
      <p id="d2e3478">Finally, <inline-formula><mml:math id="M117" display="inline"><mml:mrow><mml:msub><mml:mi>J</mml:mi><mml:mi>C</mml:mi></mml:msub><mml:mo>(</mml:mo><mml:msub><mml:mi>v</mml:mi><mml:mi>i</mml:mi></mml:msub><mml:mo>)</mml:mo></mml:mrow></mml:math></inline-formula> on a single element, <inline-formula><mml:math id="M118" display="inline"><mml:mi>K</mml:mi></mml:math></inline-formula>, becomes

                <disp-formula specific-use="align" content-type="numbered"><mml:math id="M119" display="block"><mml:mtable displaystyle="true"><mml:mlabeledtr id="Ch1.E23"><mml:mtd><mml:mtext>23</mml:mtext></mml:mtd><mml:mtd><mml:mrow><mml:mstyle class="stylechange" displaystyle="true"/><mml:msub><mml:mi>J</mml:mi><mml:mi>C</mml:mi></mml:msub><mml:mo>(</mml:mo><mml:msub><mml:mi>v</mml:mi><mml:mi>i</mml:mi></mml:msub><mml:mo>)</mml:mo><mml:mo>=</mml:mo><mml:mo>-</mml:mo><mml:mi>f</mml:mi><mml:munderover><mml:mo movablelimits="false">∑</mml:mo><mml:mrow><mml:mi>i</mml:mi><mml:mo>=</mml:mo><mml:mn mathvariant="normal">1</mml:mn></mml:mrow><mml:mn mathvariant="normal">3</mml:mn></mml:munderover><mml:msub><mml:mi>v</mml:mi><mml:mi>i</mml:mi></mml:msub><mml:munder><mml:mo movablelimits="false">∫</mml:mo><mml:mi>K</mml:mi></mml:munder><mml:msub><mml:mi mathvariant="italic">ϕ</mml:mi><mml:mi>i</mml:mi></mml:msub></mml:mrow></mml:mtd></mml:mlabeledtr><mml:mlabeledtr id="Ch1.E24"><mml:mtd><mml:mtext>24</mml:mtext></mml:mtd><mml:mtd><mml:mrow><mml:mstyle class="stylechange" displaystyle="true"/><mml:msub><mml:mi>J</mml:mi><mml:mi>C</mml:mi></mml:msub><mml:mo>(</mml:mo><mml:msub><mml:mi>v</mml:mi><mml:mi>i</mml:mi></mml:msub><mml:mo>)</mml:mo><mml:mo>=</mml:mo><mml:mo>-</mml:mo><mml:mi>f</mml:mi><mml:mstyle displaystyle="true"><mml:mfrac style="display"><mml:mrow><mml:msub><mml:mi>a</mml:mi><mml:mi>K</mml:mi></mml:msub></mml:mrow><mml:mn mathvariant="normal">3</mml:mn></mml:mfrac></mml:mstyle><mml:msup><mml:mn mathvariant="bold">1</mml:mn><mml:mi mathvariant="normal">T</mml:mi></mml:msup><mml:msub><mml:mi mathvariant="bold-italic">v</mml:mi><mml:mi>K</mml:mi></mml:msub><mml:mo>.</mml:mo></mml:mrow></mml:mtd></mml:mlabeledtr></mml:mtable></mml:math></disp-formula>

          Here <inline-formula><mml:math id="M120" display="inline"><mml:mrow><mml:mi>f</mml:mi><mml:mo>=</mml:mo><mml:mi mathvariant="italic">ρ</mml:mi><mml:mi>g</mml:mi><mml:mi>sin⁡</mml:mi><mml:mi mathvariant="italic">α</mml:mi></mml:mrow></mml:math></inline-formula> is body-force density (N m<sup>−3</sup>), <inline-formula><mml:math id="M122" display="inline"><mml:mrow><mml:msub><mml:mi>a</mml:mi><mml:mi>K</mml:mi></mml:msub><mml:mo>=</mml:mo><mml:mo>|</mml:mo><mml:mi>K</mml:mi><mml:mo>|</mml:mo></mml:mrow></mml:math></inline-formula> is the geometric triangle area (the local d<inline-formula><mml:math id="M123" display="inline"><mml:mi>K</mml:mi></mml:math></inline-formula>), and <inline-formula><mml:math id="M124" display="inline"><mml:mn mathvariant="bold">1</mml:mn></mml:math></inline-formula> is the three-component vector of ones. Since <inline-formula><mml:math id="M125" display="inline"><mml:mrow><mml:msub><mml:mi>J</mml:mi><mml:mi>C</mml:mi></mml:msub><mml:mo>(</mml:mo><mml:msub><mml:mi>v</mml:mi><mml:mi>i</mml:mi></mml:msub><mml:mo>)</mml:mo></mml:mrow></mml:math></inline-formula> is an affine function on element velocities, <inline-formula><mml:math id="M126" display="inline"><mml:mrow><mml:msub><mml:mi mathvariant="bold-italic">v</mml:mi><mml:mi>K</mml:mi></mml:msub></mml:mrow></mml:math></inline-formula>, it is therefore convex.</p>
      <p id="d2e3714">The summation of convex functions is convex, proving that the fully discretized energy functional (i.e., the summation of each of these terms over all elements and boundary edges) is convex in <inline-formula><mml:math id="M127" display="inline"><mml:mi>v</mml:mi></mml:math></inline-formula>. Simplifying and vectorizing, the fully discretized energy becomes

            <disp-formula id="Ch1.E25" content-type="numbered"><label>25</label><mml:math id="M128" display="block"><mml:mtable rowspacing="0.2ex" class="split" displaystyle="true" columnalign="right left"><mml:mtr><mml:mtd><mml:mrow><mml:mi>J</mml:mi><mml:mo>(</mml:mo><mml:mi mathvariant="bold-italic">v</mml:mi><mml:mo>)</mml:mo></mml:mrow></mml:mtd><mml:mtd><mml:mrow><mml:mo>=</mml:mo><mml:mstyle displaystyle="true"><mml:mfrac style="display"><mml:mi mathvariant="italic">η</mml:mi><mml:mi>p</mml:mi></mml:mfrac></mml:mstyle><mml:mspace linebreak="nobreak" width="0.125em"/><mml:msup><mml:mi mathvariant="bold-italic">a</mml:mi><mml:mi mathvariant="normal">T</mml:mi></mml:msup><mml:msup><mml:mfenced open="|" close="|"><mml:mrow><mml:mo>[</mml:mo><mml:msub><mml:mi mathvariant="bold">D</mml:mi><mml:mi>x</mml:mi></mml:msub><mml:mi mathvariant="bold-italic">v</mml:mi><mml:mo>,</mml:mo><mml:msub><mml:mi mathvariant="bold">D</mml:mi><mml:mi>y</mml:mi></mml:msub><mml:mi mathvariant="bold-italic">v</mml:mi><mml:mo>]</mml:mo></mml:mrow></mml:mfenced><mml:mi>p</mml:mi></mml:msup><mml:mo>+</mml:mo><mml:msup><mml:mfenced open="[" close="]"><mml:mrow><mml:msub><mml:mi mathvariant="bold-italic">w</mml:mi><mml:mi mathvariant="normal">b</mml:mi></mml:msub><mml:mo>∘</mml:mo><mml:msub><mml:mi mathvariant="bold-italic">q</mml:mi><mml:mi mathvariant="normal">b</mml:mi></mml:msub><mml:mo>(</mml:mo><mml:msub><mml:mi mathvariant="bold-italic">v</mml:mi><mml:mi mathvariant="normal">b</mml:mi></mml:msub><mml:mo>)</mml:mo></mml:mrow></mml:mfenced><mml:mi mathvariant="normal">T</mml:mi></mml:msup><mml:msub><mml:mi mathvariant="bold-italic">v</mml:mi><mml:mi mathvariant="normal">b</mml:mi></mml:msub></mml:mrow></mml:mtd></mml:mtr><mml:mtr><mml:mtd/><mml:mtd><mml:mrow><mml:mo>-</mml:mo><mml:mo>(</mml:mo><mml:mi mathvariant="bold-italic">w</mml:mi><mml:mo>∘</mml:mo><mml:mi mathvariant="bold-italic">f</mml:mi><mml:msup><mml:mo>)</mml:mo><mml:mi mathvariant="normal">T</mml:mi></mml:msup><mml:mi mathvariant="bold-italic">v</mml:mi><mml:mo>.</mml:mo></mml:mrow></mml:mtd></mml:mtr></mml:mtable></mml:math></disp-formula>

          Let <inline-formula><mml:math id="M129" display="inline"><mml:mrow><mml:mi mathvariant="bold-italic">a</mml:mi><mml:mo>=</mml:mo><mml:mo>(</mml:mo><mml:mo>|</mml:mo><mml:msub><mml:mi>K</mml:mi><mml:mi mathvariant="normal">e</mml:mi></mml:msub><mml:mo>|</mml:mo><mml:msub><mml:mo>)</mml:mo><mml:mi mathvariant="normal">e</mml:mi></mml:msub></mml:mrow></mml:math></inline-formula> contain the geometric element areas. Let <inline-formula><mml:math id="M130" display="inline"><mml:mi mathvariant="bold">P</mml:mi></mml:math></inline-formula> be the element-by-node averaging matrix, with <inline-formula><mml:math id="M131" display="inline"><mml:mrow><mml:msub><mml:mi>P</mml:mi><mml:mrow><mml:mi>e</mml:mi><mml:mi>j</mml:mi></mml:mrow></mml:msub><mml:mo>=</mml:mo><mml:mn mathvariant="normal">1</mml:mn><mml:mo>/</mml:mo><mml:mn mathvariant="normal">3</mml:mn></mml:mrow></mml:math></inline-formula> when node j belongs to triangle <inline-formula><mml:math id="M132" display="inline"><mml:mrow><mml:msub><mml:mi>K</mml:mi><mml:mi mathvariant="normal">e</mml:mi></mml:msub></mml:mrow></mml:math></inline-formula> and zero otherwise. The assembled nodal volume weights are <inline-formula><mml:math id="M133" display="inline"><mml:mrow><mml:mi mathvariant="bold-italic">w</mml:mi><mml:mo>=</mml:mo><mml:msup><mml:mi mathvariant="bold">P</mml:mi><mml:mi mathvariant="normal">T</mml:mi></mml:msup><mml:mi mathvariant="bold-italic">a</mml:mi></mml:mrow></mml:math></inline-formula>. For each basal node <inline-formula><mml:math id="M134" display="inline"><mml:mi>j</mml:mi></mml:math></inline-formula>, <inline-formula><mml:math id="M135" display="inline"><mml:mrow><mml:mo>(</mml:mo><mml:msub><mml:mi>w</mml:mi><mml:mi mathvariant="normal">b</mml:mi></mml:msub><mml:msub><mml:mo>)</mml:mo><mml:mi>j</mml:mi></mml:msub><mml:mo>=</mml:mo><mml:msub><mml:mo>∑</mml:mo><mml:mrow><mml:mi mathvariant="italic">γ</mml:mi><mml:mo>∋</mml:mo><mml:mi>j</mml:mi></mml:mrow></mml:msub><mml:msub><mml:mi mathvariant="normal">ℓ</mml:mi><mml:mi mathvariant="italic">γ</mml:mi></mml:msub><mml:mo>/</mml:mo><mml:mn mathvariant="normal">2</mml:mn></mml:mrow></mml:math></inline-formula> sums incident basal half-edge lengths; <inline-formula><mml:math id="M136" display="inline"><mml:mrow><mml:msub><mml:mi mathvariant="bold-italic">w</mml:mi><mml:mi mathvariant="normal">b</mml:mi></mml:msub></mml:mrow></mml:math></inline-formula> is the nodal <inline-formula><mml:math id="M137" display="inline"><mml:mrow><mml:mi mathvariant="bold-italic">d</mml:mi><mml:mi mathvariant="bold-italic">γ</mml:mi></mml:mrow></mml:math></inline-formula> used in Eq. (14). For nodal coefficient vectors <inline-formula><mml:math id="M138" display="inline"><mml:mrow><mml:mi mathvariant="bold-italic">η</mml:mi><mml:mo>(</mml:mo><mml:mi mathvariant="bold-italic">T</mml:mi><mml:mo>)</mml:mo></mml:mrow></mml:math></inline-formula> and <inline-formula><mml:math id="M139" display="inline"><mml:mi mathvariant="bold-italic">k</mml:mi></mml:math></inline-formula>, set <inline-formula><mml:math id="M140" display="inline"><mml:mrow><mml:msub><mml:mi mathvariant="bold-italic">η</mml:mi><mml:mi mathvariant="normal">e</mml:mi></mml:msub><mml:mo>(</mml:mo><mml:mi mathvariant="bold-italic">T</mml:mi><mml:mo>)</mml:mo><mml:mo>=</mml:mo><mml:mi mathvariant="bold">P</mml:mi><mml:mi mathvariant="bold-italic">η</mml:mi><mml:mo>(</mml:mo><mml:mi mathvariant="bold-italic">T</mml:mi><mml:mo>)</mml:mo></mml:mrow></mml:math></inline-formula> and <inline-formula><mml:math id="M141" display="inline"><mml:mrow><mml:msub><mml:mi mathvariant="bold-italic">k</mml:mi><mml:mi mathvariant="normal">e</mml:mi></mml:msub><mml:mo>=</mml:mo><mml:mi mathvariant="bold">P</mml:mi><mml:mi mathvariant="bold-italic">k</mml:mi></mml:mrow></mml:math></inline-formula>; <inline-formula><mml:math id="M142" display="inline"><mml:mrow><mml:msub><mml:mi mathvariant="italic">η</mml:mi><mml:mi>K</mml:mi></mml:msub><mml:mo>(</mml:mo><mml:mi>T</mml:mi><mml:mo>)</mml:mo></mml:mrow></mml:math></inline-formula> in Eq. (16) is the corresponding element coefficient. The nodal vector <inline-formula><mml:math id="M143" display="inline"><mml:mi mathvariant="bold-italic">f</mml:mi></mml:math></inline-formula> has entries <inline-formula><mml:math id="M144" display="inline"><mml:mrow><mml:mi mathvariant="italic">ρ</mml:mi><mml:mi>g</mml:mi><mml:mi>sin⁡</mml:mi><mml:mi mathvariant="italic">α</mml:mi></mml:mrow></mml:math></inline-formula>, and <inline-formula><mml:math id="M145" display="inline"><mml:mrow><mml:msub><mml:mi mathvariant="bold-italic">τ</mml:mi><mml:mi>E</mml:mi></mml:msub><mml:mo>(</mml:mo><mml:mi mathvariant="bold-italic">u</mml:mi><mml:mo>)</mml:mo></mml:mrow></mml:math></inline-formula> and <inline-formula><mml:math id="M146" display="inline"><mml:mrow><mml:msub><mml:mover accent="true"><mml:mi mathvariant="bold-italic">ϵ</mml:mi><mml:mo mathvariant="normal">˙</mml:mo></mml:mover><mml:mi>E</mml:mi></mml:msub><mml:mo>(</mml:mo><mml:mi mathvariant="bold-italic">u</mml:mi><mml:mo>)</mml:mo></mml:mrow></mml:math></inline-formula> are element vectors. Equation (25) is the constant-<inline-formula><mml:math id="M147" display="inline"><mml:mi mathvariant="italic">η</mml:mi></mml:math></inline-formula> form; Eq. (26) uses <inline-formula><mml:math id="M148" display="inline"><mml:mrow><mml:msub><mml:mi mathvariant="bold-italic">η</mml:mi><mml:mi mathvariant="normal">e</mml:mi></mml:msub><mml:mo>(</mml:mo><mml:mi mathvariant="bold-italic">T</mml:mi><mml:mo>)</mml:mo></mml:mrow></mml:math></inline-formula>. With element-wise multiplication, <inline-formula><mml:math id="M149" display="inline"><mml:mo>∘</mml:mo></mml:math></inline-formula>, boundary nodal values, <inline-formula><mml:math id="M150" display="inline"><mml:mrow><mml:msub><mml:mo>⋅</mml:mo><mml:mi mathvariant="normal">b</mml:mi></mml:msub></mml:mrow></mml:math></inline-formula>, and second-dimension vector-wise norm, <inline-formula><mml:math id="M151" display="inline"><mml:mrow><mml:mfenced open="|" close="|"><mml:mo>⋅</mml:mo></mml:mfenced></mml:mrow></mml:math></inline-formula>. Therefore, the fully discretized energy functional is a convex function on the vectorized speed, <inline-formula><mml:math id="M152" display="inline"><mml:mi mathvariant="bold-italic">v</mml:mi></mml:math></inline-formula>.</p>
      <p id="d2e4192">The resulting vectorized, fully discretized energy becomes

            <disp-formula id="Ch1.E26" content-type="numbered"><label>26</label><mml:math id="M153" display="block"><mml:mrow><mml:mtable rowspacing="0.2ex" class="split" displaystyle="true" columnalign="right left"><mml:mtr><mml:mtd><mml:mrow><mml:mi>J</mml:mi><mml:mo>(</mml:mo><mml:mi mathvariant="bold-italic">u</mml:mi><mml:mo>)</mml:mo></mml:mrow></mml:mtd><mml:mtd><mml:mrow><mml:mo>=</mml:mo><mml:mstyle displaystyle="true"><mml:mfrac style="display"><mml:mn mathvariant="normal">1</mml:mn><mml:mi>p</mml:mi></mml:mfrac></mml:mstyle><mml:msup><mml:mfenced close="]" open="["><mml:mrow><mml:msub><mml:mi mathvariant="bold-italic">η</mml:mi><mml:mi mathvariant="normal">e</mml:mi></mml:msub><mml:mo>(</mml:mo><mml:mi mathvariant="bold-italic">T</mml:mi><mml:mo>)</mml:mo><mml:mo>∘</mml:mo><mml:mi mathvariant="bold-italic">a</mml:mi></mml:mrow></mml:mfenced><mml:mi mathvariant="normal">T</mml:mi></mml:msup><mml:msup><mml:mfenced open="|" close="|"><mml:mrow><mml:mo>[</mml:mo><mml:msub><mml:mi mathvariant="bold">D</mml:mi><mml:mi>x</mml:mi></mml:msub><mml:mi mathvariant="bold-italic">u</mml:mi><mml:mo>,</mml:mo><mml:msub><mml:mi mathvariant="bold">D</mml:mi><mml:mi>y</mml:mi></mml:msub><mml:mi mathvariant="bold-italic">u</mml:mi><mml:mo>]</mml:mo></mml:mrow></mml:mfenced><mml:mi>p</mml:mi></mml:msup><mml:mo>+</mml:mo><mml:msup><mml:mfenced open="[" close="]"><mml:mrow><mml:msub><mml:mi mathvariant="bold-italic">w</mml:mi><mml:mi mathvariant="normal">b</mml:mi></mml:msub><mml:mo>∘</mml:mo><mml:msub><mml:mi mathvariant="bold-italic">q</mml:mi><mml:mi mathvariant="normal">b</mml:mi></mml:msub><mml:mo>(</mml:mo><mml:msub><mml:mi mathvariant="bold-italic">u</mml:mi><mml:mi mathvariant="normal">b</mml:mi></mml:msub><mml:mo>)</mml:mo></mml:mrow></mml:mfenced><mml:mi mathvariant="normal">T</mml:mi></mml:msup><mml:msub><mml:mi mathvariant="bold-italic">u</mml:mi><mml:mi mathvariant="normal">b</mml:mi></mml:msub></mml:mrow></mml:mtd></mml:mtr><mml:mtr><mml:mtd/><mml:mtd><mml:mrow><mml:mo>-</mml:mo><mml:mo>(</mml:mo><mml:mi mathvariant="bold-italic">w</mml:mi><mml:mo>∘</mml:mo><mml:mi mathvariant="bold-italic">f</mml:mi><mml:msup><mml:mo>)</mml:mo><mml:mi mathvariant="normal">T</mml:mi></mml:msup><mml:mi mathvariant="bold-italic">u</mml:mi><mml:mo>.</mml:mo></mml:mrow></mml:mtd></mml:mtr></mml:mtable></mml:mrow></mml:math></disp-formula>

          with discretized gradient operators, <inline-formula><mml:math id="M154" display="inline"><mml:mrow><mml:msub><mml:mi mathvariant="bold">D</mml:mi><mml:mi>x</mml:mi></mml:msub></mml:mrow></mml:math></inline-formula> and <inline-formula><mml:math id="M155" display="inline"><mml:mrow><mml:msub><mml:mi mathvariant="bold">D</mml:mi><mml:mi>y</mml:mi></mml:msub></mml:mrow></mml:math></inline-formula>, element-wise multiplication, <inline-formula><mml:math id="M156" display="inline"><mml:mo>∘</mml:mo></mml:math></inline-formula>, boundary nodal values, <inline-formula><mml:math id="M157" display="inline"><mml:mrow><mml:msub><mml:mo>⋅</mml:mo><mml:mi mathvariant="normal">b</mml:mi></mml:msub></mml:mrow></mml:math></inline-formula>, and second-dimension vector-wise norm, <inline-formula><mml:math id="M158" display="inline"><mml:mrow><mml:mfenced close="|" open="|"><mml:mo>⋅</mml:mo></mml:mfenced></mml:mrow></mml:math></inline-formula>, where we have substituted the minimizer <inline-formula><mml:math id="M159" display="inline"><mml:mrow><mml:mi mathvariant="bold-italic">v</mml:mi><mml:mo>=</mml:mo><mml:mi mathvariant="bold-italic">u</mml:mi></mml:mrow></mml:math></inline-formula>. The fully discretized energy functional is a convex function on the vectorized speed.</p>
      <p id="d2e4388">A similar process applied to the thermal minimization problem in Eq. (<xref ref-type="disp-formula" rid="Ch1.E13"/>), results in a  vectorized, fully discretized energy,

            <disp-formula id="Ch1.E27" content-type="numbered"><label>27</label><mml:math id="M160" display="block"><mml:mtable rowspacing="0.2ex" class="split" displaystyle="true" columnalign="right left"><mml:mtr><mml:mtd><mml:mrow><mml:mi>L</mml:mi><mml:mo>(</mml:mo><mml:mi mathvariant="bold-italic">T</mml:mi><mml:mo>)</mml:mo></mml:mrow></mml:mtd><mml:mtd><mml:mrow><mml:mo>=</mml:mo><mml:mstyle displaystyle="true"><mml:mfrac style="display"><mml:mn mathvariant="normal">1</mml:mn><mml:mn mathvariant="normal">2</mml:mn></mml:mfrac></mml:mstyle><mml:mo>(</mml:mo><mml:msub><mml:mi mathvariant="bold-italic">k</mml:mi><mml:mi mathvariant="normal">e</mml:mi></mml:msub><mml:mo>∘</mml:mo><mml:mi mathvariant="bold-italic">a</mml:mi><mml:msup><mml:mo>)</mml:mo><mml:mi mathvariant="normal">T</mml:mi></mml:msup><mml:msup><mml:mfenced open="|" close="|"><mml:mrow><mml:mo>[</mml:mo><mml:msub><mml:mi mathvariant="bold">D</mml:mi><mml:mi>x</mml:mi></mml:msub><mml:mi mathvariant="bold-italic">T</mml:mi><mml:mo>,</mml:mo><mml:msub><mml:mi mathvariant="bold">D</mml:mi><mml:mi>y</mml:mi></mml:msub><mml:mi mathvariant="bold-italic">T</mml:mi><mml:mo>]</mml:mo></mml:mrow></mml:mfenced><mml:mn mathvariant="normal">2</mml:mn></mml:msup><mml:mo>-</mml:mo><mml:msub><mml:mi>G</mml:mi><mml:mi mathvariant="normal">geo</mml:mi></mml:msub><mml:msubsup><mml:mi mathvariant="bold-italic">w</mml:mi><mml:mi mathvariant="normal">b</mml:mi><mml:mi mathvariant="normal">T</mml:mi></mml:msubsup><mml:msub><mml:mi mathvariant="bold-italic">T</mml:mi><mml:mi mathvariant="normal">b</mml:mi></mml:msub></mml:mrow></mml:mtd></mml:mtr><mml:mtr><mml:mtd/><mml:mtd><mml:mrow><mml:mo>-</mml:mo><mml:mn mathvariant="normal">2</mml:mn><mml:msup><mml:mfenced close="]" open="["><mml:mrow><mml:msup><mml:mi mathvariant="bold">P</mml:mi><mml:mi mathvariant="normal">T</mml:mi></mml:msup><mml:mfenced close=")" open="("><mml:mrow><mml:mi mathvariant="bold-italic">a</mml:mi><mml:mo>∘</mml:mo><mml:msub><mml:mi mathvariant="bold-italic">τ</mml:mi><mml:mi>E</mml:mi></mml:msub><mml:mo>(</mml:mo><mml:mi mathvariant="bold-italic">u</mml:mi><mml:mo>)</mml:mo><mml:mo>∘</mml:mo><mml:msub><mml:mover accent="true"><mml:mi mathvariant="bold-italic">ϵ</mml:mi><mml:mo mathvariant="normal">˙</mml:mo></mml:mover><mml:mi>E</mml:mi></mml:msub><mml:mo>(</mml:mo><mml:mi mathvariant="bold-italic">u</mml:mi><mml:mo>)</mml:mo></mml:mrow></mml:mfenced></mml:mrow></mml:mfenced><mml:mi mathvariant="normal">T</mml:mi></mml:msup><mml:mi mathvariant="bold-italic">T</mml:mi><mml:mo>.</mml:mo></mml:mrow></mml:mtd></mml:mtr></mml:mtable></mml:math></disp-formula>

          We solve both minimization problems through the Disciplined Convex Programming software library, CVX <xref ref-type="bibr" rid="bib1.bibx21" id="paren.47"/>.</p>
      <p id="d2e4539">The governing equations are coupled through temperature-dependent ice viscosity, <inline-formula><mml:math id="M161" display="inline"><mml:mrow><mml:mi mathvariant="italic">η</mml:mi><mml:mo>(</mml:mo><mml:mi>T</mml:mi><mml:mo>)</mml:mo></mml:mrow></mml:math></inline-formula>, and viscous shear heating, <inline-formula><mml:math id="M162" display="inline"><mml:mrow><mml:mn mathvariant="normal">2</mml:mn><mml:msub><mml:mi mathvariant="italic">τ</mml:mi><mml:mi mathvariant="normal">E</mml:mi></mml:msub><mml:mo>(</mml:mo><mml:mi>u</mml:mi><mml:mo>)</mml:mo><mml:msub><mml:mover accent="true"><mml:mi mathvariant="italic">ϵ</mml:mi><mml:mo mathvariant="normal">˙</mml:mo></mml:mover><mml:mi mathvariant="normal">E</mml:mi></mml:msub><mml:mo>(</mml:mo><mml:mi>u</mml:mi><mml:mo>)</mml:mo></mml:mrow></mml:math></inline-formula>. We implement this coupling through an under-relaxed fixed-point iteration between the thermal and mechanical models, as illustrated in Fig. <xref ref-type="fig" rid="F3"/>. At each iteration, we relax the thermal solution by a relaxation parameter, <inline-formula><mml:math id="M163" display="inline"><mml:mrow><mml:mi mathvariant="italic">ω</mml:mi><mml:mo>∈</mml:mo><mml:mo>[</mml:mo><mml:mn mathvariant="normal">0</mml:mn><mml:mo>,</mml:mo><mml:mn mathvariant="normal">1</mml:mn><mml:mo>]</mml:mo></mml:mrow></mml:math></inline-formula>, and iterate until the residual falls below a threshold, <inline-formula><mml:math id="M164" display="inline"><mml:mi mathvariant="italic">ϵ</mml:mi></mml:math></inline-formula>. The result of this coupling is an antiplane speed <inline-formula><mml:math id="M165" display="inline"><mml:mi>u</mml:mi></mml:math></inline-formula>, and a temperature field <inline-formula><mml:math id="M166" display="inline"><mml:mi>T</mml:mi></mml:math></inline-formula>, which are fully consistent.</p>

      <fig id="F3"><label>Figure 3</label><caption><p id="d2e4636">Flowchart of our thermomechanical coupling scheme. We solve the mechanical and thermal free-boundary problems sequentially using CVX <xref ref-type="bibr" rid="bib1.bibx21" id="paren.48"/> and couple them through an under-relaxed fixed-point iteration of temperature. The temperature indices are schematic: convergence is checked after the thermal solve, and the relaxed temperature is used in the next mechanical solve if needed. The thermal minimization enforces <inline-formula><mml:math id="M167" display="inline"><mml:mrow><mml:mi>T</mml:mi><mml:mi mathvariant="normal">≤</mml:mi><mml:msub><mml:mi>T</mml:mi><mml:mi mathvariant="normal">melt</mml:mi></mml:msub></mml:mrow></mml:math></inline-formula>; the displayed diffusion equation applies where <inline-formula><mml:math id="M168" display="inline"><mml:mrow><mml:mi>T</mml:mi><mml:mo>&lt;</mml:mo><mml:msub><mml:mi>T</mml:mi><mml:mi mathvariant="normal">melt</mml:mi></mml:msub></mml:mrow></mml:math></inline-formula>. We iterate until the temperature residual falls below the prescribed tolerance, <inline-formula><mml:math id="M169" display="inline"><mml:mi mathvariant="italic">ε</mml:mi></mml:math></inline-formula>.</p></caption>
          <graphic xlink:href="https://gmd.copernicus.org/articles/19/11041/2026/gmd-19-11041-2026-f03.png"/>

        </fig>

</sec>
<sec id="Ch1.S2.SS3">
  <label>2.3</label><title>Verification</title>
      <p id="d2e4693">To verify the accuracy of our implementation of the mechanical model, we benchmark against the analytical solution derived for an idealized valley glacier in <xref ref-type="bibr" rid="bib1.bibx65" id="text.49"/>. This test case consists of a simplified glacier valley geometry underlain by saturated sediments, with variable pore pressure resulting from a specified elevation of the water table. Different failure regions become apparent when the elevation of the water table and the driving stress are varied. Figure <xref ref-type="fig" rid="F4"/>a shows the model setup of a triangular computational domain, with the elevation of the water table denoted by a dashed line and the failure patches shown in red. In Fig. <xref ref-type="fig" rid="F4"/>a, we reproduce Fig. 5 from <xref ref-type="bibr" rid="bib1.bibx65" id="text.50"/> with the developed method by simulating the four distinct failure regimes. We demonstrate that our solver using a nominal grid spacing of <inline-formula><mml:math id="M170" display="inline"><mml:mrow><mml:mi mathvariant="normal">Δ</mml:mi><mml:mi>x</mml:mi><mml:mo>=</mml:mo><mml:mn mathvariant="normal">0.02</mml:mn></mml:mrow></mml:math></inline-formula> reproduces the velocity field and failure patches of <xref ref-type="bibr" rid="bib1.bibx65" id="text.51"/>. The small deviations between our numerical solution and the benchmark arise from our finite discretization of the domain.</p>

      <fig id="F4" specific-use="star"><label>Figure 4</label><caption><p id="d2e4726">Verification of our numerical implementation of the mechanical model against <xref ref-type="bibr" rid="bib1.bibx65" id="text.52"/>. <bold>(a)</bold> Contours of flow speed for an idealized valley glacier as specified in <xref ref-type="bibr" rid="bib1.bibx65" id="text.53"/>. Mimicking Fig. 5 in <xref ref-type="bibr" rid="bib1.bibx65" id="text.54"/>, the dashed line signifies the water table, with each flow regime chosen to exhibit markedly different regions of till failure (red lines). <bold>(b)</bold> Convergence plots of the normalized <inline-formula><mml:math id="M171" display="inline"><mml:mrow><mml:msub><mml:mi>L</mml:mi><mml:mn mathvariant="normal">2</mml:mn></mml:msub></mml:mrow></mml:math></inline-formula>-error on surface velocities, computed against the most refined solution with the four curves representing the four benchmark examples shown in <bold>(a)</bold>. We observe the expected second-order convergence of linear finite elements.</p></caption>
          <graphic xlink:href="https://gmd.copernicus.org/articles/19/11041/2026/gmd-19-11041-2026-f04.png"/>

        </fig>

      <p id="d2e4765">We perform a numerical convergence analysis of Ortholine for each of these benchmark solutions, shown in Fig. <xref ref-type="fig" rid="F4"/>b. This is a self-convergence study: we compute the normalized <inline-formula><mml:math id="M172" display="inline"><mml:mrow><mml:msub><mml:mi>L</mml:mi><mml:mn mathvariant="normal">2</mml:mn></mml:msub></mml:mrow></mml:math></inline-formula>-error of the surface velocities against our own most refined simulation as the grid is refined, which isolates the numerical discretization error and confirms that the solver converges at the expected rate. The error exhibits second-order convergence with grid refinement. The linear finite elements have polynomial degree <inline-formula><mml:math id="M173" display="inline"><mml:mrow><mml:mi>p</mml:mi><mml:mo>=</mml:mo><mml:mn mathvariant="normal">1</mml:mn></mml:mrow></mml:math></inline-formula> and an expected <inline-formula><mml:math id="M174" display="inline"><mml:mrow><mml:mi>p</mml:mi><mml:mo>+</mml:mo><mml:mn mathvariant="normal">1</mml:mn></mml:mrow></mml:math></inline-formula>, or second-order, convergence rate in the <inline-formula><mml:math id="M175" display="inline"><mml:mrow><mml:msup><mml:mi>L</mml:mi><mml:mn mathvariant="normal">2</mml:mn></mml:msup></mml:mrow></mml:math></inline-formula> norm <xref ref-type="bibr" rid="bib1.bibx7" id="paren.55"/>. The observed second-order numerical convergence makes us confident that our implementation of the numerical method is correct and consistent. We are not aware of a verification case of a thermomechanically coupled free-boundary problem that would allow us to perform a similar analysis for thermally variable ice.</p>
</sec>
</sec>
<sec id="Ch1.S3">
  <label>3</label><title>Application of Ortholine to a specific field site</title>
      <p id="d2e4829">Ortholine is intended as a general framework that can in principle be applied to any field site, where the properties of the shear margins are of specific interest. However, the model has more potential to lead to valuable scientific insights at field sites where the shear margin is only partially controlled by topography, where subglacial conditions are thought to vary, and where field observations are available to constrain bed conditions. Figure <xref ref-type="fig" rid="F5"/> outlines the workflow for applying Ortholine to a given field site. We illustrate these steps below for the example of Institute Ice Stream located in the Ronne Ice Shelf region of West Antarctica.</p>

      <fig id="F5" specific-use="star"><label>Figure 5</label><caption><p id="d2e4836">Flowchart for the application of Ortholine to a given field site. Each of the four main steps is discussed in detail in Sect. <xref ref-type="sec" rid="Ch1.S3"/>. The only necessary data set for setting up the model is surface-velocity data (labeled “speed data” in the diagram), typically derived from satellite-based measurements. These data provide both direction and magnitude for computing ortho-flow lines. Other data can be valuable to integrate (blue arrow), but only surface-velocity data are necessary (orange arrow).</p></caption>
        <graphic xlink:href="https://gmd.copernicus.org/articles/19/11041/2026/gmd-19-11041-2026-f05.png"/>

      </fig>

<sec id="Ch1.S3.SS1">
  <label>3.1</label><title>Compute ortho-flow lines</title>
      <p id="d2e4854">The classical conceptualization of ice streams by <xref ref-type="bibr" rid="bib1.bibx53" id="text.56"/> posits a plug-like motion in the trunk of the ice stream bordered by two narrow shear margins on the sides, yielding a flow configuration that is symmetric around the center line of the stream. In that limit, a simple linear transect that is orthogonal to the main flow direction and cuts through both shear margins is in a state of antiplane shear, as assumed in many idealized shear-margin models <xref ref-type="bibr" rid="bib1.bibx33 bib1.bibx72 bib1.bibx50 bib1.bibx26" id="paren.57"/>. Although there are examples of ice streams that fit this original conceptualization, such as Kamb Ice Stream <xref ref-type="bibr" rid="bib1.bibx13" id="paren.58"/>, the majority of field sites will not.</p>
      <p id="d2e4866">To analyze shear-margin properties in the field, the first step is to compute the curves that represent the closest equivalent to the simple cross-sectional transect envisioned by <xref ref-type="bibr" rid="bib1.bibx53" id="text.59"/>. Because the antiplane approximation requires the model plane to be orthogonal to the local ice-flow direction (see Sect. 1), these curves are not in general straight transects but follow the local flow geometry. We refer to this set of curves as ortho-flow lines to highlight that their orientation is orthogonal to the local direction of ice velocity, For irrotational flow, these lines correspond to isocontours of the velocity potential, <inline-formula><mml:math id="M176" display="inline"><mml:mrow><mml:mi mathvariant="italic">φ</mml:mi><mml:mo>(</mml:mo><mml:mi mathvariant="normal">∇</mml:mi><mml:mi mathvariant="italic">φ</mml:mi><mml:mo>=</mml:mo><mml:msub><mml:mi mathvariant="bold-italic">u</mml:mi><mml:mi>d</mml:mi></mml:msub><mml:mo>)</mml:mo></mml:mrow></mml:math></inline-formula>. We numerically generate ortho-flow lines from the velocity field derived from existing satellite data <xref ref-type="bibr" rid="bib1.bibx55" id="paren.60"/>. The velocity field need not satisfy the irrotationality condition <inline-formula><mml:math id="M177" display="inline"><mml:mrow><mml:mi mathvariant="normal">∇</mml:mi><mml:mo>×</mml:mo><mml:msub><mml:mi mathvariant="bold-italic">u</mml:mi><mml:mi>d</mml:mi></mml:msub><mml:mo>=</mml:mo><mml:mn mathvariant="normal">0</mml:mn></mml:mrow></mml:math></inline-formula>: physical shear as well as observational noise can produce rotational flow. If the flow is not irrotational, we cannot directly solve for the velocity potential. Instead, we compute ortho-flow lines through numerical integration starting from a seed point in the center of the ice stream, <inline-formula><mml:math id="M178" display="inline"><mml:mrow><mml:msub><mml:mi>x</mml:mi><mml:mn mathvariant="normal">0</mml:mn></mml:msub></mml:mrow></mml:math></inline-formula>. Velocity vectors are perpendicular to the ortho-flow line, which we generate from the stream center outward through Euler integration,

            <disp-formula id="Ch1.E28" content-type="numbered"><label>28</label><mml:math id="M179" display="block"><mml:mrow><mml:msub><mml:mi mathvariant="bold-italic">x</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:msub><mml:mi mathvariant="bold-italic">x</mml:mi><mml:mi>i</mml:mi></mml:msub><mml:mo>+</mml:mo><mml:mi mathvariant="normal">Δ</mml:mi><mml:mi>x</mml:mi><mml:mstyle displaystyle="true"><mml:mfrac style="display"><mml:mrow><mml:msubsup><mml:mi mathvariant="bold-italic">u</mml:mi><mml:mi>d</mml:mi><mml:mo>⟂</mml:mo></mml:msubsup><mml:mo>(</mml:mo><mml:msub><mml:mi>x</mml:mi><mml:mi>i</mml:mi></mml:msub><mml:mo>)</mml:mo></mml:mrow><mml:mrow><mml:mo>|</mml:mo><mml:msubsup><mml:mi mathvariant="bold-italic">u</mml:mi><mml:mi>d</mml:mi><mml:mo>⟂</mml:mo></mml:msubsup><mml:mo>(</mml:mo><mml:msub><mml:mi>x</mml:mi><mml:mi>i</mml:mi></mml:msub><mml:mo>)</mml:mo><mml:mo>|</mml:mo></mml:mrow></mml:mfrac></mml:mstyle></mml:mrow></mml:math></disp-formula>

          with step size, <inline-formula><mml:math id="M180" display="inline"><mml:mrow><mml:mi mathvariant="normal">Δ</mml:mi><mml:mi>x</mml:mi></mml:mrow></mml:math></inline-formula>. The resolution of the computation is smaller than the resolution of the observed velocity field, <inline-formula><mml:math id="M181" display="inline"><mml:mrow><mml:msub><mml:mi mathvariant="bold-italic">u</mml:mi><mml:mi>d</mml:mi></mml:msub></mml:mrow></mml:math></inline-formula>. This procedure generates an ortho-flow line, <inline-formula><mml:math id="M182" display="inline"><mml:mi mathvariant="bold-italic">x</mml:mi></mml:math></inline-formula>, with a maximal antiplane velocity component and a near-zero in-plane velocity component. When the downstream ice speed drops below 10 m yr<sup>−1</sup>, this approach becomes noisy and we extend the ortho-flow line linearly.</p>
      <p id="d2e5038">Figure <xref ref-type="fig" rid="F6"/>a shows a computed ortho-flow line cutting across the trunk of Institute Ice Stream as a gray line. We emphasize that this ortho-flow line is one of many possible ones. Our ortho-flow line is approximately linear near the center line of the stream, highlighting that Institute partially abides by the original conceptualization of ice streams by <xref ref-type="bibr" rid="bib1.bibx53" id="text.61"/>. However, as the ortho-flow line approaches the margins of the ice stream, it increasingly curves to partially align with the main flow direction, particularly near the western shear margin where we intentionally extend it until ice speed drops below 10 m yr<sup>−1</sup>. The pronounced curvature of the ortho-flow line in the western shear margin is the consequence of lateral influx of ice through several smaller tributaries visible in Fig. <xref ref-type="fig" rid="F6"/>a. It highlights that the ortho-flow line A–A' is dynamically coupled to stresses tens of kilometers further downstream through an extended marginal zone highlighted in Fig. <xref ref-type="fig" rid="F6"/>c.</p>

      <fig id="F6" specific-use="star"><label>Figure 6</label><caption><p id="d2e5065">Application of Ortholine to Institute Ice Stream. <bold>(a)</bold> Map of the ice speed in the study area derived from satellite radar data <xref ref-type="bibr" rid="bib1.bibx55" id="paren.62"/> and generated with the Antarctic Mapping Toolbox <xref ref-type="bibr" rid="bib1.bibx23" id="paren.63"/>. The flow direction is highlighted as a thick, transparent arrow. The shear margins are the distinct lateral transitions from fast (red) to slow flow (blue). The inferred rock-sediment transition from <xref ref-type="bibr" rid="bib1.bibx69" id="text.64"/> is indicated as a white line. We denote the main features of the subglacial hydrology in blue, including the active subglacial lake E1 <xref ref-type="bibr" rid="bib1.bibx71" id="paren.65"/>, hydropotential flow paths <xref ref-type="bibr" rid="bib1.bibx69" id="paren.66"/>, and a meltwater channel carved into the ice shelf <xref ref-type="bibr" rid="bib1.bibx40" id="paren.67"/>. Our ortho-flow line is indicated in gray and extends from A to A'. <bold>(b)</bold> Radar cross-section along C–C' from <xref ref-type="bibr" rid="bib1.bibx69" id="text.68"/>, showing basal topography and inferred subglacial conditions. We superimpose the downstream ice speed as an orange line. Inferred hydropotential gradients <xref ref-type="bibr" rid="bib1.bibx69" id="paren.69"/> suggest that water flow propagates from the shear margin on top of sediment to the shear margin on hard rock. We choose to focus on the ortho-flow line A–A' because of its proximity to the radar cross-section C–C' from Siegert et al. (2016). <bold>(c)</bold> Observed surface-speed profile along A–A', showing the extended marginal zone and the western and eastern shear margins.</p></caption>
          <graphic xlink:href="https://gmd.copernicus.org/articles/19/11041/2026/gmd-19-11041-2026-f06.png"/>

        </fig>

</sec>
<sec id="Ch1.S3.SS2">
  <label>3.2</label><title>Set up the model domain</title>
      <p id="d2e5117">To set up the Ortholine model domain, we chose the ortho-flow line A–A'. In principle, Ortholine could represent any ortho-flow line, but some ortho-flow lines might be more suitable than others. For example, it might not be ideal to select an ortho-flow line that crosses the grounding line, because grounding-line processes are not currently included in the model. Also, it can be valuable to select an ortho-flow line where data from previous field campaigns can shed additional light on basal processes. While only a surface-velocity data set is strictly necessary for setting up Ortholine, other constraints could elevate the scientific contribution of the modeling. We illustrate this process through a short literature review for Institute Ice Stream.</p>
      <p id="d2e5120">At Institute, streaming ice is bounded to the east by Bungenstock Ice Rise, and fed from the west by tributaries from the Ellsworth Mountains <xref ref-type="bibr" rid="bib1.bibx56 bib1.bibx58" id="paren.70"/>. As evident from the ice-speed map in Fig. <xref ref-type="fig" rid="F6"/>a, the eastern shear margin near the Bungenstock Ice Rise is very sharp, but the western shear margin is much less localized <xref ref-type="bibr" rid="bib1.bibx61" id="paren.71"/>. This asymmetry on the surface is mirrored by different subglacial conditions at the base of the ice, with the eastern shear margin underlain by sediments and the western shear margin underlain by a hard-rock bed, as inferred from the radar profile C–C' shown in Fig. <xref ref-type="fig" rid="F6"/>b <xref ref-type="bibr" rid="bib1.bibx69" id="paren.72"/>. The basal topography across this profile is relatively subdued, making it unlikely that topographic control is the primary factor determining the shear margin position.</p>
      <p id="d2e5136">In the 2010–2011 field season, the British Antarctic Survey completed extensive airborne geophysical surveys of the area, including radar sounding, gravity, magnetics, and LIDAR <xref ref-type="bibr" rid="bib1.bibx35" id="paren.73"/>. These geophysical data constrain the various factors that contribute to the dynamics of Institute Ice Stream, including topography, lithology, and hydrology. <xref ref-type="bibr" rid="bib1.bibx69" id="text.74"/> highlight subglacial hydrology as an important factor that could induce rapid rearrangement in Institute's trunk. The trunk is downstream of an active subglacial lake, Institute E1 <xref ref-type="bibr" rid="bib1.bibx71" id="paren.75"/> highlighted in Fig. <xref ref-type="fig" rid="F6"/>a, and meltwater emanates from the western shear margin at the grounding line <xref ref-type="bibr" rid="bib1.bibx40 bib1.bibx1" id="paren.76"/> in the form of a channel. The hydraulic potential of Institute Ice Stream suggests that subglacial meltwater is routed from the eastern shear margin to the western shear margin <xref ref-type="bibr" rid="bib1.bibx69" id="paren.77"/>. The white line in Fig. <xref ref-type="fig" rid="F6"/>a represents a transition between subglacial sediments and hard rock, as inferred by <xref ref-type="bibr" rid="bib1.bibx69" id="text.78"/> from radar sounding data.</p>
      <p id="d2e5162">For the remainder of this section, we choose to focus on ortho-flow line A–A', because of its proximity to the radar profile C–C' <xref ref-type="bibr" rid="bib1.bibx69" id="paren.79"/>. At this location, the main ice stream trunk is approximately 60 km wide, with a maximum ice thickness of about 1700 m near the western margin. We simplify the geometry of the actual bed profile by assuming a model domain with a flat surface and a bed that is piecewise-linear, neglecting roughness. For the characteristic driving stress of the region we use a representative ice thickness, <inline-formula><mml:math id="M185" display="inline"><mml:mrow><mml:mi>H</mml:mi><mml:mo>=</mml:mo><mml:mn mathvariant="normal">1275</mml:mn><mml:mspace linebreak="nobreak" width="0.125em"/><mml:mtext>m</mml:mtext></mml:mrow></mml:math></inline-formula>, giving <inline-formula><mml:math id="M186" display="inline"><mml:mrow><mml:msub><mml:mi mathvariant="italic">τ</mml:mi><mml:mi>d</mml:mi></mml:msub><mml:mo>=</mml:mo><mml:mi mathvariant="italic">ρ</mml:mi><mml:mi>g</mml:mi><mml:mi>H</mml:mi><mml:mi>sin⁡</mml:mi><mml:mi mathvariant="italic">α</mml:mi><mml:mo>=</mml:mo><mml:mn mathvariant="normal">27.5</mml:mn><mml:mspace linebreak="nobreak" width="0.125em"/><mml:mtext>kPa</mml:mtext></mml:mrow></mml:math></inline-formula>, with ice density, <inline-formula><mml:math id="M187" display="inline"><mml:mrow><mml:mi mathvariant="italic">ρ</mml:mi><mml:mo>=</mml:mo><mml:mn mathvariant="normal">917</mml:mn></mml:mrow></mml:math></inline-formula> kg m<sup>−3</sup>, acceleration due to gravity, <inline-formula><mml:math id="M189" display="inline"><mml:mrow><mml:mi>g</mml:mi><mml:mo>=</mml:mo><mml:mn mathvariant="normal">9.8</mml:mn></mml:mrow></mml:math></inline-formula> m s<sup>−2</sup>, and surface slope, <inline-formula><mml:math id="M191" display="inline"><mml:mrow><mml:mi mathvariant="italic">α</mml:mi><mml:mo>=</mml:mo><mml:mn mathvariant="normal">0.0024</mml:mn></mml:mrow></mml:math></inline-formula>. The full thickness variation across the trunk (from approximately 1000 m near the eastern to about 1700 m near the western margin) enters the overburden-dependent bed strength through Eq. (<xref ref-type="disp-formula" rid="Ch1.E31"/>).</p>
      <p id="d2e5277">Based on the interpretation of radio-echo sounding data, <xref ref-type="bibr" rid="bib1.bibx69" id="text.80"/> suggest that the lithology underneath Institute transitions from sediment to hard rock in the across-flow direction (see Fig. <xref ref-type="fig" rid="F6"/>b). Both the bed roughness and the return power of the radio-echo sound provide evidence in favor of this interpretation <xref ref-type="bibr" rid="bib1.bibx69" id="paren.81"/>. We separate the bed into two boundary conditions specified by regions identified by <xref ref-type="bibr" rid="bib1.bibx69" id="text.82"/> as hard rock and saturated sediments, respectively. Within the sediment-occupied zone, we impose a sharp increase in the prescribed till strength <inline-formula><mml:math id="M192" display="inline"><mml:mrow><mml:msub><mml:mi mathvariant="italic">τ</mml:mi><mml:mi>c</mml:mi></mml:msub></mml:mrow></mml:math></inline-formula> outside of the eastern shear margin to represent a transition from temperate to frozen conditions, as inferred by <xref ref-type="bibr" rid="bib1.bibx69" id="text.83"/>. Rather than raising the total strength <inline-formula><mml:math id="M193" display="inline"><mml:mrow><mml:msub><mml:mi mathvariant="italic">τ</mml:mi><mml:mi>c</mml:mi></mml:msub></mml:mrow></mml:math></inline-formula> directly, one could equivalently adjust individual parameters such as the cohesion <inline-formula><mml:math id="M194" display="inline"><mml:mi>c</mml:mi></mml:math></inline-formula> where independent information exists for how they change under frozen conditions.</p>
      <p id="d2e5324">The trunk of Institute Ice Stream (see Fig. <xref ref-type="fig" rid="F6"/>a) is underlain by a complex subglacial drainage system involving a lake upstream <xref ref-type="bibr" rid="bib1.bibx71" id="paren.84"/>, lateral meltwater drainage in the main trunk, and a meltwater channel emanating from the grounding line in the vicinity of the western shear margin <xref ref-type="bibr" rid="bib1.bibx40" id="paren.85"/>. We assume that distributed drainage can be approximated as a thin film of meltwater and that efficient drainage occurs through a Röthlisberger channel <xref ref-type="bibr" rid="bib1.bibx59" id="paren.86"/> operating at low pore pressure <xref ref-type="bibr" rid="bib1.bibx29 bib1.bibx50 bib1.bibx13" id="paren.87"/>. The reduction in pore pressure in the vicinity of the channel depends on the channel discharge and the decay length of the pressure perturbation scales with the film thickness of the distributed system, as derived in more detail in <xref ref-type="bibr" rid="bib1.bibx50" id="text.88"/>.</p>
      <p id="d2e5345">We model the decline of bed strength away from a subglacial drainage channel located at <inline-formula><mml:math id="M195" display="inline"><mml:mi>Y</mml:mi></mml:math></inline-formula> using a localized pore-pressure deficit, <inline-formula><mml:math id="M196" display="inline"><mml:mrow><mml:mi>p</mml:mi><mml:mo>(</mml:mo><mml:mi>y</mml:mi><mml:mo>)</mml:mo><mml:mo>=</mml:mo><mml:msub><mml:mi>p</mml:mi><mml:mi mathvariant="normal">∞</mml:mi></mml:msub><mml:mo>-</mml:mo><mml:mi mathvariant="normal">Δ</mml:mi><mml:mi>p</mml:mi><mml:mi>exp⁡</mml:mi><mml:mspace width="-0.125em" linebreak="nobreak"/><mml:mfenced open="(" close=")"><mml:mrow><mml:mo>-</mml:mo><mml:mstyle displaystyle="false"><mml:mfrac style="text"><mml:mrow><mml:mo>|</mml:mo><mml:mi>y</mml:mi><mml:mo>-</mml:mo><mml:mi>Y</mml:mi><mml:mo>|</mml:mo></mml:mrow><mml:mi mathvariant="italic">γ</mml:mi></mml:mfrac></mml:mstyle></mml:mrow></mml:mfenced><mml:mo>,</mml:mo></mml:mrow></mml:math></inline-formula>, where <inline-formula><mml:math id="M197" display="inline"><mml:mrow><mml:msub><mml:mi>p</mml:mi><mml:mi mathvariant="normal">∞</mml:mi></mml:msub></mml:mrow></mml:math></inline-formula> is the background pore pressure, <inline-formula><mml:math id="M198" display="inline"><mml:mrow><mml:mi mathvariant="normal">Δ</mml:mi><mml:mi>p</mml:mi><mml:mo>&gt;</mml:mo><mml:mn mathvariant="normal">0</mml:mn></mml:mrow></mml:math></inline-formula> is the pressure reduction at the channel, and <inline-formula><mml:math id="M199" display="inline"><mml:mi mathvariant="italic">γ</mml:mi></mml:math></inline-formula> is its decay length.

            <disp-formula id="Ch1.E29" content-type="numbered"><label>29</label><mml:math id="M200" display="block"><mml:mrow><mml:msub><mml:mi mathvariant="italic">τ</mml:mi><mml:mi>c</mml:mi></mml:msub><mml:mo>(</mml:mo><mml:mi>y</mml:mi><mml:mo>)</mml:mo><mml:mo>=</mml:mo><mml:msub><mml:mi>C</mml:mi><mml:mi mathvariant="normal">f</mml:mi></mml:msub><mml:mfenced close="]" open="["><mml:mrow><mml:msub><mml:mi mathvariant="italic">σ</mml:mi><mml:mi>n</mml:mi></mml:msub><mml:mo>-</mml:mo><mml:msub><mml:mi>p</mml:mi><mml:mi mathvariant="normal">∞</mml:mi></mml:msub><mml:mo>+</mml:mo><mml:mi mathvariant="normal">Δ</mml:mi><mml:mi>p</mml:mi><mml:mi>exp⁡</mml:mi><mml:mspace linebreak="nobreak" width="-0.125em"/><mml:mfenced close=")" open="("><mml:mrow><mml:mo>-</mml:mo><mml:mstyle displaystyle="true"><mml:mfrac style="display"><mml:mrow><mml:mo>|</mml:mo><mml:mi>y</mml:mi><mml:mo>-</mml:mo><mml:mi>Y</mml:mi><mml:mo>|</mml:mo></mml:mrow><mml:mi mathvariant="italic">γ</mml:mi></mml:mfrac></mml:mstyle></mml:mrow></mml:mfenced></mml:mrow></mml:mfenced><mml:mo>+</mml:mo><mml:mi>c</mml:mi><mml:mo>.</mml:mo></mml:mrow></mml:math></disp-formula>

          similar to a previous study of the effect of channelized drainage on shear margins <xref ref-type="bibr" rid="bib1.bibx13" id="paren.89"/>. The exponential form follows the treatment of <xref ref-type="bibr" rid="bib1.bibx50" id="text.90"/>: the pore-pressure reduction near a channel reflects the competition between drainage into the channel and resupply from the surrounding distributed system, which to leading order yields a perturbation that decays exponentially away from the channel, with a decay length set by the film thickness of the distributed system.</p>
      <p id="d2e5518">Another factor modifying subglacial drainage in our domain is the variable overburden of the ice. The approximately linear basal topography across the trunk implies that ice thickness increases from approximately 1000 m near the eastern to about 1700 m near the western margin. We capture the effect of overburden pressure on the underlying hydrological system by making the basal strength, <inline-formula><mml:math id="M201" display="inline"><mml:mrow><mml:msub><mml:mi mathvariant="italic">τ</mml:mi><mml:mi>c</mml:mi></mml:msub></mml:mrow></mml:math></inline-formula>, a function of the effective pressure <inline-formula><mml:math id="M202" display="inline"><mml:mrow><mml:mi>N</mml:mi><mml:mo>=</mml:mo><mml:mi mathvariant="italic">ρ</mml:mi><mml:mi>g</mml:mi><mml:mi>H</mml:mi><mml:mo>-</mml:mo><mml:mi>p</mml:mi></mml:mrow></mml:math></inline-formula> at the bed:

            <disp-formula id="Ch1.E30" content-type="numbered"><label>30</label><mml:math id="M203" display="block"><mml:mrow><mml:msub><mml:mi mathvariant="italic">τ</mml:mi><mml:mi>c</mml:mi></mml:msub><mml:mo>(</mml:mo><mml:mi>y</mml:mi><mml:mo>)</mml:mo><mml:mo>=</mml:mo><mml:msub><mml:mi>C</mml:mi><mml:mi mathvariant="normal">f</mml:mi></mml:msub><mml:mfenced close="]" open="["><mml:mrow><mml:mi mathvariant="italic">ρ</mml:mi><mml:mi>g</mml:mi><mml:mi>H</mml:mi><mml:mo>(</mml:mo><mml:mi>y</mml:mi><mml:mo>)</mml:mo><mml:mo>-</mml:mo><mml:mi>p</mml:mi><mml:mo>(</mml:mo><mml:mi>y</mml:mi><mml:mo>)</mml:mo></mml:mrow></mml:mfenced><mml:mo>+</mml:mo><mml:mi>c</mml:mi><mml:mo>.</mml:mo></mml:mrow></mml:math></disp-formula>

          To link the pore pressure to the overburden, we take the pore pressure to be a fixed fraction of the overburden, <inline-formula><mml:math id="M204" display="inline"><mml:mrow><mml:mi>p</mml:mi><mml:mo>(</mml:mo><mml:mi>y</mml:mi><mml:mo>)</mml:mo><mml:mo>=</mml:mo><mml:msub><mml:mi>k</mml:mi><mml:mi>p</mml:mi></mml:msub><mml:mspace linebreak="nobreak" width="0.125em"/><mml:mi mathvariant="italic">ρ</mml:mi><mml:mi>g</mml:mi><mml:mi>H</mml:mi><mml:mo>(</mml:mo><mml:mi>y</mml:mi><mml:mo>)</mml:mo></mml:mrow></mml:math></inline-formula>, which corresponds to a spatially near-constant flotation fraction across the trunk. Substituting this relation gives

            <disp-formula id="Ch1.E31" content-type="numbered"><label>31</label><mml:math id="M205" display="block"><mml:mrow><mml:msub><mml:mi mathvariant="italic">τ</mml:mi><mml:mi>c</mml:mi></mml:msub><mml:mo>(</mml:mo><mml:mi>y</mml:mi><mml:mo>)</mml:mo><mml:mo>=</mml:mo><mml:msub><mml:mi>C</mml:mi><mml:mi mathvariant="normal">f</mml:mi></mml:msub><mml:mspace width="0.125em" linebreak="nobreak"/><mml:mi mathvariant="italic">ρ</mml:mi><mml:mi>g</mml:mi><mml:mi>H</mml:mi><mml:mo>(</mml:mo><mml:mi>y</mml:mi><mml:mo>)</mml:mo><mml:mo>(</mml:mo><mml:mn mathvariant="normal">1</mml:mn><mml:mo>-</mml:mo><mml:msub><mml:mi>k</mml:mi><mml:mi>p</mml:mi></mml:msub><mml:mo>)</mml:mo><mml:mo>+</mml:mo><mml:mi>c</mml:mi><mml:mo>,</mml:mo></mml:mrow></mml:math></disp-formula>

          with friction parameter, <inline-formula><mml:math id="M206" display="inline"><mml:mrow><mml:msub><mml:mi>C</mml:mi><mml:mi mathvariant="normal">f</mml:mi></mml:msub><mml:mo>≈</mml:mo><mml:mn mathvariant="normal">0.5</mml:mn><mml:mo>-</mml:mo><mml:mn mathvariant="normal">0.6</mml:mn></mml:mrow></mml:math></inline-formula> <xref ref-type="bibr" rid="bib1.bibx52" id="paren.91"/>, sediment cohesion, <inline-formula><mml:math id="M207" display="inline"><mml:mrow><mml:mi>c</mml:mi><mml:mo>≈</mml:mo><mml:mn mathvariant="normal">1</mml:mn></mml:mrow></mml:math></inline-formula>–2 kPa <xref ref-type="bibr" rid="bib1.bibx36" id="paren.92"/>, and flotation fraction, <inline-formula><mml:math id="M208" display="inline"><mml:mrow><mml:msub><mml:mi>k</mml:mi><mml:mi>p</mml:mi></mml:msub></mml:mrow></mml:math></inline-formula>, which varies from 0 (no flotation) to 1 (full flotation). We do not treat <inline-formula><mml:math id="M209" display="inline"><mml:mrow><mml:msub><mml:mi>k</mml:mi><mml:mi>p</mml:mi></mml:msub></mml:mrow></mml:math></inline-formula> as an independent free parameter in the convex solve for lack of constraints on how it might vary. Instead, we prescribe a basal-strength profile that varies linearly with ice thickness <inline-formula><mml:math id="M210" display="inline"><mml:mi>H</mml:mi></mml:math></inline-formula> and adjust the corresponding strength coefficients so that the modeled surface speed and transverse strain rate best match the MEaSUREs observations along the ortho-flow line.</p>
</sec>
<sec id="Ch1.S3.SS3">
  <label>3.3</label><title>Iterative hypothesis testing</title>
      <p id="d2e5774">The central idea of iterative hypothesis testing with Ortholine is that different basal processes generate distinct spatial profiles of basal strength, but only some of these profiles are consistent with observed surface velocities. Similarly, different basal processes have distinct implications for the subglacial melt rate that could be evaluated against observations of basal hydrology. For example, subglacial drainage can occur in multiple configurations <xref ref-type="bibr" rid="bib1.bibx14" id="paren.93"><named-content content-type="pre">e.g.,</named-content></xref>, each implying a different basal-strength profile. A channelized system would produce localized strengthening that decays away from the channel <xref ref-type="bibr" rid="bib1.bibx50" id="paren.94"/>, whereas a distributed thin film would yield a more uniform bed strength. By testing which basal strength profiles reproduce measured surface speeds, Ortholine constrains which basal processes are plausible at a given field site. Also, the hypothesized hydrological drainage elements need to be fed by melting of ice and can hence be evaluated for consistency with ice dynamics.</p>
      <p id="d2e5785">In Fig. <xref ref-type="fig" rid="F7"/>, we illustrate how to test the hypothesis that basal strength underneath the fast-moving portion of Institute Ice Stream is dominated by ice overburden. We compare two simulations, one in which we assume that ice overburden does not affect the basal strength (panels a1–a5) and one in which ice overburden is the primary control (panels b1–b5). More specifically, we assume that the bed strength is constant for the simulation on the left, as would be the case if variable ice thickness did not matter. For the simulations on the right, we assume that pore pressure increases with ice overburden (see Eq. <xref ref-type="disp-formula" rid="Ch1.E31"/>), leading to a bed strength that varies linearly with ice thickness <inline-formula><mml:math id="M211" display="inline"><mml:mi>H</mml:mi></mml:math></inline-formula> and is maximal at the topographic low. In both cases, we only show the simulation that best reproduces the observed surface speed.</p>

      <fig id="F7" specific-use="star"><label>Figure 7</label><caption><p id="d2e5801">Numerical experiment investigating the relative importance of ice overburden. <bold>(a)</bold> Counterfactual simulation assuming uniform basal strength in the sediment indicative of ice overburden not being important. <bold>(b)</bold> Simulation assuming linearly increasing basal strength in the sediment as a function of overburden pressure. Rows illustrate: <bold>(1)</bold> Comparison of modeled velocities to observed MEaSUREs surface velocities across profile A–A'. <bold>(2)</bold> The modeled and observed transverse shear strain-rate component, <inline-formula><mml:math id="M212" display="inline"><mml:mrow><mml:msub><mml:mover accent="true"><mml:mi mathvariant="italic">ϵ</mml:mi><mml:mo mathvariant="normal">˙</mml:mo></mml:mover><mml:mi>x</mml:mi></mml:msub><mml:mi>y</mml:mi><mml:mo>=</mml:mo><mml:mo>(</mml:mo><mml:mn mathvariant="normal">1</mml:mn><mml:mo>/</mml:mo><mml:mn mathvariant="normal">2</mml:mn><mml:mo>)</mml:mo><mml:mo>∂</mml:mo><mml:mi>u</mml:mi><mml:mo>/</mml:mo><mml:mo>∂</mml:mo><mml:mi>y</mml:mi></mml:mrow></mml:math></inline-formula>, across A–A' (yr<sup>−1</sup>). <bold>(3)</bold> Tuned basal conditions, with locations specified from observations. Hard-rock sliding parameter (blue), saturated sediment strength (red), and locked sediments (gray). <bold>(4)</bold> Englacial temperature field within the ice column. <bold>(5)</bold> Melt rate at the base of the ice.</p></caption>
          <graphic xlink:href="https://gmd.copernicus.org/articles/19/11041/2026/gmd-19-11041-2026-f07.png"/>

        </fig>

      <p id="d2e5884">For both simulations, the rows in Fig. <xref ref-type="fig" rid="F7"/> show the calculated downstream speed compared to the observed surface speed (panels a1 and b1), the calculated and observed transverse strain rates (panels a2 and b2), the hard-rock sliding parameter and sediment strength of the best-fitting computation (panels a3 and b3), the englacial temperature distribution (panels a4 and b4) and the resulting subglacial melt rate (panels a5 and b5). In both cases, we assume that the lithologic transition from rock (blue line in a3 and b3) to sediment (red line in a3 and b3) occurs in the vicinity of the western shear margin and that the ice trunk is bordered by areas of frozen basal conditions (shown in gray). In general, the computation based on the linear decrease in sediment strength from the western to the eastern shear margin fits observed surface speed much better (Fig. <xref ref-type="fig" rid="F7"/>b1 vs a1). The best-fit strength profile varies linearly from 29.85 to 18.35 kPa across the ice stream trunk, mimicking the approximately linear basal topography across the trunk for an approximately constant average flotation factor of <inline-formula><mml:math id="M214" display="inline"><mml:mrow><mml:msub><mml:mi>k</mml:mi><mml:mi>p</mml:mi></mml:msub><mml:mo>=</mml:mo><mml:mn mathvariant="normal">0.996</mml:mn></mml:mrow></mml:math></inline-formula>. We note that this flotation fraction is diagnosed from this best-fit strength profile, together with representative values of the friction coefficient <inline-formula><mml:math id="M215" display="inline"><mml:mrow><mml:msub><mml:mi>C</mml:mi><mml:mi mathvariant="normal">f</mml:mi></mml:msub></mml:mrow></mml:math></inline-formula> and sediment cohesion <inline-formula><mml:math id="M216" display="inline"><mml:mi>c</mml:mi></mml:math></inline-formula>, and would vary if other values for <inline-formula><mml:math id="M217" display="inline"><mml:mrow><mml:msub><mml:mi>C</mml:mi><mml:mi mathvariant="normal">f</mml:mi></mml:msub></mml:mrow></mml:math></inline-formula> and <inline-formula><mml:math id="M218" display="inline"><mml:mi>c</mml:mi></mml:math></inline-formula> were chosen.</p>
      <p id="d2e5943">An interesting nuance of the analysis is that the ice overburden not only affects the basal strength, but also exhibits an indirect control on the englacial temperature and on subglacial melt rate. The reason is that for the computation on the left (panels a1–a4), the sediment portion of the bed is relatively weak and the hard-rock portion of the bed plays a more important role in the overall force balance. In contrast, sediment strength increases significantly from the eastern to the western shear margin for the computation on the right (panels b1–b4). The high basal strength in parts of the sediment portion means that the sliding coefficient of the rock portion of the bed is relatively lower, reducing shear heating near the western shear margin (see Fig. <xref ref-type="fig" rid="F7"/>b4) compared to the case where the sliding coefficient is high in both segments (see Fig. <xref ref-type="fig" rid="F7"/>a4). As a consequence, meltwater is generated in different locations with melt concentrated near the topographic low in Fig. <xref ref-type="fig" rid="F7"/>a5 as opposed to near the eastern shear margin in Fig. <xref ref-type="fig" rid="F7"/>b5.</p>
      <p id="d2e5954">Although Fig. <xref ref-type="fig" rid="F7"/>b1 fits the measured ice speed well for most of the ice-stream trunk, it reproduces the surface speed in the extended marginal zone less well. This disconnect suggests that a physical process other than ice overburden controls the basal strength distribution in the extended marginal zone. A likely candidate is subglacial hydrology, because the meltwater channel that emanates from below the western shear margin at the grounding line (see Fig. <xref ref-type="fig" rid="F6"/>a) likely extends upstream into the trough near the western shear margin. The meltwater channel locally strengthens the bed (see Eq. <xref ref-type="disp-formula" rid="Ch1.E29"/>), but little is known about how quickly the pressure perturbation induced by a meltwater channel decays spatially in the presence of both lithological and topographic variations. For a plastic bed, we assume a constant decay length, <inline-formula><mml:math id="M219" display="inline"><mml:mi mathvariant="italic">γ</mml:mi></mml:math></inline-formula>, everywhere, implying that the channel alters the bed strength similarly irrespective of the lithology and local topography.</p>
      <p id="d2e5970">The existence of a meltwater channel near the western shear margin is well supported by observational data (e.g., see Fig. <xref ref-type="fig" rid="F6"/> and accompanying text). Less clear is how this meltwater channel affects basal strength. In Fig. <xref ref-type="fig" rid="F8"/>, we test the hypothesis that a channel leads to an exponential increase in basal strength near the western shear margin. We compare two simulations: one represents velocity-strengthening sliding over hard rock (Fig. <xref ref-type="fig" rid="F8"/>a1–a5), and the other assumes Coulomb sliding (Fig. <xref ref-type="fig" rid="F8"/>b1–b5). In the sediment-dominated portion of the domain, we superimpose the basal strengthening of the channel with a linear decrease in bed strength towards the eastern shear margin as a result of decreasing overburden (see Fig. <xref ref-type="fig" rid="F7"/>). The agreement between modeled and observed surface speeds (Fig. <xref ref-type="fig" rid="F8"/>a1, b1) and strain rates (Fig. <xref ref-type="fig" rid="F8"/>a2, b2) in the extended marginal zone improves significantly compared to Fig. <xref ref-type="fig" rid="F7"/> The quality of fit is comparable for both simulations, highlighting that both basal-strength distributions have comparable explanatory potential.</p>

      <fig id="F8" specific-use="star"><label>Figure 8</label><caption><p id="d2e5992">Numerical experiment to assess the relative importance of a channel located in the trough near the western shear margin. <bold>(a)</bold> Simulation assuming velocity-strengthening sliding over hard rock. <bold>(b)</bold> Simulation assuming Coulomb sliding in the extended marginal zone. Both simulations include channel-related strengthening near the western shear margin. Rows illustrate: <bold>(1)</bold> Comparison of modeled velocities to observed MEaSUREs surface velocities across profile A–A′. <bold>(2)</bold> The modeled and observed transverse shear strain-rate component, <inline-formula><mml:math id="M220" display="inline"><mml:mrow><mml:msub><mml:mover accent="true"><mml:mi mathvariant="italic">ϵ</mml:mi><mml:mo mathvariant="normal">˙</mml:mo></mml:mover><mml:mi>x</mml:mi></mml:msub><mml:mi>y</mml:mi><mml:mo>=</mml:mo><mml:mo>(</mml:mo><mml:mn mathvariant="normal">1</mml:mn><mml:mo>/</mml:mo><mml:mn mathvariant="normal">2</mml:mn><mml:mo>)</mml:mo><mml:mo>∂</mml:mo><mml:mi>u</mml:mi><mml:mo>/</mml:mo><mml:mo>∂</mml:mo><mml:mi>y</mml:mi></mml:mrow></mml:math></inline-formula>, across A–A' (yr<sup>−1</sup>). <bold>(3)</bold> Tuned basal conditions, showing the hard-rock sliding parameter in blue and plastic-bed strength in red. <bold>(4)</bold> Temperature field in the ice column. <bold>(5)</bold> Melt rate at the base of the ice.</p></caption>
          <graphic xlink:href="https://gmd.copernicus.org/articles/19/11041/2026/gmd-19-11041-2026-f08.png"/>

        </fig>

      <p id="d2e6075">To create a counterfactual for Fig. <xref ref-type="fig" rid="F8"/> that represents a scenario without a meltwater channel in the trough below the western shear margin, we assume that the presence of pervasive subglacial water leads to regularized Coulomb sliding in the hard-rock portion of our model domain <xref ref-type="bibr" rid="bib1.bibx41 bib1.bibx32 bib1.bibx15 bib1.bibx16 bib1.bibx63 bib1.bibx18" id="paren.95"/>. In Fig. <xref ref-type="fig" rid="F9"/>, we test whether a plastic bed with variable strength improves the fit to observed surface speeds in the western shear margin and extended marginal zone. Similarly to Fig. <xref ref-type="fig" rid="F7"/>, we restrict our tests to constant and linearly varying basal strength profiles. To highlight that regularized Coulomb sliding is equivalent to plastic behavior of the bed, we plot both the hard-rock and the sediment-dominated portion of the domain in red in Fig. <xref ref-type="fig" rid="F9"/>a3 and b3. Building on our analysis from Fig. <xref ref-type="fig" rid="F7"/>, we assume that the basal strength of the sediments decreases linearly from the western to the eastern shear margin.</p>

      <fig id="F9" specific-use="star"><label>Figure 9</label><caption><p id="d2e6094">Counterfactual simulations assuming the absence of a meltwater channel near the western shear margin. <bold>(a)</bold> Simulation assuming that regularized Coulomb sliding over hard rock leads to approximately uniform basal strength. <bold>(b)</bold> Simulation assuming regularized Coulomb sliding with a linearly increasing basal strength in the extended marginal zone. Rows illustrate: <bold>(1)</bold> Comparison of modeled velocities to observed MEaSUREs surface velocities across profile A–A'. <bold>(2)</bold> The modeled and observed transverse shear strain-rate component, <inline-formula><mml:math id="M222" display="inline"><mml:mrow><mml:msub><mml:mover accent="true"><mml:mi mathvariant="italic">ϵ</mml:mi><mml:mo mathvariant="normal">˙</mml:mo></mml:mover><mml:mi>x</mml:mi></mml:msub><mml:mi>y</mml:mi><mml:mo>=</mml:mo><mml:mo>(</mml:mo><mml:mn mathvariant="normal">1</mml:mn><mml:mo>/</mml:mo><mml:mn mathvariant="normal">2</mml:mn><mml:mo>)</mml:mo><mml:mo>∂</mml:mo><mml:mi>u</mml:mi><mml:mo>/</mml:mo><mml:mo>∂</mml:mo><mml:mi>y</mml:mi></mml:mrow></mml:math></inline-formula>, across A–A' (yr<sup>−1</sup>). <bold>(3)</bold> Basal strength of the plastically behaving bed. <bold>(4)</bold> Temperature field in the ice column. <bold>(5)</bold> Melt rate at the base of the ice.</p></caption>
          <graphic xlink:href="https://gmd.copernicus.org/articles/19/11041/2026/gmd-19-11041-2026-f09.png"/>

        </fig>

      <p id="d2e6176">The two best fitting simulations shown in Fig. <xref ref-type="fig" rid="F9"/> demonstrate that switching the nature of sliding does not improve the fit to observed surface speed and strain rates. When assuming a constant basal strength, we are unable to reproduce the slow but finite sliding outside of the western shear margin (see Fig. <xref ref-type="fig" rid="F9"/>a1). In comparison, velocity-strengthening sliding (see Fig. <xref ref-type="fig" rid="F7"/>) is consistent with the observed slow slip. A linearly increasing basal strength profile provides a better fit in the extended marginal zone, but fails to capture the western shear margin (see Fig. <xref ref-type="fig" rid="F9"/>b1 and b2). We emphasize that, contrary to Fig. <xref ref-type="fig" rid="F7"/>, it is also not clear why the basal strength in the extended marginal zone would increase linearly, because ice overburden does not vary much in the extended marginal zone.</p>
</sec>
<sec id="Ch1.S3.SS4">
  <label>3.4</label><title>Drawing inferences about the physical processes controlling shear-margin position and stability</title>
      <p id="d2e6197">Figure <xref ref-type="fig" rid="F7"/> suggests that the simplest hydrological system capable of reproducing the observed surface speed at Institute Ice Stream is a spatially well-connected drainage network near flotation. The simplest physical process that could explain the properties of this drainage network is variable ice overburden on a temperate bed with variable basal topography. Needless to say, there is almost certainly considerable small-scale variability not captured in Ortholine. However, this small-scale variability might not affect surface speed because ice acts as a low-pass filter on basal topography and slipperiness, as analyzed in more detail in <xref ref-type="bibr" rid="bib1.bibx24" id="text.96"/>.</p>
      <p id="d2e6205">Transfer functions, pioneered by <xref ref-type="bibr" rid="bib1.bibx24" id="text.97"/>, are an approach to estimating the wavelength below which small-scale variability in basal topography or slipperiness is likely not relevant for surface velocities unless this small-scale variability is coupled with other basal processes. For example, a small topographic low containing a meltwater channel can still be dynamically relevant, not because of its own spatial scale but because of its large-scale effects on water routing. For example, the best fitting simulation shown in Fig. <xref ref-type="fig" rid="F7"/>b1–b4 entails a temperate zone in the eastern shear margin that could feed a meltwater channel causing local strengthening of the bed <xref ref-type="bibr" rid="bib1.bibx49 bib1.bibx51 bib1.bibx72 bib1.bibx13" id="paren.98"/>. Alternatively, the increase in bed strength could result from freezing <xref ref-type="bibr" rid="bib1.bibx62 bib1.bibx67 bib1.bibx72 bib1.bibx50 bib1.bibx26" id="paren.99"/>. The latter explanation would require lateral advection of cold ice into the eastern shear margin to counteract the heating entailed in the measured strain rates (Fig. <xref ref-type="fig" rid="F7"/>b2) and an additional englacial process like fabric to explain the narrow margin width.</p>
      <p id="d2e6221">Further evidence for the importance of small-scale variability in subglacial hydrology comes from Fig. <xref ref-type="fig" rid="F8"/>, which suggests that the basal strength distribution over the hard-rock bed is dominated by the existence of a meltwater channel in or close to the western shear margin. However, Fig. <xref ref-type="fig" rid="F7"/> demonstrates that the presence of this channel is not a necessary component to obtain a satisfactory fit to the surface speed and strain rates in most of the domain. This finding suggests that the channel is an important physical control for the western but not necessarily for the eastern shear margin. Given the pronounced lateral asymmetries in topography, lithology, and subglacial hydrology at Institute (see Sect. 3.2) it is perhaps not surprising that the two shear margins could be controlled by different physical processes.</p>
      <p id="d2e6228">When drawing inferences about the physical processes controlling shear-margin properties through Ortholine, it is important to keep in mind that individual factors, such as topography, lithology, temperature, and subglacial hydrology, can be interconnected. This interconnectivity is the reason why it is challenging to derive insights into the physical processes controlling shear-margin properties from observational data alone, particularly for field sites like Institute where all of these factors vary simultaneously. Similarly, there may be large-scale variations that correlate with or follow a trend similar to overburden stress and are therefore difficult to differentiate from effects of ice thickness. We merely note here that it is not necessary to invoke this complexity in most of the domain for the model to be compatible with surface observables (Fig. <xref ref-type="fig" rid="F7"/>b1).</p>
      <p id="d2e6234">To disentangle the various dependencies, Ortholine uses simple auxiliary assumptions (e.g., Eqs. <xref ref-type="disp-formula" rid="Ch1.E29"/> and  <xref ref-type="disp-formula" rid="Ch1.E31"/>) to link spatial variability in subglacial hydrology to specific physical processes such as ice overburden or the presence of a Röthlisberger channel. These auxiliary assumptions can be adjusted by the user. Generally, we recommend keeping the number of free parameters used in these auxiliary assumptions small to reduce the risk of overfitting. A similar strategy was recently used in <xref ref-type="bibr" rid="bib1.bibx73" id="text.100"/> to constrain the physical processes controlling shear-margin behavior at Thwaites, albeit in the context of a different modeling approach. We have deliberately framed this step as forward hypothesis testing rather than as a formal inversion of the basal strength from the surface velocities. The choices that enter the auxiliary assumptions are unavoidable in either framing, but making them explicit and physically motivated, rather than folding them into the regularization and priors of an inversion, keeps the small number of free parameters transparent. The two approaches are complementary: an inversion could propose a basal-strength profile and the forward model could then test which physical processes are capable of producing it, and the balance between the two is ultimately for the user to decide.</p>
      <p id="d2e6244">Ortholine is designed to evaluate the relative importance of different englacial and subglacial processes in controlling a specific shear-margin position. The main value of this insight for making inferences about future shear-margin stability lies in realizing that different physical processes evolve on different time scales. Topographically controlled shear margins are the most stable, the shear margin at Lambert Glacier being one example <xref ref-type="bibr" rid="bib1.bibx60" id="paren.101"/>. Thermal processes such as thawing of the subglacial bed can lead to shear-margin migration <xref ref-type="bibr" rid="bib1.bibx67" id="paren.102"/> that is much faster than topographic change or erosion, but still comparatively slower than hydrological processes that can lead to sudden rearrangements of shear-margin position <xref ref-type="bibr" rid="bib1.bibx3 bib1.bibx9 bib1.bibx13" id="paren.103"/>.</p>
</sec>
</sec>
<sec id="Ch1.S4">
  <label>4</label><title>Discussion</title>
      <p id="d2e6265">Different models have different strengths and weaknesses. Ice-sheet models such as the Ice Sheet System Model <xref ref-type="bibr" rid="bib1.bibx39" id="paren.104"/> or the Parallel Ice Sheet Model <xref ref-type="bibr" rid="bib1.bibx8" id="paren.105"/> aim for realism by capturing the full two- or three-dimensional stress balance. They can represent lateral drag, tributary inflow, and grounding line migration in complex geometries, but these advantages come at the cost of high computational demand and significant data requirements to constrain all model parameters <xref ref-type="bibr" rid="bib1.bibx37" id="paren.106"/>. Flow-line models, by contrast, neglect lateral drag and assume simplified geometries, but are computationally efficient <xref ref-type="bibr" rid="bib1.bibx46" id="paren.107"/>. They are tools for hypothesis testing by isolating a particular aspect of ice dynamics and focusing on analyzing its impact and have advanced our understanding of how longitudinal stress gradients change the transient responses of glaciers to external forcing <xref ref-type="bibr" rid="bib1.bibx48" id="paren.108"/> or how grounding-line stability varies with basal drag and rheology <xref ref-type="bibr" rid="bib1.bibx66" id="paren.109"/>.</p>
      <p id="d2e6287">Conceptually, Ortholine follows the modeling philosophy of flow-line models, in which some aspects of ice dynamics are intentionally simplified or neglected to isolate the effect of others, and is complementary to them: instead of focusing on the longitudinal stress balance along a flow line while neglecting lateral effects, Ortholine focuses on lateral stresses while neglecting longitudinal effects.</p>
      <p id="d2e6290">Currently, evaluations of shear-margin stability are typically based on a qualitative comparison of whether the shear-margin coincides with distinctive topographic features or not <xref ref-type="bibr" rid="bib1.bibx42 bib1.bibx60" id="paren.110"><named-content content-type="pre">e.g.,</named-content></xref>. For example, <xref ref-type="bibr" rid="bib1.bibx42" id="text.111"/> argued that the eastern shear margin at Thwaites Glacier could be prone to migration because the topography below the eastern shear margin is flat. However, a more in-depth quantitative analysis of the shear margins at Thwaites <xref ref-type="bibr" rid="bib1.bibx73" id="paren.112"/> shows that this criterion alone is not sufficient to assess shear-margin stability. Many other processes affect shear-margin dynamics sensitively, from basal processes such as subglacial hydrology to regional-scale mass balance leading to rapid ice thinning at Thwaites <xref ref-type="bibr" rid="bib1.bibx70" id="paren.113"/>.</p>
      <p id="d2e6307">Similarly, the presence of pronounced topography does not necessarily imply margin stability. For example, <xref ref-type="bibr" rid="bib1.bibx73" id="text.114"/> show that the western shear margin at Thwaites Glacier could also be prone to migration despite being located in an area of much more pronounced topography. Our simulations for Institute support this more nuanced view of shear-margin controls: despite the western shear margin at Institute being approximately co-located with a topographic trough, our simulations suggest that the presence of a meltwater channel in this trough is important, as evidenced by simulations with a meltwater channel (see Fig. <xref ref-type="fig" rid="F8"/>a1 and b1) fitting observed surface speeds much better than simulations without a channel (see Fig. <xref ref-type="fig" rid="F9"/>a1 and b1).</p>
      <p id="d2e6318">In addition to the inspiration we take from flow-line models, our model is motivated by the insights of previous antiplane shear models of shear margins <xref ref-type="bibr" rid="bib1.bibx33 bib1.bibx62 bib1.bibx72 bib1.bibx49 bib1.bibx26 bib1.bibx13" id="paren.115"/>. At first, it might seem counterintuitive to adapt these highly idealized models to a particular field site, but this step is warranted by our advancing scientific understanding. We now have both observational <xref ref-type="bibr" rid="bib1.bibx25 bib1.bibx5 bib1.bibx11 bib1.bibx10 bib1.bibx9" id="paren.116"/> and theoretical evidence <xref ref-type="bibr" rid="bib1.bibx33 bib1.bibx62 bib1.bibx72 bib1.bibx49 bib1.bibx26 bib1.bibx13" id="paren.117"/>  that shear-margin migration is possible, in principle. The challenge now becomes to assess if and where shear-margin migration might occur not only in principle, but also in reality. This challenge is probably most acute for Thwaites Glacier <xref ref-type="bibr" rid="bib1.bibx42 bib1.bibx73" id="paren.118"/>, but also relevant for other regions, such as Institute, where accelerated melting at the grounding line is expected later this century <xref ref-type="bibr" rid="bib1.bibx28" id="paren.119"/>.</p>
      <p id="d2e6336">In addition to our growing scientific understanding, the design of field-site-specific shear-margin models through Ortholine is enabled by the substantial increase in the resolution of remotely detected data probing the ice sheets <xref ref-type="bibr" rid="bib1.bibx68 bib1.bibx69 bib1.bibx45 bib1.bibx38" id="paren.120"/>. Radio-echo sounding <xref ref-type="bibr" rid="bib1.bibx68 bib1.bibx69" id="paren.121"/> and seismic <xref ref-type="bibr" rid="bib1.bibx45 bib1.bibx38" id="paren.122"/> surveys have advanced our ability to map heterogeneities in subglacial conditions. These increasingly detailed observations highlight that heterogeneity in basal conditions is the norm rather than the exception. The observed diversity is not surprising in light of the numerous physical processes at play in a heterogeneous subglacial environment <xref ref-type="bibr" rid="bib1.bibx68 bib1.bibx69 bib1.bibx45 bib1.bibx38" id="paren.123"/>. Increasingly, these detailed observations provide an important opportunity for models to better constrain the relative importance of the various physical processes that contribute to the dynamics of ice streams and their shear margins.</p>
</sec>
<sec id="Ch1.S5" sec-type="conclusions">
  <label>5</label><title>Conclusions</title>
      <p id="d2e6360">In this paper, we develop Ortholine, a thermomechanical, antiplane shear model. Ortholine is designed to evaluate the relative importance of englacial and subglacial processes controlling shear-margin location. Building on the variational formulation by <xref ref-type="bibr" rid="bib1.bibx65" id="text.124"/>, we frame the problem in more flexible Disciplined Convex Programming <xref ref-type="bibr" rid="bib1.bibx22" id="paren.125"/> and implement a numerical solution using the software package CVX <xref ref-type="bibr" rid="bib1.bibx21" id="paren.126"/>. We verify our numerical implementation against the idealized mountain glacier case proposed by <xref ref-type="bibr" rid="bib1.bibx65" id="text.127"/>. We then illustrate the four key steps for applying the model to a given field site, using Institute Ice Stream as a proof-of-concept. Our simulations show that identifying the physical processes governing shear-margin stability through spatial association alone is not reliable. The presence of a topographic trough below a shear margin does not necessarily imply that the position of this shear margin is topographically controlled, as highlighted by the example of the western shear margin at Institute. Similarly, the presence of a gently sloping topography does not mean that topographic effects are irrelevant, as demonstrated in Fig. <xref ref-type="fig" rid="F7"/>. Ortholine provides a framework for evaluating the relative importance of different physical processes across an antiplane cross-section in a transparent, if simplified, way.</p>
</sec>

      
      </body>
    <back><notes notes-type="codedataavailability"><title>Code and data availability</title>

      <p id="d2e6381">Observational data including MEaSUREs surface velocity measurements and locations of subglacially sourced channels are publicly available through the National Snow &amp; Ice Data Center (<uri>https://nsidc.org/data/</uri>, last access: 21 September 2026). The code developed in this paper is open source (<uri>https://github.com/coopere/InstituteIceStream2D</uri>, last access: 21 September 2026) under the GNU General Public License, version 3 (GPL-3.0). The specific version of the code used to create the results in this manuscript is ortholine_v1.1 <xref ref-type="bibr" rid="bib1.bibx12" id="paren.128"/>, archived at  <ext-link xlink:href="https://doi.org/10.5281/zenodo.20858511" ext-link-type="DOI">10.5281/zenodo.20858511</ext-link>.</p>
  </notes><notes notes-type="authorcontribution"><title>Author contributions</title>

      <p id="d2e6399">JS supervised the different stages of the study, participated in the conceptualization of the model, wrote the manuscript, and provided funding for all authors involved. CE led the early model and numerical method development stages, implemented the CVX-based version of the code, performed the verification analysis, and wrote the Methods section of the manuscript. NA performed the reproduction analysis and helped improve the code. All authors have reviewed and approved the final version of the manuscript.</p>
  </notes><notes notes-type="competinginterests"><title>Competing interests</title>

      <p id="d2e6405">The contact author has declared that none of the authors has any competing interests.</p>
  </notes><notes notes-type="disclaimer"><title>Disclaimer</title>

      <p id="d2e6411">Publisher's note: Copernicus Publications remains neutral with regard to jurisdictional claims made in the text, published maps, institutional affiliations, or any other geographical representation in this paper. The authors bear the ultimate responsibility for providing appropriate place names. Views expressed in the text are those of the authors and do not necessarily reflect the views of the publisher.</p>
  </notes><ack><title>Acknowledgements</title><p id="d2e6417">This work benefited greatly from the input of Paul Summers, Anuj Shetty, and Samantha Hordyk. The authors thank Dustin Schroeder, Neal Iverson, Lucas Zoet, Knut Christianson, and Mark Fahnestock for fruitful discussions.</p></ack><notes notes-type="financialsupport"><title>Financial support</title>

      <p id="d2e6422">This research has been supported by the National Science Foundation, Directorate for Geosciences, Office of Polar Programs (grant nos. 1341499, 1744758, and 1739027).</p>
  </notes><notes notes-type="reviewstatement"><title>Review statement</title>

      <p id="d2e6428">This paper was edited by Ludovic Räss and reviewed by two anonymous referees.</p>
  </notes><ref-list>
    <title>References</title>

      <ref id="bib1.bibx1"><label>Alley et al.(2016)Alley, Scambos, Siegfried, and Fricker</label><mixed-citation>Alley, K. E., Scambos, T. A., Siegfried, M. R., and Fricker, H. A.: Impacts of warm water on Antarctic ice shelf stability through basal channel formation, Nat. Geosci., 9, 290–293, <ext-link xlink:href="https://doi.org/10.1038/ngeo2675" ext-link-type="DOI">10.1038/ngeo2675</ext-link>, 2016.</mixed-citation></ref>
      <ref id="bib1.bibx2"><label>Alley et al.(2018)Alley, Scambos, S., Rajaram, Pope, and Haran</label><mixed-citation>Alley, K. E., Scambos, T. A., Anderson, R. S., Rajaram, H., Pope, A., and Haran, T. M.: Continent-wide estimates of Antarctic strain rates from Landsat 8-derived velocity grids, J. Glaciol., 64, 321–332, <ext-link xlink:href="https://doi.org/10.1017/jog.2018.23" ext-link-type="DOI">10.1017/jog.2018.23</ext-link>, 2018.</mixed-citation></ref>
      <ref id="bib1.bibx3"><label>Alley et al.(1994)Alley, Anandakrishnan, Bentley, and Lord</label><mixed-citation>Alley, R. B., Anandakrishnan, S., Bentley, C. R., and Lord, N.: A water-piracy hypothesis for the stagnation of Ice Stream C, Antarctica, Ann. Glaciol., 20, 187–194, <ext-link xlink:href="https://doi.org/10.3189/172756494794587032" ext-link-type="DOI">10.3189/172756494794587032</ext-link>, 1994.</mixed-citation></ref>
      <ref id="bib1.bibx4"><label>Bamber et al.(2000)Bamber, Vaughan, and Joughin</label><mixed-citation>Bamber, J. L., Vaughan, D. G., and Joughin, I.: Widespread complex flow in the interior of the Antarctic ice sheet, Science, 287, 1248–1250, <ext-link xlink:href="https://doi.org/10.1126/science.287.5456.1248" ext-link-type="DOI">10.1126/science.287.5456.1248</ext-link>, 2000.</mixed-citation></ref>
      <ref id="bib1.bibx5"><label>Bindschadler and Vornberger(1998)</label><mixed-citation>Bindschadler, R. and Vornberger, P. L.: Changes in the West Antarctic Ice Sheet Since 1963 from Declassified Satellite Photography, Science, 279, 689–692, <ext-link xlink:href="https://doi.org/10.1126/science.279.5351.689" ext-link-type="DOI">10.1126/science.279.5351.689</ext-link>, 1998.</mixed-citation></ref>
      <ref id="bib1.bibx6"><label>Bingham and Siegert(2007)</label><mixed-citation>Bingham, R. G. and Siegert, M. J.: Radar-derived bed roughness characterization of Institute and Möller ice streams, West Antarctica, and comparison with Siple Coast ice streams, Geophys. Res. Lett., 34, L21504, <ext-link xlink:href="https://doi.org/10.1029/2007GL031483" ext-link-type="DOI">10.1029/2007GL031483</ext-link>, 2007.</mixed-citation></ref>
      <ref id="bib1.bibx7"><label>Brenner and Scott(2008)</label><mixed-citation>Brenner, S. C. and Scott, L. R.: The Mathematical Theory of Finite Element Methods, 3rd edn., Texts in Applied Mathematics, vol. 15, Springer, <ext-link xlink:href="https://doi.org/10.1007/978-0-387-75934-0" ext-link-type="DOI">10.1007/978-0-387-75934-0</ext-link>, 2008.</mixed-citation></ref>
      <ref id="bib1.bibx8"><label>Bueler and Brown(2009)</label><mixed-citation>Bueler, E. and Brown, J.: Shallow shelf approximation as a “sliding law” in a thermomechanically coupled ice sheet model, J. Geophys. Res.-Earth, 114, F03008, <ext-link xlink:href="https://doi.org/10.1029/2008JF001179" ext-link-type="DOI">10.1029/2008JF001179</ext-link>, 2009.</mixed-citation></ref>
      <ref id="bib1.bibx9"><label>Catania et al.(2006)Catania, Scambos, Conway, and Raymond</label><mixed-citation>Catania, G. A., Scambos, T. A., Conway, H., and Raymond, C. F.: Sequential stagnation of Kamb Ice Stream, West Antarctica, Geophys. Res. Lett., 33, L14502, <ext-link xlink:href="https://doi.org/10.1029/2006GL026430" ext-link-type="DOI">10.1029/2006GL026430</ext-link>, 2006.</mixed-citation></ref>
      <ref id="bib1.bibx10"><label>Clarke et al.(2000)Clarke, Liu, Lord, and Bentley</label><mixed-citation>Clarke, T. S., Liu, C., Lord, N. E., and Bentley, C. R.: Evidence for a recently abandoned shear margin adjacent to ice stream B2, Antarctica, from ice-penetrating radar measurements, J. Geophys. Res.: Solid Earth, 105, 13409–13422, <ext-link xlink:href="https://doi.org/10.1029/2000JB900037" ext-link-type="DOI">10.1029/2000JB900037</ext-link>, 2000.</mixed-citation></ref>
      <ref id="bib1.bibx11"><label>Echelmeyer and Harrison(1999)</label><mixed-citation>Echelmeyer, K. and Harrison, W.: Ongoing margin migration of ice stream B, Antarctica, J. Glaciol., 45, 361–369, <ext-link xlink:href="https://doi.org/10.3189/002214399793377059" ext-link-type="DOI">10.3189/002214399793377059</ext-link>, 1999.</mixed-citation></ref>
      <ref id="bib1.bibx12"><label>Elsworth(2026)</label><mixed-citation>Elsworth, C. W.: coopere/InstituteIceStream2D: ortholine_v1.1, Zenodo [code], <ext-link xlink:href="https://doi.org/10.5281/zenodo.20858511" ext-link-type="DOI">10.5281/zenodo.20858511</ext-link>, 2026.</mixed-citation></ref>
      <ref id="bib1.bibx13"><label>Elsworth and Suckale(2016)</label><mixed-citation>Elsworth, C. W. and Suckale, J.: Rapid ice flow rearrangement induced by subglacial drainage in West Antarctica, Geophys. Res. Lett., 43, 11697–11707, <ext-link xlink:href="https://doi.org/10.1002/2016GL070430" ext-link-type="DOI">10.1002/2016GL070430</ext-link>, 2016.</mixed-citation></ref>
      <ref id="bib1.bibx14"><label>Flowers(2015)</label><mixed-citation>Flowers, G. E.: Modelling water flow under glaciers and ice sheets, P. Roy. Soc. A-Math. Phy., 471, 20140907, <ext-link xlink:href="https://doi.org/10.1098/rspa.2014.0907" ext-link-type="DOI">10.1098/rspa.2014.0907</ext-link>, 2015.</mixed-citation></ref>
      <ref id="bib1.bibx15"><label>Fowler(1986)</label><mixed-citation>Fowler, A.: A sliding law for glaciers of constant viscosity in the presence of subglacial cavitation, Proc. R. Soc. Lond. Ser A Math. Phys. Sci., 407, 147–170, <ext-link xlink:href="https://doi.org/10.1098/rspa.1986.0090" ext-link-type="DOI">10.1098/rspa.1986.0090</ext-link>, 1986.</mixed-citation></ref>
      <ref id="bib1.bibx16"><label>Fowler(1987)</label><mixed-citation>Fowler, A.: Sliding with cavity formation, J. Glaciol., 33, 255–267, <ext-link xlink:href="https://doi.org/10.3189/S0022143000008820" ext-link-type="DOI">10.3189/S0022143000008820</ext-link>, 1987.</mixed-citation></ref>
      <ref id="bib1.bibx17"><label>Friedman(2010)</label><mixed-citation> Friedman, A.: Variational principles and free-boundary problems, Dover Publications, Mineola, NY, ISBN 978-0-486-47853-1, 2010.</mixed-citation></ref>
      <ref id="bib1.bibx18"><label>Gagliardini et al.(2007)Gagliardini, Cohen, Råback, and Zwinger</label><mixed-citation>Gagliardini, O., Cohen, D., Råback, P., and Zwinger, T.: Finite-element modeling of subglacial cavities and related friction law, J. Geophys. Res.-Earth, 112, F02027, <ext-link xlink:href="https://doi.org/10.1029/2006JF000576" ext-link-type="DOI">10.1029/2006JF000576</ext-link>, 2007.</mixed-citation></ref>
      <ref id="bib1.bibx19"><label>Geuzaine and Remacle(2009)</label><mixed-citation>Geuzaine, C. and Remacle, J.-F.: Gmsh: A three-dimensional finite element mesh generator with built-in pre- and post-processing facilities, Int. J. Numer. Methods Eng., 79, 1309–1331, <ext-link xlink:href="https://doi.org/10.1002/nme.2579" ext-link-type="DOI">10.1002/nme.2579</ext-link>, 2009.</mixed-citation></ref>
      <ref id="bib1.bibx20"><label>Glen(1955)</label><mixed-citation>Glen, J. W.: The creep of polycrystalline ice, Proc. R. Soc. Lond. Ser. A Math Phys. Sci., 228, 519–538, <ext-link xlink:href="https://doi.org/10.1098/rspa.1955.0066" ext-link-type="DOI">10.1098/rspa.1955.0066</ext-link>, 1955.</mixed-citation></ref>
      <ref id="bib1.bibx21"><label>Grant and Boyd(2014)</label><mixed-citation>Grant, M. and Boyd, S.: CVX: Matlab Software for Disciplined Convex Programming, version 2.1, CVX Research [code], <uri>https://cvxr.com/cvx/</uri> (last access: 5 January 2026), 2014.</mixed-citation></ref>
      <ref id="bib1.bibx22"><label>Grant et al.(2006)Grant, Boyd, and Ye</label><mixed-citation>Grant, M., Boyd, S., and Ye, Y.: Disciplined Convex Programming, in: Global Optimization: From Theory to Implementation, edited by Liberti, L. and Maculan, N., 155–210, Springer, Boston, MA, <ext-link xlink:href="https://doi.org/10.1007/0-387-30065-1_7" ext-link-type="DOI">10.1007/0-387-30065-1_7</ext-link>, 2006.</mixed-citation></ref>
      <ref id="bib1.bibx23"><label>Greene et al.(2017)Greene, Gwyther, and Blankenship</label><mixed-citation>Greene, C. A., Gwyther, D. E., and Blankenship, D. D.: Antarctic Mapping Tools for MATLAB, Comp. Geosci., 104, 151–157, <ext-link xlink:href="https://doi.org/10.1016/j.cageo.2016.08.003" ext-link-type="DOI">10.1016/j.cageo.2016.08.003</ext-link>, 2017.</mixed-citation></ref>
      <ref id="bib1.bibx24"><label>Gudmundsson(2003)</label><mixed-citation>Gudmundsson, G. H.: Transmission of basal variability to a glacier surface, J. Geophys. Res.: Solid Earth, 108,  2253, <ext-link xlink:href="https://doi.org/10.1029/2002JB002107" ext-link-type="DOI">10.1029/2002JB002107</ext-link>, 2003.</mixed-citation></ref>
      <ref id="bib1.bibx25"><label>Harrison et al.(1998)Harrison, Echelmeyer, and Larsen</label><mixed-citation>Harrison, W., Echelmeyer, K., and Larsen, C.: Measurement of temperature in a margin of Ice Stream B, Antarctica: implications for margin migration and lateral drag, J. Glaciol., 44, 615–624, <ext-link xlink:href="https://doi.org/10.3189/S0022143000002112" ext-link-type="DOI">10.3189/S0022143000002112</ext-link>, 1998.</mixed-citation></ref>
      <ref id="bib1.bibx26"><label>Haseloff et al.(2015)Haseloff, Schoof, and Gagliardini</label><mixed-citation>Haseloff, M., Schoof, C., and Gagliardini, O.: A boundary layer model for ice stream margins, J. Fluid Mech., 781, 353–387, <ext-link xlink:href="https://doi.org/10.1017/jfm.2015.503" ext-link-type="DOI">10.1017/jfm.2015.503</ext-link>, 2015.</mixed-citation></ref>
      <ref id="bib1.bibx27"><label>Haseloff et al.(2019)Haseloff, Hewitt, and Katz</label><mixed-citation>Haseloff, M., Hewitt, I., and Katz, R.: Englacial pore water localizes shear in temperate ice stream margins, J. Geophys. Res.-Earth, 124, 2521–2541, <ext-link xlink:href="https://doi.org/10.1029/2019JF005399" ext-link-type="DOI">10.1029/2019JF005399</ext-link>, 2019.</mixed-citation></ref>
      <ref id="bib1.bibx28"><label>Hellmer et al.(2012)Hellmer, Kauker, Timmermann, Determann, and Rae</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.bibx29"><label>Hewitt(2011)</label><mixed-citation>Hewitt, I. J.: Modelling distributed and channelized subglacial drainage: the spacing of channels, J. Glaciol., 57, 302–314, <ext-link xlink:href="https://doi.org/10.3189/002214311796405951" ext-link-type="DOI">10.3189/002214311796405951</ext-link>, 2011.</mixed-citation></ref>
      <ref id="bib1.bibx30"><label>Holland et al.(2023)Holland, Bevan, and Luckman</label><mixed-citation>Holland, P. R., Bevan, S. L., and Luckman, A. J.: Strong ocean melting feedback during the recent retreat of Thwaites Glacier, Geophys. Res. Lett., 50, e2023GL103088, <ext-link xlink:href="https://doi.org/10.1029/2023GL103088" ext-link-type="DOI">10.1029/2023GL103088</ext-link>, 2023.</mixed-citation></ref>
      <ref id="bib1.bibx31"><label>Hooke et al.(1979)Hooke, Raymond, Hotchkiss, and Gustafson</label><mixed-citation>Hooke, R. L., Raymond, C. F., Hotchkiss, R. L., and Gustafson, R. J.: Calculations of velocity and temperature in a polar glacier using the finite-element method, J. Glaciol., 24, 131–146, <ext-link xlink:href="https://doi.org/10.3189/S0022143000014696" ext-link-type="DOI">10.3189/S0022143000014696</ext-link>, 1979.</mixed-citation></ref>
      <ref id="bib1.bibx32"><label>Iken(1981)</label><mixed-citation>Iken, A.: The effect of the subglacial water pressure on the sliding velocity of a glacier in an idealized numerical model, J. Glaciol., 27, 407–421, <ext-link xlink:href="https://doi.org/10.3189/S0022143000011448" ext-link-type="DOI">10.3189/S0022143000011448</ext-link>, 1981.</mixed-citation></ref>
      <ref id="bib1.bibx33"><label>Jacobson and Raymond(1998)</label><mixed-citation>Jacobson, H. and Raymond, C.: Thermal effects on the location of ice stream margins, J. Geophys. Res.: Solid Earth, 103, 12111–12122, <ext-link xlink:href="https://doi.org/10.1029/98JB00574" ext-link-type="DOI">10.1029/98JB00574</ext-link>, 1998.</mixed-citation></ref>
      <ref id="bib1.bibx34"><label>Jamieson et al.(2012)Jamieson, Vieli, Livingstone, Cofaigh, Stokes, Hillenbrand, and Dowdeswell</label><mixed-citation>Jamieson, S. S. R., Vieli, A., Livingstone, S. J., Ó Cofaigh, C., Stokes, C., Hillenbrand, C.-D., and Dowdeswell, J. A.: Ice-stream stability on a reverse bed slope, Nat. Geosci., 5, 799–802, <ext-link xlink:href="https://doi.org/10.1038/ngeo1600" ext-link-type="DOI">10.1038/ngeo1600</ext-link>, 2012.</mixed-citation></ref>
      <ref id="bib1.bibx35"><label>Jordan et al.(2013)Jordan, Ferraccioli, Ross, Corr, Leat, Bingham, Rippin, le Brocq, and Siegert</label><mixed-citation>Jordan, T. A., Ferraccioli, F., Ross, N., Corr, H. F., Leat, P. T., Bingham, R. G., Rippin, D. M., le Brocq, A., and Siegert, M. J.: Inland extent of the Weddell Sea Rift imaged by new aerogeophysical data, Tectonophysics, 585, 137–160, <ext-link xlink:href="https://doi.org/10.1016/j.tecto.2012.09.010" ext-link-type="DOI">10.1016/j.tecto.2012.09.010</ext-link>, 2013.</mixed-citation></ref>
      <ref id="bib1.bibx36"><label>Kamb(2001)</label><mixed-citation>Kamb, B.: Basal zone of the West Antarctic ice streams and its role in lubrication of their rapid motion, in: The West Antarctic ice sheet: behavior and environment, 77, 157–199, <ext-link xlink:href="https://doi.org/10.1029/AR077p0157" ext-link-type="DOI">10.1029/AR077p0157</ext-link>, 2001.</mixed-citation></ref>
      <ref id="bib1.bibx37"><label>Kirchner et al.(2011)Kirchner, Hutter, Jakobsson, and Gyllencreutz</label><mixed-citation>Kirchner, N., Hutter, K., Jakobsson, M., and Gyllencreutz, R.: Capabilities and limitations of numerical ice sheet models: A discussion for Earth-scientists and modelers, Quat. Sci. Rev., 30, 3691–3704, <ext-link xlink:href="https://doi.org/10.1016/j.quascirev.2011.09.012" ext-link-type="DOI">10.1016/j.quascirev.2011.09.012</ext-link>, 2011.</mixed-citation></ref>
      <ref id="bib1.bibx38"><label>Koellner et al.(2019)Koellner, Parizek, Alley, Muto, Holschuh, and Nowicki</label><mixed-citation>Koellner, S., Parizek, B., Alley, R., Muto, A., Holschuh, N., and Nowicki, S.: The impact of spatially-variable basal properties on outlet glacier flow, Earth Planet. Sci. Lett., 515, 200–208, <ext-link xlink:href="https://doi.org/10.1016/j.epsl.2019.03.026" ext-link-type="DOI">10.1016/j.epsl.2019.03.026</ext-link>, 2019.</mixed-citation></ref>
      <ref id="bib1.bibx39"><label>Larour et al.(2012)Larour, Seroussi, Morlighem, and Rignot</label><mixed-citation>Larour, E., Seroussi, H., Morlighem, M., and Rignot, E.: Continental scale, high order, high spatial resolution, ice sheet modeling using the Ice Sheet System Model (ISSM), J. Geophys. Res.-Earth, 117,  F01022, <ext-link xlink:href="https://doi.org/10.1029/2011JF002140" ext-link-type="DOI">10.1029/2011JF002140</ext-link>, 2012.</mixed-citation></ref>
      <ref id="bib1.bibx40"><label>Le Brocq et al.(2013)LeBrocq, Ross, Griggs, Bingham, Corr, Ferraccioli, Jenkins, Jordan, Payne, Rippin, and Siegert</label><mixed-citation>Le Brocq, A. M., Ross, N., Griggs, J. A., Bingham, R. G., Corr, H. F. J., Ferraccioli, F., Jenkins, A., Jordan, T. A., Payne, A. J., Rippin, D. M., and Siegert, M. J.: Evidence from ice shelves for channelized meltwater flow beneath the Antarctic Ice Sheet, Nat. Geosci., 6, 945–948, <ext-link xlink:href="https://doi.org/10.1038/ngeo1977" ext-link-type="DOI">10.1038/ngeo1977</ext-link>, 2013.</mixed-citation></ref>
      <ref id="bib1.bibx41"><label>Lliboutry(1968)</label><mixed-citation>Lliboutry, L.: General theory of subglacial cavitation and sliding of temperate glaciers, J. Glaciol., 7, 21–58, <ext-link xlink:href="https://doi.org/10.3189/S0022143000020396" ext-link-type="DOI">10.3189/S0022143000020396</ext-link>, 1968.</mixed-citation></ref>
      <ref id="bib1.bibx42"><label>MacGregor et al.(2013)MacGregor, Catania, Conway, Schroeder, Joughin, Young, Kempf, and Blankenship</label><mixed-citation>MacGregor, J. A., Catania, G. A., Conway, H., Schroeder, D. M., Joughin, I., Young, D. A., Kempf, S. D., and Blankenship, D. D.: Weak bed control of the eastern shear margin of Thwaites Glacier, West Antarctica, J. Glaciol., 59, 900–912, <ext-link xlink:href="https://doi.org/10.3189/2013JoG13J050" ext-link-type="DOI">10.3189/2013JoG13J050</ext-link>, 2013.</mixed-citation></ref>
      <ref id="bib1.bibx43"><label>Meyer and Minchew(2018)</label><mixed-citation>Meyer, C. R. and Minchew, B. M.: Temperate ice in the shear margins of the Antarctic Ice Sheet: Controlling processes and preliminary locations, Earth Planet. Sci. Lett., 498, 17–26, <ext-link xlink:href="https://doi.org/10.1016/j.epsl.2018.06.028" ext-link-type="DOI">10.1016/j.epsl.2018.06.028</ext-link>, 2018.</mixed-citation></ref>
      <ref id="bib1.bibx44"><label>Morland and Johnson(1980)</label><mixed-citation>Morland, L. and Johnson, I.: Steady motion of ice sheets, J. Glaciol., 25, 229–246, <ext-link xlink:href="https://doi.org/10.3189/S0022143000010467" ext-link-type="DOI">10.3189/S0022143000010467</ext-link>, 1980.</mixed-citation></ref>
      <ref id="bib1.bibx45"><label>Muto et al.(2019)Muto, Anandakrishnan, Alley, Horgan, Parizek, Koellner, Christianson, and Holschuh</label><mixed-citation>Muto, A., Anandakrishnan, S., Alley, R. B., Horgan, H. J., Parizek, B. R., Koellner, S., Christianson, K., and Holschuh, N.: Relating bed character and subglacial morphology using seismic data from Thwaites Glacier, West Antarctica, Earth Planet. Sci. Lett., 507, 199–206, <ext-link xlink:href="https://doi.org/10.1016/j.epsl.2018.12.008" ext-link-type="DOI">10.1016/j.epsl.2018.12.008</ext-link>, 2019.</mixed-citation></ref>
      <ref id="bib1.bibx46"><label>Nye(1952)</label><mixed-citation>Nye, J. F.: The mechanics of glacier flow, J. Glaciol., 2, 82–93, <ext-link xlink:href="https://doi.org/10.3189/S0022143000033967" ext-link-type="DOI">10.3189/S0022143000033967</ext-link>, 1952.</mixed-citation></ref>
      <ref id="bib1.bibx47"><label>Nye(1957)</label><mixed-citation>Nye, J. F.: The distribution of stress and velocity in glaciers and ice-sheets, Proc. R. Soc. Lond. Ser. A Math Phys. Sci., 239, 113–133, <ext-link xlink:href="https://doi.org/10.1098/rspa.1957.0026" ext-link-type="DOI">10.1098/rspa.1957.0026</ext-link>, 1957.</mixed-citation></ref>
      <ref id="bib1.bibx48"><label>Pattyn(2002)</label><mixed-citation>Pattyn, F.: Transient glacier response with a higher-order numerical ice-flow model, J. Glaciol., 48, 467–477, <ext-link xlink:href="https://doi.org/10.3189/172756502781831278" ext-link-type="DOI">10.3189/172756502781831278</ext-link>, 2002.</mixed-citation></ref>
      <ref id="bib1.bibx49"><label>Perol and Rice(2015)</label><mixed-citation>Perol, T. and Rice, J. R.: Shear heating and weakening of the margins of West Antarctic ice streams, Geophys. Res. Lett., 42, 3406–3413, <ext-link xlink:href="https://doi.org/10.1002/2015GL063638" ext-link-type="DOI">10.1002/2015GL063638</ext-link>, 2015.</mixed-citation></ref>
      <ref id="bib1.bibx50"><label>Perol et al.(2015)Perol, Rice, Platt, and Suckale</label><mixed-citation>Perol, T., Rice, J. R., Platt, J. D., and Suckale, J.: Subglacial hydrology and ice stream margin locations, J. Geophys. Res.-Earth, 120, 1352–1368, <ext-link xlink:href="https://doi.org/10.1002/2015JF003542" ext-link-type="DOI">10.1002/2015JF003542</ext-link>, 2015.</mixed-citation></ref>
      <ref id="bib1.bibx51"><label>Platt et al.(2016)Platt, Perol, Suckale, and Rice</label><mixed-citation>Platt, J. D., Perol, T., Suckale, J., and Rice, J. R.: Determining conditions that allow a shear margin to coincide with a Röthlisberger channel, J. Geophys. Res.-Earth, 121, 1273–1294, <ext-link xlink:href="https://doi.org/10.1002/2015JF003707" ext-link-type="DOI">10.1002/2015JF003707</ext-link>, 2016.</mixed-citation></ref>
      <ref id="bib1.bibx52"><label>Rathbun et al.(2008)Rathbun, Marone, Alley, and Anandakrishnan</label><mixed-citation>Rathbun, A. P., Marone, C., Alley, R. B., and Anandakrishnan, S.: Laboratory study of the frictional rheology of sheared till, J. Geophys. Res.-Earth, 113,  F02020, <ext-link xlink:href="https://doi.org/10.1029/2007JF000815" ext-link-type="DOI">10.1029/2007JF000815</ext-link>, 2008.</mixed-citation></ref>
      <ref id="bib1.bibx53"><label>Raymond(1996)</label><mixed-citation>Raymond, C.: Shear margins in glaciers and ice sheets, J. Glaciol., 42, 90–102, <ext-link xlink:href="https://doi.org/10.3189/S0022143000030550" ext-link-type="DOI">10.3189/S0022143000030550</ext-link>, 1996.</mixed-citation></ref>
      <ref id="bib1.bibx54"><label>Reeh and Paterson(1988)</label><mixed-citation>Reeh, N. and Paterson, W.: Application of a flow model to the ice-divide region of Devon Island ice cap, Canada, J. Glaciol., 34, 55–63, <ext-link xlink:href="https://doi.org/10.3189/S0022143000009060" ext-link-type="DOI">10.3189/S0022143000009060</ext-link>, 1988.</mixed-citation></ref>
      <ref id="bib1.bibx55"><label>Rignot et al.(2011)Rignot, Mouginot, and Scheuchl</label><mixed-citation>Rignot, E., Mouginot, J., and Scheuchl, B.: Ice flow of the Antarctic ice sheet, Science, 333, 1427–1430, <ext-link xlink:href="https://doi.org/10.1126/science.1208336" ext-link-type="DOI">10.1126/science.1208336</ext-link>, 2011.</mixed-citation></ref>
      <ref id="bib1.bibx56"><label>Ross et al.(2011)Ross, Siegert, Woodward, Smith, Corr, Bentley, Hindmarsh, King, and Rivera</label><mixed-citation>Ross, N., Siegert, M. J., Woodward, J., Smith, A. M., Corr, H. F., Bentley, M. J., Hindmarsh, R. C., King, E. C., and Rivera, A.: Holocene stability of the Amundsen-Weddell ice divide, West Antarctica, Geology, 39, 935–938, <ext-link xlink:href="https://doi.org/10.1130/G31920.1" ext-link-type="DOI">10.1130/G31920.1</ext-link>, 2011.</mixed-citation></ref>
      <ref id="bib1.bibx57"><label>Ross et al.(2012)Ross, Bingham, Corr, Ferraccioli, Jordan, Le Brocq, Rippin, Young, Blankenship, and Siegert</label><mixed-citation>Ross, N., Bingham, R., Corr, H. F. J., Ferraccioli, F., Jordan, T., Le Brocq, A., Rippin, D., Young, D., Blankenship, D., and Siegert, M. J.: Steep reverse bed slope at the grounding line of the Weddell Sea sector in West Antarctica, Nat. Geosci., 5, 393–396, <ext-link xlink:href="https://doi.org/10.1038/ngeo1468" ext-link-type="DOI">10.1038/ngeo1468</ext-link>, 2012.</mixed-citation></ref>
      <ref id="bib1.bibx58"><label>Ross et al.(2014)Ross, Jordan, Bingham, Corr, Ferraccioli, Brocq, Rippin, Wright, and Siegert</label><mixed-citation>Ross, N., Jordan, T. A., Bingham, R. G., Corr, H. F. J., Ferraccioli, F., Le Brocq, A., Rippin, D. M., Wright, A. P., and Siegert, M. J.: The Ellsworth Subglacial Highlands: Inception and retreat of the West Antarctic Ice Sheet, Geol. Soc. Am. Bull., 126, 3–15, <ext-link xlink:href="https://doi.org/10.1130/B30794.1" ext-link-type="DOI">10.1130/B30794.1</ext-link>, 2014.</mixed-citation></ref>
      <ref id="bib1.bibx59"><label>Röthlisberger(1972)</label><mixed-citation>Röthlisberger, H.: Water pressure in intra- and subglacial channels, J. Glaciol., 11, 177–203, <ext-link xlink:href="https://doi.org/10.3189/S0022143000022188" ext-link-type="DOI">10.3189/S0022143000022188</ext-link>, 1972.</mixed-citation></ref>
      <ref id="bib1.bibx60"><label>Sanderson et al.(2023)Sanderson, Winter, Callard, Napoleoni, Ross, Jordan, and Bingham</label><mixed-citation>Sanderson, R. J., Winter, K., Callard, S. L., Napoleoni, F., Ross, N., Jordan, T. A., and Bingham, R. G.: Englacial architecture of Lambert Glacier, East Antarctica, The Cryosphere, 17, 4853–4871, <ext-link xlink:href="https://doi.org/10.5194/tc-17-4853-2023" ext-link-type="DOI">10.5194/tc-17-4853-2023</ext-link>, 2023.</mixed-citation></ref>
      <ref id="bib1.bibx61"><label>Scambos et al.(2004)Scambos, Bohlander, Raup, and Haran</label><mixed-citation>Scambos, T., Bohlander, J., Raup, B., and Haran, T.: Glaciological characteristics of Institute Ice Stream using remote sensing, Antarct. Sci., 16, 205–213, <ext-link xlink:href="https://doi.org/10.1017/S0954102004001919" ext-link-type="DOI">10.1017/S0954102004001919</ext-link>, 2004.</mixed-citation></ref>
      <ref id="bib1.bibx62"><label>Schoof(2004)</label><mixed-citation>Schoof, C.: On the mechanics of ice-stream shear margins, J. Glaciol., 50, 208–218, <ext-link xlink:href="https://doi.org/10.3189/172756504781830024" ext-link-type="DOI">10.3189/172756504781830024</ext-link>, 2004.</mixed-citation></ref>
      <ref id="bib1.bibx63"><label>Schoof(2005)</label><mixed-citation>Schoof, C.: The effect of cavitation on glacier sliding, Proc. R. Soc. Lond. Ser. A Math Phys. Sci., 461, 609–627, <ext-link xlink:href="https://doi.org/10.1098/rspa.2004.1350" ext-link-type="DOI">10.1098/rspa.2004.1350</ext-link>, 2005.</mixed-citation></ref>
      <ref id="bib1.bibx64"><label>Schoof(2006a)</label><mixed-citation>Schoof, C.: A variational approach to ice stream flow, J. Fluid Mech., 556, 227–251, <ext-link xlink:href="https://doi.org/10.1017/S0022112006009591" ext-link-type="DOI">10.1017/S0022112006009591</ext-link>, 2006a.</mixed-citation></ref>
      <ref id="bib1.bibx65"><label>Schoof(2006b)</label><mixed-citation>Schoof, C.: Variational methods for glacier flow over plastic till, J. Fluid Mech., 555, 299–320, <ext-link xlink:href="https://doi.org/10.1017/S0022112006009104" ext-link-type="DOI">10.1017/S0022112006009104</ext-link>, 2006b.</mixed-citation></ref>
      <ref id="bib1.bibx66"><label>Schoof(2007)</label><mixed-citation>Schoof, C.: Ice sheet grounding line dynamics: Steady states, stability, and hysteresis, J. Geophys. Res., 112, F03S28, <ext-link xlink:href="https://doi.org/10.1029/2006JF000664" ext-link-type="DOI">10.1029/2006JF000664</ext-link>, 2007.</mixed-citation></ref>
      <ref id="bib1.bibx67"><label>Schoof(2012)</label><mixed-citation>Schoof, C.: Thermally driven migration of ice-stream shear margins, J. Fluid Mech., 712, 552–578, <ext-link xlink:href="https://doi.org/10.1017/jfm.2012.438" ext-link-type="DOI">10.1017/jfm.2012.438</ext-link>, 2012.</mixed-citation></ref>
      <ref id="bib1.bibx68"><label>Schroeder et al.(2014)Schroeder, Blankenship, Young, and Quartini</label><mixed-citation>Schroeder, D. M., Blankenship, D. D., Young, D. A., and Quartini, E.: Evidence for elevated and spatially variable geothermal flux beneath the West Antarctic Ice Sheet, P. Natl. Acad. Sci. USA, 111, 9070–9072, <ext-link xlink:href="https://doi.org/10.1073/pnas.1405184111" ext-link-type="DOI">10.1073/pnas.1405184111</ext-link>, 2014.</mixed-citation></ref>
      <ref id="bib1.bibx69"><label>Siegert et al.(2016)Siegert, Ross, Li, Schroeder, Rippin, Ashmore, Bingham, and Gogineni</label><mixed-citation>Siegert, M. J., Ross, N., Li, J., Schroeder, D. M., Rippin, D., Ashmore, D., Bingham, R., and Gogineni, P.: Subglacial controls on the flow of Institute Ice Stream, West Antarctica, Ann. Glaciol., 57, 19–24, <ext-link xlink:href="https://doi.org/10.1017/aog.2016.17" ext-link-type="DOI">10.1017/aog.2016.17</ext-link>, 2016.</mixed-citation></ref>
      <ref id="bib1.bibx70"><label>Smith et al.(2020)Smith, Fricker, Gardner, Medley, Nilsson, Paolo, Holschuh, Adusumilli, Brunt, Csatho et al.</label><mixed-citation>Smith, B., Fricker, H. A., Gardner, A. S., Medley, B., Nilsson, J., Paolo, F. S., Holschuh, N., Adusumilli, S., Brunt, K., Csatho, B., Harbeck, K., Markus, T., Neumann, T., Siegfried, M. R., and Zwally, H. J.: Pervasive ice sheet mass loss reflects competing ocean and atmosphere processes, Science, 368, 1239–1242, <ext-link xlink:href="https://doi.org/10.1126/science.aaz5845" ext-link-type="DOI">10.1126/science.aaz5845</ext-link>, 2020.</mixed-citation></ref>
      <ref id="bib1.bibx71"><label>Smith et al.(2009)Smith, A. Fricker, Joughin, and Slawek</label><mixed-citation>Smith, B. E., Fricker, H. A., Joughin, I. R., and Tulaczyk, S.: An inventory of active subglacial lakes in Antarctica detected by ICESat (2003–2008), J. Glaciol., 55, 573–595, <ext-link xlink:href="https://doi.org/10.3189/002214309789470879" ext-link-type="DOI">10.3189/002214309789470879</ext-link>, 2009.</mixed-citation></ref>
      <ref id="bib1.bibx72"><label>Suckale et al.(2014)Suckale, Platt, Perol, and Rice</label><mixed-citation>Suckale, J., Platt, J. D., Perol, T., and Rice, J. R.: Deformation-induced melting in the margins of the West Antarctic ice streams, J. Geophys. Res., 119, 1004–1025, <ext-link xlink:href="https://doi.org/10.1002/2013JF003008" ext-link-type="DOI">10.1002/2013JF003008</ext-link>, 2014.</mixed-citation></ref>
      <ref id="bib1.bibx73"><label>Summers et al.(2023)Summers, Elsworth, Dow, and Suckale</label><mixed-citation>Summers, P. T., Elsworth, C. W., Dow, C. F., and Suckale, J.: Migration of the shear margins at Thwaites Glacier: Dependence on basal conditions and testability against field data, J. Geophys. Res.-Earth, 128, e2022JF006958, <ext-link xlink:href="https://doi.org/10.1029/2022JF006958" ext-link-type="DOI">10.1029/2022JF006958</ext-link>, 2023.</mixed-citation></ref>
      <ref id="bib1.bibx74"><label>Thomas(1979)</label><mixed-citation>Thomas, R. H.: The dynamics of marine ice sheets, J. Glaciol., 24, 167–177, <ext-link xlink:href="https://doi.org/10.3189/S0022143000014726" ext-link-type="DOI">10.3189/S0022143000014726</ext-link>, 1979.</mixed-citation></ref>
      <ref id="bib1.bibx75"><label>Tulaczyk et al.(2000a)Tulaczyk, Kamb, and Engelhardt</label><mixed-citation>Tulaczyk, S., Kamb, W. B., and Engelhardt, H. F.: Basal mechanics of Ice Stream B, west Antarctica: 2. Undrained plastic bed model, J. Geophys. Res., 105, 483–494, <ext-link xlink:href="https://doi.org/10.1029/1999JB900328" ext-link-type="DOI">10.1029/1999JB900328</ext-link>, 2000a.</mixed-citation></ref>
      <ref id="bib1.bibx76"><label>Tulaczyk et al.(2000b)Tulaczyk, Kamb, and Engelhardt</label><mixed-citation>Tulaczyk, S., Kamb, W. B., and Engelhardt, H. F.: Basal mechanics of Ice Stream B, west Antarctica: 1. Till mechanics, J. Geophys. Res., 105, 463–481, <ext-link xlink:href="https://doi.org/10.1029/1999JB900329" ext-link-type="DOI">10.1029/1999JB900329</ext-link>, 2000b.</mixed-citation></ref>
      <ref id="bib1.bibx77"><label>Weertman(1957)</label><mixed-citation>Weertman, J.: On the sliding of glaciers, J. Glaciol., 3, 33–38, <ext-link xlink:href="https://doi.org/10.3189/S0022143000024709" ext-link-type="DOI">10.3189/S0022143000024709</ext-link>, 1957.</mixed-citation></ref>
      <ref id="bib1.bibx78"><label>Weertman(1974)</label><mixed-citation>Weertman, J.: Stability of the junction of an ice sheet and an ice shelf, J. Glaciol., 13, 3–11, <ext-link xlink:href="https://doi.org/10.3189/S0022143000023327" ext-link-type="DOI">10.3189/S0022143000023327</ext-link>, 1974.</mixed-citation></ref>
      <ref id="bib1.bibx79"><label>Zoet and Iverson(2016)</label><mixed-citation>Zoet, L. K. and Iverson, N. R.: Rate-weakening drag during glacier sliding, J. Geophys. Res.-Earth, 121, 1206–1217, <ext-link xlink:href="https://doi.org/10.1002/2016JF003909" ext-link-type="DOI">10.1002/2016JF003909</ext-link>, 2016.</mixed-citation></ref>

  </ref-list></back>
    <!--<article-title-html>An antiplane shear model for evaluating shear-margin stability (Ortholine v1.0)</article-title-html>
<abstract-html/>
<ref-html id="bib1.bib1"><label>Alley et al.(2016)Alley, Scambos, Siegfried, and Fricker</label><mixed-citation>
      
Alley, K. E., Scambos, T. A., Siegfried, M. R., and Fricker, H. A.: Impacts of warm water on Antarctic ice shelf stability through basal channel formation, Nat. Geosci., 9, 290–293, <a href="https://doi.org/10.1038/ngeo2675" target="_blank">https://doi.org/10.1038/ngeo2675</a>, 2016.

    </mixed-citation></ref-html>
<ref-html id="bib1.bib2"><label>Alley et al.(2018)Alley, Scambos, S., Rajaram, Pope, and Haran</label><mixed-citation>
      
Alley, K. E., Scambos, T. A., Anderson, R. S., Rajaram, H., Pope, A., and Haran, T. M.: Continent-wide estimates of Antarctic strain rates from Landsat 8-derived velocity grids, J. Glaciol., 64, 321–332, <a href="https://doi.org/10.1017/jog.2018.23" target="_blank">https://doi.org/10.1017/jog.2018.23</a>, 2018.

    </mixed-citation></ref-html>
<ref-html id="bib1.bib3"><label>Alley et al.(1994)Alley, Anandakrishnan, Bentley, and Lord</label><mixed-citation>
      
Alley, R. B., Anandakrishnan, S., Bentley, C. R., and Lord, N.: A water-piracy hypothesis for the stagnation of Ice Stream C, Antarctica, Ann. Glaciol., 20, 187–194, <a href="https://doi.org/10.3189/172756494794587032" target="_blank">https://doi.org/10.3189/172756494794587032</a>, 1994.

    </mixed-citation></ref-html>
<ref-html id="bib1.bib4"><label>Bamber et al.(2000)Bamber, Vaughan, and Joughin</label><mixed-citation>
      
Bamber, J. L., Vaughan, D. G., and Joughin, I.: Widespread complex flow in the interior of the Antarctic ice sheet, Science, 287, 1248–1250, <a href="https://doi.org/10.1126/science.287.5456.1248" target="_blank">https://doi.org/10.1126/science.287.5456.1248</a>, 2000.

    </mixed-citation></ref-html>
<ref-html id="bib1.bib5"><label>Bindschadler and Vornberger(1998)</label><mixed-citation>
      
Bindschadler, R. and Vornberger, P. L.: Changes in the West Antarctic Ice Sheet Since 1963 from Declassified Satellite Photography, Science, 279, 689–692, <a href="https://doi.org/10.1126/science.279.5351.689" target="_blank">https://doi.org/10.1126/science.279.5351.689</a>, 1998.

    </mixed-citation></ref-html>
<ref-html id="bib1.bib6"><label>Bingham and Siegert(2007)</label><mixed-citation>
      
Bingham, R. G. and Siegert, M. J.: Radar-derived bed roughness characterization of Institute and Möller ice streams, West Antarctica, and comparison with Siple Coast ice streams, Geophys. Res. Lett., 34, L21504, <a href="https://doi.org/10.1029/2007GL031483" target="_blank">https://doi.org/10.1029/2007GL031483</a>, 2007.

    </mixed-citation></ref-html>
<ref-html id="bib1.bib7"><label>Brenner and Scott(2008)</label><mixed-citation>
      
Brenner, S. C. and Scott, L. R.: The Mathematical Theory of Finite Element Methods, 3rd edn., Texts in Applied Mathematics, vol. 15, Springer, <a href="https://doi.org/10.1007/978-0-387-75934-0" target="_blank">https://doi.org/10.1007/978-0-387-75934-0</a>, 2008.

    </mixed-citation></ref-html>
<ref-html id="bib1.bib8"><label>Bueler and Brown(2009)</label><mixed-citation>
      
Bueler, E. and Brown, J.: Shallow shelf approximation as a “sliding law” in a thermomechanically coupled ice sheet model, J. Geophys. Res.-Earth, 114, F03008, <a href="https://doi.org/10.1029/2008JF001179" target="_blank">https://doi.org/10.1029/2008JF001179</a>, 2009.

    </mixed-citation></ref-html>
<ref-html id="bib1.bib9"><label>Catania et al.(2006)Catania, Scambos, Conway, and Raymond</label><mixed-citation>
      
Catania, G. A., Scambos, T. A., Conway, H., and Raymond, C. F.: Sequential stagnation of Kamb Ice Stream, West Antarctica, Geophys. Res. Lett., 33, L14502, <a href="https://doi.org/10.1029/2006GL026430" target="_blank">https://doi.org/10.1029/2006GL026430</a>, 2006.

    </mixed-citation></ref-html>
<ref-html id="bib1.bib10"><label>Clarke et al.(2000)Clarke, Liu, Lord, and Bentley</label><mixed-citation>
      
Clarke, T. S., Liu, C., Lord, N. E., and Bentley, C. R.: Evidence for a recently abandoned shear margin adjacent to ice stream B2, Antarctica, from ice-penetrating radar measurements, J. Geophys. Res.: Solid Earth, 105, 13409–13422, <a href="https://doi.org/10.1029/2000JB900037" target="_blank">https://doi.org/10.1029/2000JB900037</a>, 2000.

    </mixed-citation></ref-html>
<ref-html id="bib1.bib11"><label>Echelmeyer and Harrison(1999)</label><mixed-citation>
      
Echelmeyer, K. and Harrison, W.: Ongoing margin migration of ice stream B, Antarctica, J. Glaciol., 45, 361–369, <a href="https://doi.org/10.3189/002214399793377059" target="_blank">https://doi.org/10.3189/002214399793377059</a>, 1999.

    </mixed-citation></ref-html>
<ref-html id="bib1.bib12"><label>Elsworth(2026)</label><mixed-citation>
      
Elsworth, C. W.: coopere/InstituteIceStream2D: ortholine_v1.1, Zenodo [code], <a href="https://doi.org/10.5281/zenodo.20858511" target="_blank">https://doi.org/10.5281/zenodo.20858511</a>, 2026.

    </mixed-citation></ref-html>
<ref-html id="bib1.bib13"><label>Elsworth and Suckale(2016)</label><mixed-citation>
      
Elsworth, C. W. and Suckale, J.: Rapid ice flow rearrangement induced by subglacial drainage in West Antarctica, Geophys. Res. Lett., 43, 11697–11707, <a href="https://doi.org/10.1002/2016GL070430" target="_blank">https://doi.org/10.1002/2016GL070430</a>, 2016.

    </mixed-citation></ref-html>
<ref-html id="bib1.bib14"><label>Flowers(2015)</label><mixed-citation>
      
Flowers, G. E.: Modelling water flow under glaciers and ice sheets, P. Roy. Soc. A-Math. Phy., 471, 20140907, <a href="https://doi.org/10.1098/rspa.2014.0907" target="_blank">https://doi.org/10.1098/rspa.2014.0907</a>, 2015.

    </mixed-citation></ref-html>
<ref-html id="bib1.bib15"><label>Fowler(1986)</label><mixed-citation>
      
Fowler, A.: A sliding law for glaciers of constant viscosity in the presence of subglacial cavitation, Proc. R. Soc. Lond. Ser A Math. Phys. Sci., 407, 147–170, <a href="https://doi.org/10.1098/rspa.1986.0090" target="_blank">https://doi.org/10.1098/rspa.1986.0090</a>, 1986.

    </mixed-citation></ref-html>
<ref-html id="bib1.bib16"><label>Fowler(1987)</label><mixed-citation>
      
Fowler, A.: Sliding with cavity formation, J. Glaciol., 33, 255–267, <a href="https://doi.org/10.3189/S0022143000008820" target="_blank">https://doi.org/10.3189/S0022143000008820</a>, 1987.

    </mixed-citation></ref-html>
<ref-html id="bib1.bib17"><label>Friedman(2010)</label><mixed-citation>
      
Friedman, A.: Variational principles and free-boundary problems, Dover Publications, Mineola, NY, ISBN 978-0-486-47853-1, 2010.

    </mixed-citation></ref-html>
<ref-html id="bib1.bib18"><label>Gagliardini et al.(2007)Gagliardini, Cohen, Råback, and Zwinger</label><mixed-citation>
      
Gagliardini, O., Cohen, D., Råback, P., and Zwinger, T.: Finite-element modeling of subglacial cavities and related friction law, J. Geophys. Res.-Earth, 112, F02027, <a href="https://doi.org/10.1029/2006JF000576" target="_blank">https://doi.org/10.1029/2006JF000576</a>, 2007.

    </mixed-citation></ref-html>
<ref-html id="bib1.bib19"><label>Geuzaine and Remacle(2009)</label><mixed-citation>
      
Geuzaine, C. and Remacle, J.-F.: Gmsh: A three-dimensional finite element mesh generator with built-in pre- and post-processing facilities, Int. J. Numer. Methods Eng., 79, 1309–1331, <a href="https://doi.org/10.1002/nme.2579" target="_blank">https://doi.org/10.1002/nme.2579</a>, 2009.

    </mixed-citation></ref-html>
<ref-html id="bib1.bib20"><label>Glen(1955)</label><mixed-citation>
      
Glen, J. W.: The creep of polycrystalline ice, Proc. R. Soc. Lond. Ser. A Math Phys. Sci., 228, 519–538, <a href="https://doi.org/10.1098/rspa.1955.0066" target="_blank">https://doi.org/10.1098/rspa.1955.0066</a>, 1955.

    </mixed-citation></ref-html>
<ref-html id="bib1.bib21"><label>Grant and Boyd(2014)</label><mixed-citation>
      
Grant, M. and Boyd, S.: CVX: Matlab Software for Disciplined Convex Programming, version 2.1, CVX Research [code], <a href="https://cvxr.com/cvx/" target="_blank"/> (last access: 5 January 2026), 2014.

    </mixed-citation></ref-html>
<ref-html id="bib1.bib22"><label>Grant et al.(2006)Grant, Boyd, and Ye</label><mixed-citation>
      
Grant, M., Boyd, S., and Ye, Y.: Disciplined Convex Programming, in: Global Optimization: From Theory to Implementation, edited by Liberti, L. and Maculan, N., 155–210, Springer, Boston, MA, <a href="https://doi.org/10.1007/0-387-30065-1_7" target="_blank">https://doi.org/10.1007/0-387-30065-1_7</a>, 2006.

    </mixed-citation></ref-html>
<ref-html id="bib1.bib23"><label>Greene et al.(2017)Greene, Gwyther, and Blankenship</label><mixed-citation>
      
Greene, C. A., Gwyther, D. E., and Blankenship, D. D.: Antarctic Mapping Tools for MATLAB, Comp. Geosci., 104, 151–157, <a href="https://doi.org/10.1016/j.cageo.2016.08.003" target="_blank">https://doi.org/10.1016/j.cageo.2016.08.003</a>, 2017.

    </mixed-citation></ref-html>
<ref-html id="bib1.bib24"><label>Gudmundsson(2003)</label><mixed-citation>
      
Gudmundsson, G. H.: Transmission of basal variability to a glacier surface, J. Geophys. Res.: Solid Earth, 108,  2253, <a href="https://doi.org/10.1029/2002JB002107" target="_blank">https://doi.org/10.1029/2002JB002107</a>, 2003.

    </mixed-citation></ref-html>
<ref-html id="bib1.bib25"><label>Harrison et al.(1998)Harrison, Echelmeyer, and Larsen</label><mixed-citation>
      
Harrison, W., Echelmeyer, K., and Larsen, C.: Measurement of temperature in a margin of Ice Stream B, Antarctica: implications for margin migration and lateral drag, J. Glaciol., 44, 615–624, <a href="https://doi.org/10.3189/S0022143000002112" target="_blank">https://doi.org/10.3189/S0022143000002112</a>, 1998.

    </mixed-citation></ref-html>
<ref-html id="bib1.bib26"><label>Haseloff et al.(2015)Haseloff, Schoof, and Gagliardini</label><mixed-citation>
      
Haseloff, M., Schoof, C., and Gagliardini, O.: A boundary layer model for ice stream margins, J. Fluid Mech., 781, 353–387, <a href="https://doi.org/10.1017/jfm.2015.503" target="_blank">https://doi.org/10.1017/jfm.2015.503</a>, 2015.

    </mixed-citation></ref-html>
<ref-html id="bib1.bib27"><label>Haseloff et al.(2019)Haseloff, Hewitt, and Katz</label><mixed-citation>
      
Haseloff, M., Hewitt, I., and Katz, R.: Englacial pore water localizes shear in temperate ice stream margins, J. Geophys. Res.-Earth, 124, 2521–2541, <a href="https://doi.org/10.1029/2019JF005399" target="_blank">https://doi.org/10.1029/2019JF005399</a>, 2019.

    </mixed-citation></ref-html>
<ref-html id="bib1.bib28"><label>Hellmer et al.(2012)Hellmer, Kauker, Timmermann, Determann, and Rae</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.bib29"><label>Hewitt(2011)</label><mixed-citation>
      
Hewitt, I. J.: Modelling distributed and channelized subglacial drainage: the spacing of channels, J. Glaciol., 57, 302–314, <a href="https://doi.org/10.3189/002214311796405951" target="_blank">https://doi.org/10.3189/002214311796405951</a>, 2011.

    </mixed-citation></ref-html>
<ref-html id="bib1.bib30"><label>Holland et al.(2023)Holland, Bevan, and Luckman</label><mixed-citation>
      
Holland, P. R., Bevan, S. L., and Luckman, A. J.: Strong ocean melting feedback during the recent retreat of Thwaites Glacier, Geophys. Res. Lett., 50, e2023GL103088, <a href="https://doi.org/10.1029/2023GL103088" target="_blank">https://doi.org/10.1029/2023GL103088</a>, 2023.

    </mixed-citation></ref-html>
<ref-html id="bib1.bib31"><label>Hooke et al.(1979)Hooke, Raymond, Hotchkiss, and Gustafson</label><mixed-citation>
      
Hooke, R. L., Raymond, C. F., Hotchkiss, R. L., and Gustafson, R. J.: Calculations of velocity and temperature in a polar glacier using the finite-element method, J. Glaciol., 24, 131–146, <a href="https://doi.org/10.3189/S0022143000014696" target="_blank">https://doi.org/10.3189/S0022143000014696</a>, 1979.

    </mixed-citation></ref-html>
<ref-html id="bib1.bib32"><label>Iken(1981)</label><mixed-citation>
      
Iken, A.: The effect of the subglacial water pressure on the sliding velocity of a glacier in an idealized numerical model, J. Glaciol., 27, 407–421, <a href="https://doi.org/10.3189/S0022143000011448" target="_blank">https://doi.org/10.3189/S0022143000011448</a>, 1981.

    </mixed-citation></ref-html>
<ref-html id="bib1.bib33"><label>Jacobson and Raymond(1998)</label><mixed-citation>
      
Jacobson, H. and Raymond, C.: Thermal effects on the location of ice stream margins, J. Geophys. Res.: Solid Earth, 103, 12111–12122, <a href="https://doi.org/10.1029/98JB00574" target="_blank">https://doi.org/10.1029/98JB00574</a>, 1998.

    </mixed-citation></ref-html>
<ref-html id="bib1.bib34"><label>Jamieson et al.(2012)Jamieson, Vieli, Livingstone, Cofaigh, Stokes, Hillenbrand, and Dowdeswell</label><mixed-citation>
      
Jamieson, S. S. R., Vieli, A., Livingstone, S. J., Ó Cofaigh, C., Stokes, C., Hillenbrand, C.-D., and Dowdeswell, J. A.: Ice-stream stability on a reverse bed slope, Nat. Geosci., 5, 799–802, <a href="https://doi.org/10.1038/ngeo1600" target="_blank">https://doi.org/10.1038/ngeo1600</a>, 2012.

    </mixed-citation></ref-html>
<ref-html id="bib1.bib35"><label>Jordan et al.(2013)Jordan, Ferraccioli, Ross, Corr, Leat, Bingham, Rippin, le Brocq, and Siegert</label><mixed-citation>
      
Jordan, T. A., Ferraccioli, F., Ross, N., Corr, H. F., Leat, P. T., Bingham, R. G., Rippin, D. M., le Brocq, A., and Siegert, M. J.: Inland extent of the Weddell Sea Rift imaged by new aerogeophysical data, Tectonophysics, 585, 137–160, <a href="https://doi.org/10.1016/j.tecto.2012.09.010" target="_blank">https://doi.org/10.1016/j.tecto.2012.09.010</a>, 2013.

    </mixed-citation></ref-html>
<ref-html id="bib1.bib36"><label>Kamb(2001)</label><mixed-citation>
      
Kamb, B.: Basal zone of the West Antarctic ice streams and its role in lubrication of their rapid motion, in: The West Antarctic ice sheet: behavior and environment, 77, 157–199, <a href="https://doi.org/10.1029/AR077p0157" target="_blank">https://doi.org/10.1029/AR077p0157</a>, 2001.

    </mixed-citation></ref-html>
<ref-html id="bib1.bib37"><label>Kirchner et al.(2011)Kirchner, Hutter, Jakobsson, and Gyllencreutz</label><mixed-citation>
      
Kirchner, N., Hutter, K., Jakobsson, M., and Gyllencreutz, R.: Capabilities and limitations of numerical ice sheet models: A discussion for Earth-scientists and modelers, Quat. Sci. Rev., 30, 3691–3704, <a href="https://doi.org/10.1016/j.quascirev.2011.09.012" target="_blank">https://doi.org/10.1016/j.quascirev.2011.09.012</a>, 2011.

    </mixed-citation></ref-html>
<ref-html id="bib1.bib38"><label>Koellner et al.(2019)Koellner, Parizek, Alley, Muto, Holschuh, and Nowicki</label><mixed-citation>
      
Koellner, S., Parizek, B., Alley, R., Muto, A., Holschuh, N., and Nowicki, S.: The impact of spatially-variable basal properties on outlet glacier flow, Earth Planet. Sci. Lett., 515, 200–208, <a href="https://doi.org/10.1016/j.epsl.2019.03.026" target="_blank">https://doi.org/10.1016/j.epsl.2019.03.026</a>, 2019.

    </mixed-citation></ref-html>
<ref-html id="bib1.bib39"><label>Larour et al.(2012)Larour, Seroussi, Morlighem, and Rignot</label><mixed-citation>
      
Larour, E., Seroussi, H., Morlighem, M., and Rignot, E.: Continental scale, high order, high spatial resolution, ice sheet modeling using the Ice Sheet System Model (ISSM), J. Geophys. Res.-Earth, 117,  F01022, <a href="https://doi.org/10.1029/2011JF002140" target="_blank">https://doi.org/10.1029/2011JF002140</a>, 2012.

    </mixed-citation></ref-html>
<ref-html id="bib1.bib40"><label>Le Brocq et al.(2013)LeBrocq, Ross, Griggs, Bingham, Corr, Ferraccioli, Jenkins, Jordan, Payne, Rippin, and Siegert</label><mixed-citation>
      
Le Brocq, A. M., Ross, N., Griggs, J. A., Bingham, R. G., Corr, H. F. J., Ferraccioli, F., Jenkins, A., Jordan, T. A., Payne, A. J., Rippin, D. M., and Siegert, M. J.: Evidence from ice shelves for channelized meltwater flow beneath the Antarctic Ice Sheet, Nat. Geosci., 6, 945–948, <a href="https://doi.org/10.1038/ngeo1977" target="_blank">https://doi.org/10.1038/ngeo1977</a>, 2013.

    </mixed-citation></ref-html>
<ref-html id="bib1.bib41"><label>Lliboutry(1968)</label><mixed-citation>
      
Lliboutry, L.: General theory of subglacial cavitation and sliding of temperate glaciers, J. Glaciol., 7, 21–58, <a href="https://doi.org/10.3189/S0022143000020396" target="_blank">https://doi.org/10.3189/S0022143000020396</a>, 1968.

    </mixed-citation></ref-html>
<ref-html id="bib1.bib42"><label>MacGregor et al.(2013)MacGregor, Catania, Conway, Schroeder, Joughin, Young, Kempf, and Blankenship</label><mixed-citation>
      
MacGregor, J. A., Catania, G. A., Conway, H., Schroeder, D. M., Joughin, I., Young, D. A., Kempf, S. D., and Blankenship, D. D.: Weak bed control of the eastern shear margin of Thwaites Glacier, West Antarctica, J. Glaciol., 59, 900–912, <a href="https://doi.org/10.3189/2013JoG13J050" target="_blank">https://doi.org/10.3189/2013JoG13J050</a>, 2013.

    </mixed-citation></ref-html>
<ref-html id="bib1.bib43"><label>Meyer and Minchew(2018)</label><mixed-citation>
      
Meyer, C. R. and Minchew, B. M.: Temperate ice in the shear margins of the Antarctic Ice Sheet: Controlling processes and preliminary locations, Earth Planet. Sci. Lett., 498, 17–26, <a href="https://doi.org/10.1016/j.epsl.2018.06.028" target="_blank">https://doi.org/10.1016/j.epsl.2018.06.028</a>, 2018.

    </mixed-citation></ref-html>
<ref-html id="bib1.bib44"><label>Morland and Johnson(1980)</label><mixed-citation>
      
Morland, L. and Johnson, I.: Steady motion of ice sheets, J. Glaciol., 25, 229–246, <a href="https://doi.org/10.3189/S0022143000010467" target="_blank">https://doi.org/10.3189/S0022143000010467</a>, 1980.

    </mixed-citation></ref-html>
<ref-html id="bib1.bib45"><label>Muto et al.(2019)Muto, Anandakrishnan, Alley, Horgan, Parizek, Koellner, Christianson, and Holschuh</label><mixed-citation>
      
Muto, A., Anandakrishnan, S., Alley, R. B., Horgan, H. J., Parizek, B. R., Koellner, S., Christianson, K., and Holschuh, N.: Relating bed character and subglacial morphology using seismic data from Thwaites Glacier, West Antarctica, Earth Planet. Sci. Lett., 507, 199–206, <a href="https://doi.org/10.1016/j.epsl.2018.12.008" target="_blank">https://doi.org/10.1016/j.epsl.2018.12.008</a>, 2019.

    </mixed-citation></ref-html>
<ref-html id="bib1.bib46"><label>Nye(1952)</label><mixed-citation>
      
Nye, J. F.: The mechanics of glacier flow, J. Glaciol., 2, 82–93, <a href="https://doi.org/10.3189/S0022143000033967" target="_blank">https://doi.org/10.3189/S0022143000033967</a>, 1952.

    </mixed-citation></ref-html>
<ref-html id="bib1.bib47"><label>Nye(1957)</label><mixed-citation>
      
Nye, J. F.: The distribution of stress and velocity in glaciers and ice-sheets, Proc. R. Soc. Lond. Ser. A Math Phys. Sci., 239, 113–133, <a href="https://doi.org/10.1098/rspa.1957.0026" target="_blank">https://doi.org/10.1098/rspa.1957.0026</a>, 1957.

    </mixed-citation></ref-html>
<ref-html id="bib1.bib48"><label>Pattyn(2002)</label><mixed-citation>
      
Pattyn, F.: Transient glacier response with a higher-order numerical ice-flow model, J. Glaciol., 48, 467–477, <a href="https://doi.org/10.3189/172756502781831278" target="_blank">https://doi.org/10.3189/172756502781831278</a>, 2002.

    </mixed-citation></ref-html>
<ref-html id="bib1.bib49"><label>Perol and Rice(2015)</label><mixed-citation>
      
Perol, T. and Rice, J. R.: Shear heating and weakening of the margins of West Antarctic ice streams, Geophys. Res. Lett., 42, 3406–3413, <a href="https://doi.org/10.1002/2015GL063638" target="_blank">https://doi.org/10.1002/2015GL063638</a>, 2015.

    </mixed-citation></ref-html>
<ref-html id="bib1.bib50"><label>Perol et al.(2015)Perol, Rice, Platt, and Suckale</label><mixed-citation>
      
Perol, T., Rice, J. R., Platt, J. D., and Suckale, J.: Subglacial hydrology and ice stream margin locations, J. Geophys. Res.-Earth, 120, 1352–1368, <a href="https://doi.org/10.1002/2015JF003542" target="_blank">https://doi.org/10.1002/2015JF003542</a>, 2015.

    </mixed-citation></ref-html>
<ref-html id="bib1.bib51"><label>Platt et al.(2016)Platt, Perol, Suckale, and Rice</label><mixed-citation>
      
Platt, J. D., Perol, T., Suckale, J., and Rice, J. R.: Determining conditions that allow a shear margin to coincide with a Röthlisberger channel, J. Geophys. Res.-Earth, 121, 1273–1294, <a href="https://doi.org/10.1002/2015JF003707" target="_blank">https://doi.org/10.1002/2015JF003707</a>, 2016.

    </mixed-citation></ref-html>
<ref-html id="bib1.bib52"><label>Rathbun et al.(2008)Rathbun, Marone, Alley, and Anandakrishnan</label><mixed-citation>
      
Rathbun, A. P., Marone, C., Alley, R. B., and Anandakrishnan, S.: Laboratory study of the frictional rheology of sheared till, J. Geophys. Res.-Earth, 113,  F02020, <a href="https://doi.org/10.1029/2007JF000815" target="_blank">https://doi.org/10.1029/2007JF000815</a>, 2008.

    </mixed-citation></ref-html>
<ref-html id="bib1.bib53"><label>Raymond(1996)</label><mixed-citation>
      
Raymond, C.: Shear margins in glaciers and ice sheets, J. Glaciol., 42, 90–102, <a href="https://doi.org/10.3189/S0022143000030550" target="_blank">https://doi.org/10.3189/S0022143000030550</a>, 1996.

    </mixed-citation></ref-html>
<ref-html id="bib1.bib54"><label>Reeh and Paterson(1988)</label><mixed-citation>
      
Reeh, N. and Paterson, W.: Application of a flow model to the ice-divide region of Devon Island ice cap, Canada, J. Glaciol., 34, 55–63, <a href="https://doi.org/10.3189/S0022143000009060" target="_blank">https://doi.org/10.3189/S0022143000009060</a>, 1988.

    </mixed-citation></ref-html>
<ref-html id="bib1.bib55"><label>Rignot et al.(2011)Rignot, Mouginot, and Scheuchl</label><mixed-citation>
      
Rignot, E., Mouginot, J., and Scheuchl, B.: Ice flow of the Antarctic ice sheet, Science, 333, 1427–1430, <a href="https://doi.org/10.1126/science.1208336" target="_blank">https://doi.org/10.1126/science.1208336</a>, 2011.

    </mixed-citation></ref-html>
<ref-html id="bib1.bib56"><label>Ross et al.(2011)Ross, Siegert, Woodward, Smith, Corr, Bentley, Hindmarsh, King, and Rivera</label><mixed-citation>
      
Ross, N., Siegert, M. J., Woodward, J., Smith, A. M., Corr, H. F., Bentley, M. J., Hindmarsh, R. C., King, E. C., and Rivera, A.: Holocene stability of the Amundsen-Weddell ice divide, West Antarctica, Geology, 39, 935–938, <a href="https://doi.org/10.1130/G31920.1" target="_blank">https://doi.org/10.1130/G31920.1</a>, 2011.

    </mixed-citation></ref-html>
<ref-html id="bib1.bib57"><label>Ross et al.(2012)Ross, Bingham, Corr, Ferraccioli, Jordan, Le Brocq, Rippin, Young, Blankenship, and Siegert</label><mixed-citation>
      
Ross, N., Bingham, R., Corr, H. F. J., Ferraccioli, F., Jordan, T., Le Brocq, A., Rippin, D., Young, D., Blankenship, D., and Siegert, M. J.: Steep reverse bed slope at the grounding line of the Weddell Sea sector in West Antarctica, Nat. Geosci., 5, 393–396, <a href="https://doi.org/10.1038/ngeo1468" target="_blank">https://doi.org/10.1038/ngeo1468</a>, 2012.

    </mixed-citation></ref-html>
<ref-html id="bib1.bib58"><label>Ross et al.(2014)Ross, Jordan, Bingham, Corr, Ferraccioli, Brocq, Rippin, Wright, and Siegert</label><mixed-citation>
      
Ross, N., Jordan, T. A., Bingham, R. G., Corr, H. F. J., Ferraccioli, F., Le Brocq, A., Rippin, D. M., Wright, A. P., and Siegert, M. J.: The Ellsworth Subglacial Highlands: Inception and retreat of the West Antarctic Ice Sheet, Geol. Soc. Am. Bull., 126, 3–15, <a href="https://doi.org/10.1130/B30794.1" target="_blank">https://doi.org/10.1130/B30794.1</a>, 2014.

    </mixed-citation></ref-html>
<ref-html id="bib1.bib59"><label>Röthlisberger(1972)</label><mixed-citation>
      
Röthlisberger, H.: Water pressure in intra- and subglacial channels, J. Glaciol., 11, 177–203, <a href="https://doi.org/10.3189/S0022143000022188" target="_blank">https://doi.org/10.3189/S0022143000022188</a>, 1972.

    </mixed-citation></ref-html>
<ref-html id="bib1.bib60"><label>Sanderson et al.(2023)Sanderson, Winter, Callard, Napoleoni, Ross, Jordan, and Bingham</label><mixed-citation>
      
Sanderson, R. J., Winter, K., Callard, S. L., Napoleoni, F., Ross, N., Jordan, T. A., and Bingham, R. G.: Englacial architecture of Lambert Glacier, East Antarctica, The Cryosphere, 17, 4853–4871, <a href="https://doi.org/10.5194/tc-17-4853-2023" target="_blank">https://doi.org/10.5194/tc-17-4853-2023</a>, 2023.

    </mixed-citation></ref-html>
<ref-html id="bib1.bib61"><label>Scambos et al.(2004)Scambos, Bohlander, Raup, and Haran</label><mixed-citation>
      
Scambos, T., Bohlander, J., Raup, B., and Haran, T.: Glaciological characteristics of Institute Ice Stream using remote sensing, Antarct. Sci., 16, 205–213, <a href="https://doi.org/10.1017/S0954102004001919" target="_blank">https://doi.org/10.1017/S0954102004001919</a>, 2004.

    </mixed-citation></ref-html>
<ref-html id="bib1.bib62"><label>Schoof(2004)</label><mixed-citation>
      
Schoof, C.: On the mechanics of ice-stream shear margins, J. Glaciol., 50, 208–218, <a href="https://doi.org/10.3189/172756504781830024" target="_blank">https://doi.org/10.3189/172756504781830024</a>, 2004.

    </mixed-citation></ref-html>
<ref-html id="bib1.bib63"><label>Schoof(2005)</label><mixed-citation>
      
Schoof, C.: The effect of cavitation on glacier sliding, Proc. R. Soc. Lond. Ser. A Math Phys. Sci., 461, 609–627, <a href="https://doi.org/10.1098/rspa.2004.1350" target="_blank">https://doi.org/10.1098/rspa.2004.1350</a>, 2005.

    </mixed-citation></ref-html>
<ref-html id="bib1.bib64"><label>Schoof(2006a)</label><mixed-citation>
      
Schoof, C.: A variational approach to ice stream flow, J. Fluid Mech., 556, 227–251, <a href="https://doi.org/10.1017/S0022112006009591" target="_blank">https://doi.org/10.1017/S0022112006009591</a>, 2006a.

    </mixed-citation></ref-html>
<ref-html id="bib1.bib65"><label>Schoof(2006b)</label><mixed-citation>
      
Schoof, C.: Variational methods for glacier flow over plastic till, J. Fluid Mech., 555, 299–320, <a href="https://doi.org/10.1017/S0022112006009104" target="_blank">https://doi.org/10.1017/S0022112006009104</a>, 2006b.

    </mixed-citation></ref-html>
<ref-html id="bib1.bib66"><label>Schoof(2007)</label><mixed-citation>
      
Schoof, C.: Ice sheet grounding line dynamics: Steady states, stability, and hysteresis, J. Geophys. Res., 112, F03S28, <a href="https://doi.org/10.1029/2006JF000664" target="_blank">https://doi.org/10.1029/2006JF000664</a>, 2007.

    </mixed-citation></ref-html>
<ref-html id="bib1.bib67"><label>Schoof(2012)</label><mixed-citation>
      
Schoof, C.: Thermally driven migration of ice-stream shear margins, J. Fluid Mech., 712, 552–578, <a href="https://doi.org/10.1017/jfm.2012.438" target="_blank">https://doi.org/10.1017/jfm.2012.438</a>, 2012.

    </mixed-citation></ref-html>
<ref-html id="bib1.bib68"><label>Schroeder et al.(2014)Schroeder, Blankenship, Young, and Quartini</label><mixed-citation>
      
Schroeder, D. M., Blankenship, D. D., Young, D. A., and Quartini, E.: Evidence for elevated and spatially variable geothermal flux beneath the West Antarctic Ice Sheet, P. Natl. Acad. Sci. USA, 111, 9070–9072, <a href="https://doi.org/10.1073/pnas.1405184111" target="_blank">https://doi.org/10.1073/pnas.1405184111</a>, 2014.

    </mixed-citation></ref-html>
<ref-html id="bib1.bib69"><label>Siegert et al.(2016)Siegert, Ross, Li, Schroeder, Rippin, Ashmore, Bingham, and Gogineni</label><mixed-citation>
      
Siegert, M. J., Ross, N., Li, J., Schroeder, D. M., Rippin, D., Ashmore, D., Bingham, R., and Gogineni, P.: Subglacial controls on the flow of Institute Ice Stream, West Antarctica, Ann. Glaciol., 57, 19–24, <a href="https://doi.org/10.1017/aog.2016.17" target="_blank">https://doi.org/10.1017/aog.2016.17</a>, 2016.

    </mixed-citation></ref-html>
<ref-html id="bib1.bib70"><label>Smith et al.(2020)Smith, Fricker, Gardner, Medley, Nilsson, Paolo, Holschuh, Adusumilli, Brunt, Csatho et al.</label><mixed-citation>
      
Smith, B., Fricker, H. A., Gardner, A. S., Medley, B., Nilsson, J., Paolo, F. S., Holschuh, N., Adusumilli, S., Brunt, K., Csatho, B., Harbeck, K., Markus, T., Neumann, T., Siegfried, M. R., and Zwally, H. J.: Pervasive ice sheet mass loss reflects competing ocean and atmosphere processes, Science, 368, 1239–1242, <a href="https://doi.org/10.1126/science.aaz5845" target="_blank">https://doi.org/10.1126/science.aaz5845</a>, 2020.

    </mixed-citation></ref-html>
<ref-html id="bib1.bib71"><label>Smith et al.(2009)Smith, A. Fricker, Joughin, and Slawek</label><mixed-citation>
      
Smith, B. E., Fricker, H. A., Joughin, I. R., and Tulaczyk, S.: An inventory of active subglacial lakes in Antarctica detected by ICESat (2003–2008), J. Glaciol., 55, 573–595, <a href="https://doi.org/10.3189/002214309789470879" target="_blank">https://doi.org/10.3189/002214309789470879</a>, 2009.

    </mixed-citation></ref-html>
<ref-html id="bib1.bib72"><label>Suckale et al.(2014)Suckale, Platt, Perol, and Rice</label><mixed-citation>
      
Suckale, J., Platt, J. D., Perol, T., and Rice, J. R.: Deformation-induced melting in the margins of the West Antarctic ice streams, J. Geophys. Res., 119, 1004–1025, <a href="https://doi.org/10.1002/2013JF003008" target="_blank">https://doi.org/10.1002/2013JF003008</a>, 2014.

    </mixed-citation></ref-html>
<ref-html id="bib1.bib73"><label>Summers et al.(2023)Summers, Elsworth, Dow, and Suckale</label><mixed-citation>
      
Summers, P. T., Elsworth, C. W., Dow, C. F., and Suckale, J.: Migration of the shear margins at Thwaites Glacier: Dependence on basal conditions and testability against field data, J. Geophys. Res.-Earth, 128, e2022JF006958, <a href="https://doi.org/10.1029/2022JF006958" target="_blank">https://doi.org/10.1029/2022JF006958</a>, 2023.

    </mixed-citation></ref-html>
<ref-html id="bib1.bib74"><label>Thomas(1979)</label><mixed-citation>
      
Thomas, R. H.: The dynamics of marine ice sheets, J. Glaciol., 24, 167–177, <a href="https://doi.org/10.3189/S0022143000014726" target="_blank">https://doi.org/10.3189/S0022143000014726</a>, 1979.

    </mixed-citation></ref-html>
<ref-html id="bib1.bib75"><label>Tulaczyk et al.(2000a)Tulaczyk, Kamb, and Engelhardt</label><mixed-citation>
      
Tulaczyk, S., Kamb, W. B., and Engelhardt, H. F.: Basal mechanics of Ice Stream B, west Antarctica: 2. Undrained plastic bed model, J. Geophys. Res., 105, 483–494, <a href="https://doi.org/10.1029/1999JB900328" target="_blank">https://doi.org/10.1029/1999JB900328</a>, 2000a.

    </mixed-citation></ref-html>
<ref-html id="bib1.bib76"><label>Tulaczyk et al.(2000b)Tulaczyk, Kamb, and Engelhardt</label><mixed-citation>
      
Tulaczyk, S., Kamb, W. B., and Engelhardt, H. F.: Basal mechanics of Ice Stream B, west Antarctica: 1. Till mechanics, J. Geophys. Res., 105, 463–481, <a href="https://doi.org/10.1029/1999JB900329" target="_blank">https://doi.org/10.1029/1999JB900329</a>, 2000b.

    </mixed-citation></ref-html>
<ref-html id="bib1.bib77"><label>Weertman(1957)</label><mixed-citation>
      
Weertman, J.: On the sliding of glaciers, J. Glaciol., 3, 33–38, <a href="https://doi.org/10.3189/S0022143000024709" target="_blank">https://doi.org/10.3189/S0022143000024709</a>, 1957.

    </mixed-citation></ref-html>
<ref-html id="bib1.bib78"><label>Weertman(1974)</label><mixed-citation>
      
Weertman, J.: Stability of the junction of an ice sheet and an ice shelf, J. Glaciol., 13, 3–11, <a href="https://doi.org/10.3189/S0022143000023327" target="_blank">https://doi.org/10.3189/S0022143000023327</a>, 1974.

    </mixed-citation></ref-html>
<ref-html id="bib1.bib79"><label>Zoet and Iverson(2016)</label><mixed-citation>
      
Zoet, L. K. and Iverson, N. R.: Rate-weakening drag during glacier sliding, J. Geophys. Res.-Earth, 121, 1206–1217, <a href="https://doi.org/10.1002/2016JF003909" target="_blank">https://doi.org/10.1002/2016JF003909</a>, 2016.

    </mixed-citation></ref-html>--></article>
