Review of the revised version of "Explicit silicate cycling in the Kiel Marine
Biogeochemistry Model, version 3 (KMBM3) embedded in the UVic ESCM
version 2.9" by Kvale et al., submitted to Geoscientific Model
Development in 2021
Main comments and recommendation
--------------------------------
This is the second version of this manuscript that I review, and the
authors have taken many of my comments to the previous version into
account. My two main points of concern about the last version were
unclear or wrong explanations in the model description and
inconsistencies in the described global Si fluxes and the comparison
with observation based estimates. I still think these two points need
a bit of work, and so I would recommend that the manuscript should be
further revised before it can be published.
While most of the remaining unclear points in the model description
are rather minor, one concerns a central part of the manuscript, the
implicitly treated dissolution of the sinking opal. As it is one of
the main points of the manuscript that this formulation of opal
dissolution improves the modeled global silicic acid distribution,
this is an important point.
The dissolution of opal is described in equation 32. The equation has
changed with respect to the last version by the multiplication with
depth (z) within the flux divergence term on the right hand side, so
that now the production Pr(Opal) and dissolution Di(Opal) have the
same unit, which makes sense. In the manuscript is is stated that the
unit of Di(Opal) is mol Si m^{-3}, but I think, as it is a rate, it
should be mol Si m^{-3} yr^{-1}. But more importantly I still do not
understand the rationale behind equation 32, for two reasons:
- Is the chosen form of the flux compatible with a reasonable
assumption on the implicit vertical distribution of sinking opal, or
of the sinking flux? I do not think so. As a simple example, assume
an ocean with uniform temperature (which is a good approximation to
the ocean below 2 km depth). Then the immediate consequence of
eq. 32 is that opal dissolution Di(Opal) is constant throughout the
water column. If the concentration of opal decreases with depth (as
it must, because of dissolution), then that would imply that the
dissolution rate needs to increase with depth. Or, of the
dissolution rate is constant, then opal concentration must also be
constant vertically. Of course, the chosen temperature dependence
will lead to a decrease of the scaled flux (the quantity within the
d/dz) with depth, and thus to a more realistic vertical behaviour of
the flux. But it still remains unclear to me how the
proportionality of the flux to depth z can follow from an
assumption on the steady-state vertical distribution of
opal biomass and dissolution rate.
- Also, if the parameterization in eq. 32 would be compatible with a
vertical distribution of sinking opal, then it would automatically
follow that the integral over the total water column dissolution
would be smaller or equal to the integrated production, independent
of the chosen dissolution rate and its temperature dependence. In
the current formulation, however, this is not the case: Here it has
to be ensured (probably by making lambda small enough) that the flux
(scaled by the vertically integrated total production) at the bottom
of the ocean lambda*depth/w * exp(T/Td) is not larger than one.
I think the authors should discuss the relationship of their chosen
flux parameterization to assumptions on sinking speed and the vertical
distribution of opal. Probably this cannot be done in a strict
analytical way (with a temperature-dependent dissolution rate, an
analytical solution for a steady state flux profile is hard to
derive). But a qualitative idea why this flux profile was chosen
should be possible.
My other main criticicm of the last manuscript version were
inconsistencies or confusing statements on the global Si fluxes. Theu
authors have changed their notation now to Tmol Si/yr, which makes the
comparison with the numbers in Treguer et al, 2021 easier. But in
doing so, also some of the stements on numbers have also changed, and
to me it was really unclear now how the different terms are defined:
- In the new manuscript (line 338-340 and table 6) it is said that the
total opal production is 127.9 Tmol Si/yr, which would be low
against the estimate of 255+-52 Tmol/yr in Treguer et al., 2021. But
in the previous manuscript version the total production was stated
as 7.68 Pg Si/yr, which approximately translates to 270 Tmol Si/yr,
close to the Treguer et al. estimate. So, was the old number simply
a miscalculation? There are more differences in numbers between the
old and the new manuscript version, e.g. in the stated total
dissolution. Is this due to a different way that the term has been
defined?
- I also do not understand the relation between production, water
column dissolution and sediment export of opal. If we take the
statement from the new table 6 of total opal production of 127.9
Tmol Si/yr, and substract the total opal water column dissolution of
64.4 Tmol Si/yr, the difference (63.6 Tmol/yr) does not dissolve in
the water column and hence must be deposited on the sediment. How is
that compatible with the statement at the vertical flux of opal at
2000 m depth is only 19.46 Tmol Si/yr? Does that mean that more than
two thirds of the opal deposition into the sediment happen above
2000 m depth, i.e. on the shelves?
I think the authors should try to make it possible for the reader to
close the opal budget by stating a) how the total production of Si is
distributed between the upper 130m and deeper, b) how much of that
production dissolves with the layers 0m-130m, 130m-2000m, and below
2000m, c) how much opal is deposited into the sediment within these
three layers, and d) how much of the deposited opal is permanently
buried within these layers.
Minor comments
--------------
- line 82-83, "any opal that reaches the seafloor is replaced by
external sources": This contradicts the formulation of the simple
sediment scheme (lines 266 ff.), where it is explained that only a
fraction of what arrives at the seafloor is permanently buried and
needs to be replaced by external sources to keep the inventory
stable. Please clarify.
- in my last review I suggested to replace the word 'bioavailability'
for the Michaelis-Menten-kinetics-like term in nutrient uptake. This
has been done. The replacement 'uptake' (lines 107, 115, 117, ..)
is, however, equally misleading: the quantity that is described is a
dimensionless limitation term, uptake would imply some material
change, probaby in mol/volume/time.
- On line 117, it is said that the nutrient limitation terms for
nitrate, phosphate and silicic acid are multiplied with the maximum
potential growth rate. This is a bit misleading because it suggests
a multiplicative limitation by the different nutrients; later on
(eq. 13) it is clarified, that only the minimum of the limitation
terms, not their product, is used.
- in equation 8, the two numbers given are not pure numbers, but
should have a unit, namely that of a silicic acid concentration. I
would suggest these numbers are replaced by a symbol, and that these
numbers with their proper units are then added to the parameter
table. Besides, I had already mentioned in my last review, that
writing a decimal number as 8E-4, when you mean 8 times 10 to the
power of minus 4 is something that fortran understands, but that is
violating typographical conventions.
- the description of the dependency of the initial slope of the PI
curve and on the Chl-a:C ratio on the degree of iron limitation is
correct, but could use a bit more justification. The neglection of
the dependency of the Chl:C ratio on the light level is justified
here, I believe, because in the model, Chl is really only needed in
the surface layer (to calculate photosynthesis rates and to compare
with satellite estimates), and not at depth, where the dependency on
light becomes crucial.
- In line 159, the temperature-dependent fast remineralization is
called 'respiration', in contrast to the linear mortality. It is
maybe a bit picky, but while from an ecostem perspective bacterial
consumption is a respiration, from the phytoplankton perspective I
would rather describe it as excretion (because it is originally
dissolved organic carbon that is excreted and then respired by the
bacteria).
- in line 162-163, the quadratic mortality is called 'non-grazing
mortality' and contrasted with a self-grazing term GZ. Again it is
maybe a bit picky, but to my knowledge the quadratic zooplankton
mortality (the so-called closure term, Steele and Henderson,
1992) is exactly meant to represent grazing by higher trophic
levels. See Edwards and Yool (2000) for a discussion of the
different forms of closure.
- equation 21 contains several numbers, but some of the numbers are
meant to have units (the 8 should be 8 mmol m^{-3}), while others
are pure numbers.
- line 198: 'organic carbon' should probably be 'organic nitrogen'
- in equations 27 and 27, should the iron scavenging rate not also
depend on the concentration of (organic and CaCO3) particulates?
Also, the term [Fe]prime is not defined. It probably means the
concentration of 'free' inorganic Fe (often denoted [Fe']), probably
calculated from total dissolved iron and ligand concentration, but
this is not stated. I also do not understand why the prime is shown
here with letters, instead of Fe', as is the conventional usage.
- in equation 30, the production of opal is made proportional to the
mortality of diatoms, which makes sense, but one term is missing,
namely the 'bacterial loop' mortality term (mu* x DT). What happens
with the opal produced by diatoms that die by that pathway?
- line 223-224: it is unclear whether in 'average surface opal:free
detritus export value of 1' the detritus is meant in units of N or
C. My guess is that the 'value' (rather a 'ratio', I would say) is
meant to be in Si:C units, which implies a Si:N of larger than
6, in the range of observations.
- But that made me wonder: what is the non-limiting Si:N ratio implied
by equation 31? If Si>kSi and Fe>kFeDT, the Si:POC ratio is
0.5 and hence a Si:PON ratio is around 3. This seems high to me, as
Si:N in most diatoms under non-limiting conditions is about one.
- the authors have chosen to keep their habit to put some terms in
square brackes, which have a meaning of a rate of change of a
concentration (e.g. on the right-hand side of equation 39). Of
course the authors are free to use any notation that they want, but
I re-iterate that this notation is a) inconsistent with the
conventional usage of square brackets in chemical texts, where the
square brackets mean 'concentration of', i.e. [Fe] = concentration
of Fe, and b) that it is used inconsistently: In equation 31, for
example, [Fe] means concentration of Fe in micromol/m^3, while in
eq. 35 [Fe]_sed means a flux in micromol/m^2/yr, and in eq. 39
[Fe]_col means a rate of change of concentration in micromol
Fe/m^3/yr.
- in eq. 38, again, the two numbers are not pure numbers, but have a
unit.
- in eq. 42, the alkalinity change is assumed to be proportional to
the phosphate change with a proportionality factor of -R_C:P. I do
not understand the rationale for that factor. Yes, biogenic
phosphate and nitrate uptake change alkalinity. In Wolf-Gladrow et
al., 2007, it is stated that "assimilation of 1 mole of nitrogen
(atoms) leads to (i) an increase of alkalinity by 1 mole when
nitrate or nitrite is the N source, (ii) to a decrease of alkalinity
by 1 mole when ammonia is used, and (iii) to no change of alkalinity
when molecular nitrogen is the N source" and "Uptake of 1 mole of
phosphate (H3PO4, H2PO4, HPO42−, or PO43−) by algae in accordance
with the + nutrient-H -compensation principle increases alkalinity
by 1 mole per mole P (...). Please note that the change of TA is
independent of the phosphate species taken up by the cell". So I
think the proportionality factor should be rather minus (Redfield
ratio between N and P plus one).
References:
Edwards, A. M., & Yool, A. (2000). The role of higher predation in
plankton population models. Journal of Plankton Research, 22(6),
1085–1112. https://doi.org/10.1093/plankt/22.6.1085
Steele,J.H. and Henderson,E.W. (1992) The role of predation in
plankton models. Journal of Plankton Research, 14, 157–172.
Wolf-Gladrow, D. A., Zeebe, R. E., Klaas, C., Körtzinger, A., &
Dickson, A. G. (2007). Total alkalinity: The explicit conservative
expression and its application to biogeochemical processes. Marine
Chemistry, 106(1-2 SPEC. ISS.),
287–300. https://doi.org/10.1016/j.marchem.2007.01.006 |