Optimisation of the marine Nd isotope scheme in the ocean component of the FAMOUS general circulation model

Robinson, Suzanne; Ivanovic, Ruza; Gregoire, Lauren; Astfalck, Lachlan; van de Flierdt, Tina; Plancherel, Yves; Pöppelmeier, Frerk; Tachikawa, Kazuyo

doi:https://doi.org/10.5194/egusphere-2022-937

Preprints

https://doi.org/10.5194/egusphere-2022-937

Preprints

07 Oct 2022

| 07 Oct 2022

Optimisation of the marine Nd isotope scheme in the ocean component of the FAMOUS general circulation model

Suzanne Robinson, Ruza Ivanovic, Lauren Gregoire, Lachlan Astfalck, Tina van de Flierdt, Yves Plancherel, Frerk Pöppelmeier, and Kazuyo Tachikawa

Abstract. The neodymium (Nd) isotope composition (ε_Nd) of seawater can be used to trace large-scale ocean circulation features. Yet, due to the elusive nature of marine Nd cycling, particularly in discerning non-conservative particle-seawater interactions, there remains considerable uncertainty surrounding a complete description of marine Nd budgets. Here, we present an optimisation of the Nd isotope scheme within the fast coupled atmosphere-ocean general circulation model (FAMOUS), using a statistical emulator to explore the parametric uncertainty and optimal combinations of three key model inputs relating to: (1) the efficiency of reversible scavenging, (2) the magnitude of the seafloor benthic flux, and (3) a riverine source scaling, accounting for release of Nd from river sourced particulate material. Furthermore, a suite of sensitivity tests provide insight on the regional mobilisation and spatial extent (i.e., testing a margin-constrained versus a seafloor-wide benthic flux) of certain reactive sediment components. In the calibrated scheme, the global marine Nd inventory totals 4.27 × 10¹² g and has a mean residence time of 727 years. Atlantic Nd isotope distributions are represented well, and the weak sensitivity of North Atlantic Deep Water to highly unradiogenic sedimentary sources implies an abyssal benthic flux is of secondary importance in determining the water mass ε_Nd properties under the modern vigorous circulation condition. On the other hand, Nd isotope distributions in the North Pacific are 3 to 4 ε_Nd-units too unradiogenic compared to water measurements, and our simulations indicate that a spatially uniform flux of bulk sediment ε_Nd does not sufficiently capture the mobile sediment components interacting with seawater. Our results of sensitivity tests suggest that there are distinct regional differences in how modern seawater acquires its ε_Nd signal, in part relating to the complex interplay of Nd addition and water advection.

Received: 15 Sep 2022 – Discussion started: 07 Oct 2022

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 preprint. The responsibility to include appropriate place names lies with the authors.

Download & links

Preprint (PDF, 6397 KB)

Supplement (8696 KB)

Download & links

Suzanne Robinson, Ruza Ivanovic, Lauren Gregoire, Lachlan Astfalck, Tina van de Flierdt, Yves Plancherel, Frerk Pöppelmeier, and Kazuyo Tachikawa

Interactive discussion

Status: closed

RC1: 'Comment on egusphere-2022-937', Anonymous Referee #1, 21 Dec 2022

Review Robinson et al, Optimisation of the marine Nd isotope scheme in the ocean component of the FAMOUS general circulation model This paper present an optimization procedure of the Neodymium scheme implemented in the FAMOUS ocean general circulation model. It provides very interesting insights on Nd cycle and sensitivity to scavenging and sedimentary and river sources. The approach is performed with a statistical emulator that is convenient for the scientific goal attempted, and derives analysis that are in accordance with the procedure. However the approach has two drawbacks which limit the scope of the article:First it does not include the Aeolian dust input in the optimization procedure. This source is also subject to large uncertainties in its parameterization, in particular in the value of the solubility in dust where the values âârange between 2 and 50%. This source has an impact on a much wider spatial extension than the rivers. It also substantially modifies the surface concentration and isotopic Nd values ââwhich have an important weight in the optimization procedure. This arbitrary choice in the method is prejudicial and not justified in the document, and potentially tends to overestimate the role of rivers (Lines 477- 490: improvement in surface Nd concentration and isotopic composition is obtained with enhanced river inputs. However, dust deposition, as it strongly impacts surface modelling results, could moderate this conclusion if it was also included in the optimization procedure)

Secondly North Atlantic simulated isotopic composition is too radiogenic compared to the observation (Lines 545-550 ; 565-568; 690-707; 840-845). Radiogenic values simulated in the North Atlantic is more probably a consequence of a too strong exchange around Island which is highly radiogenic and influences (too much) the whole north Atlantic ocean Nd isotopic composition distribution.

Sensitivity tests in the north Atlantic are a good idea (section 4), but the strategy is not satisfying. It has an interest to test the flux around the Labrador Sea, but it generates little sensitivity. Tests around Island should have generated a higher sensitivity; the signal is highly radiogenic and advected through the subpolar circulation. This strategy lead to erroneous conclusions upon the sensitivity of north Atlantic water to the local sources, that are repeated all along the paper.

For these reasons, I recommend a major revision of this paper before publication.

Minor point:

Comparison of modelled inventory with inventory derived from observation (line 499): What is the uncertainty on the inventory derived from the observations, in order to justify that modeled inventory are underestimating or overestimating it (especially REF-CONC2)

Citation: https://doi.org/10.5194/egusphere-2022-937-RC1
RC2:
'Comment on egusphere-2022-937', Jianghui Du, 17 Feb 2023
Review of “Optimisation of the marine Nd isotope scheme in the ocean component of the FAMOUS general circulation model” by Robinson et al submitted to Biogeoscience.
In this manuscript, the authors presented an optimized version of the marine Nd cycle implemented in the FAMOUS model. This work builds on the companion study that has performed parameter tuning to some degree but uses a statistical approach and ensemble simulation for more robust optimization, while further testing the model sensitivity to sedimentary flux and its eNd.
I think the authors did a good job toward a more process-oriented model sensitivity study of the marine Nd cycle, which is a step forward. The lessons we learn from such sensitivity results are more valuable than simply arriving at an optimized solution, and can guide further process-based studies. However, I think the study still missed the great opportunity to truly explore the parameter space as constrained by observations of sedimentary flux and particle scavenging. It is not clear what is the rationale for the choices of parameter space, given that they are not compatible with existing measurements of these parameters. Also, Why is a global model considered optimized, when it only tries to optimize the Atlantic, at a cost of the Pacific, which has an Nd reservoir that is 3 times that of the Atlantic? This Pacific problem in GCM models of Nd has existed for 15 years since Jones et al (2008), despite that processes-based studies and sensitivity tests point to a clear direction of a solution. The Pacific problem is not a Pacific-only problem but as the model tests have shown, a global problem that also impacts the Atlantic through circulation. I think it is time to resolve this problem and hope my suggestions will help the authors achieve this, which would be a major step in modeling the marine Nd cycle.
Major points:
Choices of parameter values

Nd inventory
This study chooses a global Nd inventory of 4.3×10¹² g from Tachikawa et al (2003) as a tuning target. Tachikawa et al (2003) did an excellent job of compiling data and estimating the ocean Nd budget using the data available to them, but it’s safe to say that most of the seawater data available today are published after 2003. So the Nd inventory needs to be updated. Du et al. (2020 QSR) updated the global budget using the data up to 2019, and the resulting global ocean Nd inventory is 5.6×10¹² g based on volumetrically binning the data, much bigger than Tachikawa et al (2003). This number needs update again in 2023 given quite a few GEOTRACE transects have been published since 2019. However, in the optimization ensemble, the target Nd inventory is a maximum 5×10¹² g. This raises the question of how reasonable the optimized results are when using the Tachikawa et al (2003) inventory as the tuning target. Is it possible that the Nd sources in the current optimized model are underestimated because of low target inventory?
[Nd_p]/[Nd_d]
The parameter range for [Nd_p]/[Nd_d] is 0.001-0.006 in this study. The authors stated in the companion paper that this range “considers the few direct observations of [Nd]_p/[Nd]_d (Jeandel et al., 1995; Stichel et al., 2020; Zhang et al., 2008)”. But this is not true. Measurements of [Nd_p]/[Nd_d] have existed since the classic studies of (Sholkovitz et al., 1994; Bertram and Elderfield, 1993), and the number of data has grown considerably in recent years thanks to GEOTRACES, for example, GP15 reported full transect of particulate and dissolved Nd (Haley et al., 2021; Lam et al., 2018). Old or new these studies consistently show that [Nd_p]/[Nd_d] is on the order of 0.01, much higher than what’s used in this study. In Stichel et al., 2020, which is cited in the companion paper, [Nd_p]/[Nd_d] in the North Atlantic has a mean of 0.03 and a median of 0.02. Note that here I use the non-lithogenic [Nd_p] in this calculation, so only the exchangeable Nd_p is considered.
I noticed the reply of the authors to the same comment raised by Stichel on the companion paper (Robinson et al, 2022). Particle studies normally report both total Nd_p and non-lithogenic exchangeable Nd_p, based on leaching the labile fraction or removing the lithogenic fraction using detrital correction. Yes, it is the [Nd_p]_exchangible/[Nd_d] not [Nd_p]_total/[Nd_d] that should be used in the model. And the values I refer to here are all [Nd_p]_exchangible/[Nd_d] and they are ~0.01 according to these studies.
Sedimentary flux
The pore water data-constrained diffusive sedimentary Nd flux is 11~16×10⁹ g/yr (Du et al., 2020; Abbott et al., 2015). This does not include advective pore water fluxes such as bio-irrigation, which can be comparable to or larger than the diffusive flux (Du et al., 2022; Deng et al., 2022). In this study, the tested range of sedimentary flux is only 1.5~6×10⁹ g/yr. Admitting the possible large uncertainty of the sedimentary flux, the parameter range should at least bracket this data-constrained value.
Residence time and regional sedimentary influence

In the optimized model both the Atlantic and Pacific seawater eNd endmembers are much less extreme than measured, resulting in a global distribution of eNd being too uniform than observed. Thus, ocean circulation plays a more important role in the model than in reality, which causes the low sensitivity of the modeled seawater eNd to localized sedimentary eNd change in the sensitivity tests.
I think this is because the parameter ranges chosen by the study are too restricted and do not bracket the natural ranges observed. Thus the residence time of Nd is too long in the model, such that the regional sedimentary influence cannot manifest clearly. I suggest that a combination of higher [Nd_p]/[Nd_d], higher sedimentary flux, and shorter residence time are needed to capture the heterogeneity of eNd distribution in the ocean. Here are my reasons.
First, although box models cannot be used to study the spatial distribution of eNd in the ocean, they are better at estimating global properties, such as residence time. This is because box models strictly follow the data-constrained requirement of global mass balance, whereas the GCMs do not, as shown by the failure of GCMs to correctly model the Pacific eNd. To me, the box model-constrained Nd residence time of 400~500 years is better than the GCM-constrained residence time of ~700 years. This shorter residence time is supported by multiple lines of evidence. Inverse model-based sensitivity constraint shows optimal results at Nd residence time of 400~500 years (Siddall et al., 2008; Pasquier et al., 2022). Th-constrained particle cycling model shows Nd residence time of ~100 years in the Atlantic (Hayes et al., 2018). Moreover, in FAMOUS the residence time decreases with an increase [Nd_p]/[Nd_d], if using the realistic value of 0.01~0.02, FAMOUS will likely produce a shorter residence time that is close to 400 years.
Abyssal sedimentary flux

The authors argue that “perhaps too much emphasis has recently been placed upon exclusively resolving nonconservative interactions of an abyssal deep seafloor flux to solve the Nd paradox”. I suggest the opposite is true.
Ever since the early GCM models (Jones et al., 2008), GCMs have struggled to correctly model the abyssal Pacific eNd, even though a solution to this problem was presented by studies of sediment and pore water Nd. Why not increase abyssal sedimentary Nd flux to what’s estimated by these studies, which will lead to better Pacific model results?
Say that the authors correctly model the Pacific eNd, then their Atlantic results will worsen, as the radiogenic Pacific water will make the AABW endmember more radiogenic, thus leading to more radiogenic NADW, which is already too radiogenic in the model compared to the data. Doesn’t this imply that the abyssal Atlantic sedimentary flux is also underestimated? Why not test the scenario of combined high abyssal sedimentary flux with extreme eN as suggested by observations? For example, the results from GEOTRACES GP03 show that along the Deep Western Boundary Current in the North Atlantic, there’s an input of highly unradiogenic sedimentary flux related to the abyssal Nepheloid layers (Jaume-Seguí et al., 2021).
Minor points:
L52. Authigenic phosphate is also a major sink.
L232. This is not an equation/function.
L315. But isn’t the abyssal Pacific the largest reservoir of Nd?
L390. In reality, fsed and the scavenged flux are not independent. Part of fsed results from the regeneration of the scavenged flux that enters the sediments; part of fsed comes from new sources (e.g. reactive detritus) within the sediments (Du et al., 2022). So sedimentary flux is closely tied to the residence time. It is only because in the model you have no interactive sedimentary component that you have to specify fsed as an independent parameter. I suggest the difference between model and reality should be made clear.
L517. (Sholkovitz et al., 1994) showed that [Nd_p]/[Nd_d] was lower at the surface than at depth in the Atlantic. This model misfit is probably caused by using a vertically constant [Nd_p]/[Nd_d].
L564. The GEOTRACES GA02 full Atlantic meridional transect of eNd was published last year (Wu et al., 2022). This should be plotted along Fig 6. It is worth noting here that the Southern ocean endmember is too unradiogenic (-11) in the model compared to data (-9).
L572. I would rather argue that AABW is too unradiogenic because the Pacific endmember is too unradiogenic in the model. With adequate Pacific endmember, the model may get the correct eNd for AABW even if the benthic flux in the Atlantic is unradiogenic.
L693-704. It is also likely that the local Nd residence time is too long for the circulation signal to overcome the sedimentary influence in the model. See major points.
L771-780. It is important to evaluate more in detail two scenarios of why NADW eNd was fitted better in this margin-only scenario: (1) PDW/AABW has become much more unradiogenic because abyssal sedimentary flux in the Pacific was reduced, propagating this signal to the abyssal Atlantic, (2) abyssal sedimentary flux in the Atlantic is not adequate. The implications are very different regarding the importance of the abyssal sedimentary source. Ideally, sensitivity tests should include ones that limit sedimentary flux to the margins only in the Pacific or only in the Atlantic.
L799. That’s very well said.
Abbott, A. N., Haley, B. A., McManus, J., and Reimers, C. E.: The sedimentary flux of dissolved rare earth elements to the ocean, Geochim. Cosmochim. Acta, 154, 186–200, https://doi.org/10.1016/j.gca.2015.01.010, 2015.
Bertram, C. J. and Elderfield, H.: The geochemical balance of the rare earth elements and neodymium isotopes in the oceans, Geochim. Cosmochim. Acta, 57, 1957–1986, https://doi.org/10.1016/0016-7037(93)90087-D, 1993.
Deng, K., Yang, S., Du, J., Lian, E., and Vance, D.: Dominance of benthic flux of REEs on continental shelves: implications for oceanic budgets, Geochem. Perspect. Lett., 22, 26–30, https://doi.org/10.7185/geochemlet.2223, 2022.
Du, J., Haley, B. A., and Mix, A. C.: Evolution of the Global Overturning Circulation since the Last Glacial Maximum based on marine authigenic neodymium isotopes, Quat. Sci. Rev., 241, 106396, https://doi.org/10.1016/j.quascirev.2020.106396, 2020.
Du, J., Haley, B. A., Mix, A. C., Abbott, A. N., McManus, J., and Vance, D.: Reactive-transport modeling of neodymium and its radiogenic isotope in deep-sea sediments: The roles of authigenesis, marine silicate weathering and reverse weathering, Earth Planet. Sci. Lett., 596, 117792, https://doi.org/10.1016/j.epsl.2022.117792, 2022.
Haley, B. A., Wu, Y., Muratli, J. M., Basak, C., Pena, L. D., and Goldstein, S. L.: Rare earth element and neodymium isotopes of the eastern US GEOTRACES Equatorial Pacific Zonal Transect (GP16), Earth Planet. Sci. Lett., 576, 117233, https://doi.org/10.1016/j.epsl.2021.117233, 2021.
Hayes, C. T., Anderson, R. F., Cheng, H., Conway, T. M., Edwards, R. L., Fleisher, M. Q., Ho, P., Huang, K.-F., John, S. G., Landing, W. M., Little, S. H., Lu, Y., Morton, P. L., Moran, S. B., Robinson, L. F., Shelley, R. U., Shiller, A. M., and Zheng, X.-Y.: Replacement Times of a Spectrum of Elements in the North Atlantic Based on Thorium Supply, Glob. Biogeochem. Cycles, 32, 1294–1311, https://doi.org/10.1029/2017GB005839, 2018.
Jaume-Seguí, M., Kim, J., Pena, L. D., Goldstein, S. L., Knudson, K. P., Yehudai, M., Hartman, A. E., Bolge, L., and Ferretti, P.: Distinguishing Glacial AMOC and Interglacial Non-AMOC Nd Isotopic Signals in the Deep Western Atlantic Over the Last 1 Myr, Paleoceanogr. Paleoclimatology, 36, e2020PA003877, https://doi.org/10.1029/2020PA003877, 2021.
Jones, K. M., Khatiwala, S. P., Goldstein, S. L., Hemming, S. R., and van de Flierdt, T.: Modeling the distribution of Nd isotopes in the oceans using an ocean general circulation model, Earth Planet. Sci. Lett., 272, 610–619, https://doi.org/10.1016/j.epsl.2008.05.027, 2008.
Lacan, F. and Jeandel, C.: Acquisition of the neodymium isotopic composition of the North Atlantic Deep Water, Geochem. Geophys. Geosystems, 6, Q12008, https://doi.org/10.1029/2005GC000956, 2005.
Lam, P. J., Lee, J.-M., Heller, M. I., Mehic, S., Xiang, Y., and Bates, N. R.: Size-fractionated distributions of suspended particle concentration and major phase composition from the U.S. GEOTRACES Eastern Pacific Zonal Transect (GP16), Mar. Chem., 201, 90–107, https://doi.org/10.1016/j.marchem.2017.08.013, 2018.
Pasquier, B., Hines, S. K. V., Liang, H., Wu, Y., Goldstein, S. L., and John, S. G.: GNOM v1.0: an optimized steady-state model of the modern marine neodymium cycle, Geosci. Model Dev., 15, 4625–4656, https://doi.org/10.5194/gmd-15-4625-2022, 2022.
Sholkovitz, E. R., Landing, W. M., and Lewis, B. L.: Ocean particle chemistry: The fractionation of rare earth elements between suspended particles and seawater, Geochim. Cosmochim. Acta, 58, 1567–1579, https://doi.org/10.1016/0016-7037(94)90559-2, 1994.
Siddall, M., Khatiwala, S., van de Flierdt, T., Jones, K., Goldstein, S. L., Hemming, S., and Anderson, R. F.: Towards explaining the Nd paradox using reversible scavenging in an ocean general circulation model, Earth Planet. Sci. Lett., 274, 448–461, https://doi.org/10.1016/j.epsl.2008.07.044, 2008.
Wu, Y., Pena, L. D., Anderson, R. F., Hartman, A. E., Bolge, L. L., Basak, C., Kim, J., Rijkenberg, M. J. A., de Baar, H. J. W., and Goldstein, S. L.: Assessing neodymium isotopes as an ocean circulation tracer in the Southwest Atlantic, Earth Planet. Sci. Lett., 599, 117846, https://doi.org/10.1016/j.epsl.2022.117846, 2022.
Citation: https://doi.org/10.5194/egusphere-2022-937-RC2
RC3:
'Comment on egusphere-2022-937', Anonymous Referee #3, 17 Feb 2023
Review of "Optimization of the marine Nd isotope scheme in the ocean component of the FAMOUS general circulation model" by Robinson et al., submitted to EGUsphere, Feb. 2023.
This work presents the incorporation of Nd isotopes into the FAMOUS GCM, including optimization and validation. They authors evaluate the model and do some sensitivity tests, particularly on the dichotomy whether a benthic Nd source flux is ubiquitous or confined to the margins. The authors conclude that the model is largely robust, but that outstanding questions regarding a few key boundary conditions and processes remain.
Overall, this is a well-written ms. I think some further editing (culling) would make it stronger, but it was not challenging to follow the discourse. The figures were well made, but I had a difficult time with them: the 3D plot (Fig.1) is without droplines and is very difficult to evaluate. Similar, but less so, for the heat data on the 2D plots (Figs.3 & 4). To the point where these figures do not add anything to help understand the text. And I have a personal dislike of model section/point data figures (Figs. 6, 8, 9, 10). I just can’t properly evaluate the data:model fit from these figures. Why not show the data as a scatterplot?
Overall, I thought the ms. was well-done and offers a contribution to the field, largely in terms of description/validation of a model (for future use) and directions suggested for future research. I recommend publication, perhaps with consideration of the following questions/comments:
L380 (and discussion): Does the evaluation of Nd(I) need to be done on a full-model, or could it be done in the simple box model the authors use to steer their GCM runs?

Why are margins defined by sediment thickness?

I am unsatisfied with the solution that the Pacific is influenced by benthic sources and the Atlantic is not. This seems like a case of too many degrees of freedom to me. Adding a dial to the model (benthic processes) simply to fix problems in the Pacific is not a real solution. If the Pacific is effected by benthic processes, so too should the Atlantic. The "ring of fire" explanation seems too thin; large areas of the Pacific are floored in carbonates and clays (the latter with known non-radiogenic signals). Likewise, there are significant sources of radiogenic Nd to the Atlantic (Iceland!). Is it possible, for instance, that the preformed end-members in the Atlantic (the Pacific has only one - Circum Polar Water) are not characterized properly? Can benthic fluxes similar to the Pacific be imposed in the Atlantic, and then testing of boundary conditions of the Atlantic be done to find a reasonable solution? Or is this not possible? I am concerned that the authors found a model solution in the Atlantic that fits, and thus stopped looking at other possibilities.

Finally, this ms. should be evaluated by an expert in modeling. I cannot and have not evaluated the mechanics of the model presented.
Citation: https://doi.org/10.5194/egusphere-2022-937-RC3

Interactive discussion

Status: closed

RC1: 'Comment on egusphere-2022-937', Anonymous Referee #1, 21 Dec 2022

Review Robinson et al, Optimisation of the marine Nd isotope scheme in the ocean component of the FAMOUS general circulation model This paper present an optimization procedure of the Neodymium scheme implemented in the FAMOUS ocean general circulation model. It provides very interesting insights on Nd cycle and sensitivity to scavenging and sedimentary and river sources. The approach is performed with a statistical emulator that is convenient for the scientific goal attempted, and derives analysis that are in accordance with the procedure. However the approach has two drawbacks which limit the scope of the article:First it does not include the Aeolian dust input in the optimization procedure. This source is also subject to large uncertainties in its parameterization, in particular in the value of the solubility in dust where the values âârange between 2 and 50%. This source has an impact on a much wider spatial extension than the rivers. It also substantially modifies the surface concentration and isotopic Nd values ââwhich have an important weight in the optimization procedure. This arbitrary choice in the method is prejudicial and not justified in the document, and potentially tends to overestimate the role of rivers (Lines 477- 490: improvement in surface Nd concentration and isotopic composition is obtained with enhanced river inputs. However, dust deposition, as it strongly impacts surface modelling results, could moderate this conclusion if it was also included in the optimization procedure)

Secondly North Atlantic simulated isotopic composition is too radiogenic compared to the observation (Lines 545-550 ; 565-568; 690-707; 840-845). Radiogenic values simulated in the North Atlantic is more probably a consequence of a too strong exchange around Island which is highly radiogenic and influences (too much) the whole north Atlantic ocean Nd isotopic composition distribution.

Sensitivity tests in the north Atlantic are a good idea (section 4), but the strategy is not satisfying. It has an interest to test the flux around the Labrador Sea, but it generates little sensitivity. Tests around Island should have generated a higher sensitivity; the signal is highly radiogenic and advected through the subpolar circulation. This strategy lead to erroneous conclusions upon the sensitivity of north Atlantic water to the local sources, that are repeated all along the paper.

For these reasons, I recommend a major revision of this paper before publication.

Minor point:

Comparison of modelled inventory with inventory derived from observation (line 499): What is the uncertainty on the inventory derived from the observations, in order to justify that modeled inventory are underestimating or overestimating it (especially REF-CONC2)

Citation: https://doi.org/10.5194/egusphere-2022-937-RC1
RC2:
'Comment on egusphere-2022-937', Jianghui Du, 17 Feb 2023
Review of “Optimisation of the marine Nd isotope scheme in the ocean component of the FAMOUS general circulation model” by Robinson et al submitted to Biogeoscience.
In this manuscript, the authors presented an optimized version of the marine Nd cycle implemented in the FAMOUS model. This work builds on the companion study that has performed parameter tuning to some degree but uses a statistical approach and ensemble simulation for more robust optimization, while further testing the model sensitivity to sedimentary flux and its eNd.
I think the authors did a good job toward a more process-oriented model sensitivity study of the marine Nd cycle, which is a step forward. The lessons we learn from such sensitivity results are more valuable than simply arriving at an optimized solution, and can guide further process-based studies. However, I think the study still missed the great opportunity to truly explore the parameter space as constrained by observations of sedimentary flux and particle scavenging. It is not clear what is the rationale for the choices of parameter space, given that they are not compatible with existing measurements of these parameters. Also, Why is a global model considered optimized, when it only tries to optimize the Atlantic, at a cost of the Pacific, which has an Nd reservoir that is 3 times that of the Atlantic? This Pacific problem in GCM models of Nd has existed for 15 years since Jones et al (2008), despite that processes-based studies and sensitivity tests point to a clear direction of a solution. The Pacific problem is not a Pacific-only problem but as the model tests have shown, a global problem that also impacts the Atlantic through circulation. I think it is time to resolve this problem and hope my suggestions will help the authors achieve this, which would be a major step in modeling the marine Nd cycle.
Major points:
Choices of parameter values

Nd inventory
This study chooses a global Nd inventory of 4.3×10¹² g from Tachikawa et al (2003) as a tuning target. Tachikawa et al (2003) did an excellent job of compiling data and estimating the ocean Nd budget using the data available to them, but it’s safe to say that most of the seawater data available today are published after 2003. So the Nd inventory needs to be updated. Du et al. (2020 QSR) updated the global budget using the data up to 2019, and the resulting global ocean Nd inventory is 5.6×10¹² g based on volumetrically binning the data, much bigger than Tachikawa et al (2003). This number needs update again in 2023 given quite a few GEOTRACE transects have been published since 2019. However, in the optimization ensemble, the target Nd inventory is a maximum 5×10¹² g. This raises the question of how reasonable the optimized results are when using the Tachikawa et al (2003) inventory as the tuning target. Is it possible that the Nd sources in the current optimized model are underestimated because of low target inventory?
[Nd_p]/[Nd_d]
The parameter range for [Nd_p]/[Nd_d] is 0.001-0.006 in this study. The authors stated in the companion paper that this range “considers the few direct observations of [Nd]_p/[Nd]_d (Jeandel et al., 1995; Stichel et al., 2020; Zhang et al., 2008)”. But this is not true. Measurements of [Nd_p]/[Nd_d] have existed since the classic studies of (Sholkovitz et al., 1994; Bertram and Elderfield, 1993), and the number of data has grown considerably in recent years thanks to GEOTRACES, for example, GP15 reported full transect of particulate and dissolved Nd (Haley et al., 2021; Lam et al., 2018). Old or new these studies consistently show that [Nd_p]/[Nd_d] is on the order of 0.01, much higher than what’s used in this study. In Stichel et al., 2020, which is cited in the companion paper, [Nd_p]/[Nd_d] in the North Atlantic has a mean of 0.03 and a median of 0.02. Note that here I use the non-lithogenic [Nd_p] in this calculation, so only the exchangeable Nd_p is considered.
I noticed the reply of the authors to the same comment raised by Stichel on the companion paper (Robinson et al, 2022). Particle studies normally report both total Nd_p and non-lithogenic exchangeable Nd_p, based on leaching the labile fraction or removing the lithogenic fraction using detrital correction. Yes, it is the [Nd_p]_exchangible/[Nd_d] not [Nd_p]_total/[Nd_d] that should be used in the model. And the values I refer to here are all [Nd_p]_exchangible/[Nd_d] and they are ~0.01 according to these studies.
Sedimentary flux
The pore water data-constrained diffusive sedimentary Nd flux is 11~16×10⁹ g/yr (Du et al., 2020; Abbott et al., 2015). This does not include advective pore water fluxes such as bio-irrigation, which can be comparable to or larger than the diffusive flux (Du et al., 2022; Deng et al., 2022). In this study, the tested range of sedimentary flux is only 1.5~6×10⁹ g/yr. Admitting the possible large uncertainty of the sedimentary flux, the parameter range should at least bracket this data-constrained value.
Residence time and regional sedimentary influence

In the optimized model both the Atlantic and Pacific seawater eNd endmembers are much less extreme than measured, resulting in a global distribution of eNd being too uniform than observed. Thus, ocean circulation plays a more important role in the model than in reality, which causes the low sensitivity of the modeled seawater eNd to localized sedimentary eNd change in the sensitivity tests.
I think this is because the parameter ranges chosen by the study are too restricted and do not bracket the natural ranges observed. Thus the residence time of Nd is too long in the model, such that the regional sedimentary influence cannot manifest clearly. I suggest that a combination of higher [Nd_p]/[Nd_d], higher sedimentary flux, and shorter residence time are needed to capture the heterogeneity of eNd distribution in the ocean. Here are my reasons.
First, although box models cannot be used to study the spatial distribution of eNd in the ocean, they are better at estimating global properties, such as residence time. This is because box models strictly follow the data-constrained requirement of global mass balance, whereas the GCMs do not, as shown by the failure of GCMs to correctly model the Pacific eNd. To me, the box model-constrained Nd residence time of 400~500 years is better than the GCM-constrained residence time of ~700 years. This shorter residence time is supported by multiple lines of evidence. Inverse model-based sensitivity constraint shows optimal results at Nd residence time of 400~500 years (Siddall et al., 2008; Pasquier et al., 2022). Th-constrained particle cycling model shows Nd residence time of ~100 years in the Atlantic (Hayes et al., 2018). Moreover, in FAMOUS the residence time decreases with an increase [Nd_p]/[Nd_d], if using the realistic value of 0.01~0.02, FAMOUS will likely produce a shorter residence time that is close to 400 years.
Abyssal sedimentary flux

The authors argue that “perhaps too much emphasis has recently been placed upon exclusively resolving nonconservative interactions of an abyssal deep seafloor flux to solve the Nd paradox”. I suggest the opposite is true.
Ever since the early GCM models (Jones et al., 2008), GCMs have struggled to correctly model the abyssal Pacific eNd, even though a solution to this problem was presented by studies of sediment and pore water Nd. Why not increase abyssal sedimentary Nd flux to what’s estimated by these studies, which will lead to better Pacific model results?
Say that the authors correctly model the Pacific eNd, then their Atlantic results will worsen, as the radiogenic Pacific water will make the AABW endmember more radiogenic, thus leading to more radiogenic NADW, which is already too radiogenic in the model compared to the data. Doesn’t this imply that the abyssal Atlantic sedimentary flux is also underestimated? Why not test the scenario of combined high abyssal sedimentary flux with extreme eN as suggested by observations? For example, the results from GEOTRACES GP03 show that along the Deep Western Boundary Current in the North Atlantic, there’s an input of highly unradiogenic sedimentary flux related to the abyssal Nepheloid layers (Jaume-Seguí et al., 2021).
Minor points:
L52. Authigenic phosphate is also a major sink.
L232. This is not an equation/function.
L315. But isn’t the abyssal Pacific the largest reservoir of Nd?
L390. In reality, fsed and the scavenged flux are not independent. Part of fsed results from the regeneration of the scavenged flux that enters the sediments; part of fsed comes from new sources (e.g. reactive detritus) within the sediments (Du et al., 2022). So sedimentary flux is closely tied to the residence time. It is only because in the model you have no interactive sedimentary component that you have to specify fsed as an independent parameter. I suggest the difference between model and reality should be made clear.
L517. (Sholkovitz et al., 1994) showed that [Nd_p]/[Nd_d] was lower at the surface than at depth in the Atlantic. This model misfit is probably caused by using a vertically constant [Nd_p]/[Nd_d].
L564. The GEOTRACES GA02 full Atlantic meridional transect of eNd was published last year (Wu et al., 2022). This should be plotted along Fig 6. It is worth noting here that the Southern ocean endmember is too unradiogenic (-11) in the model compared to data (-9).
L572. I would rather argue that AABW is too unradiogenic because the Pacific endmember is too unradiogenic in the model. With adequate Pacific endmember, the model may get the correct eNd for AABW even if the benthic flux in the Atlantic is unradiogenic.
L693-704. It is also likely that the local Nd residence time is too long for the circulation signal to overcome the sedimentary influence in the model. See major points.
L771-780. It is important to evaluate more in detail two scenarios of why NADW eNd was fitted better in this margin-only scenario: (1) PDW/AABW has become much more unradiogenic because abyssal sedimentary flux in the Pacific was reduced, propagating this signal to the abyssal Atlantic, (2) abyssal sedimentary flux in the Atlantic is not adequate. The implications are very different regarding the importance of the abyssal sedimentary source. Ideally, sensitivity tests should include ones that limit sedimentary flux to the margins only in the Pacific or only in the Atlantic.
L799. That’s very well said.
Abbott, A. N., Haley, B. A., McManus, J., and Reimers, C. E.: The sedimentary flux of dissolved rare earth elements to the ocean, Geochim. Cosmochim. Acta, 154, 186–200, https://doi.org/10.1016/j.gca.2015.01.010, 2015.
Bertram, C. J. and Elderfield, H.: The geochemical balance of the rare earth elements and neodymium isotopes in the oceans, Geochim. Cosmochim. Acta, 57, 1957–1986, https://doi.org/10.1016/0016-7037(93)90087-D, 1993.
Deng, K., Yang, S., Du, J., Lian, E., and Vance, D.: Dominance of benthic flux of REEs on continental shelves: implications for oceanic budgets, Geochem. Perspect. Lett., 22, 26–30, https://doi.org/10.7185/geochemlet.2223, 2022.
Du, J., Haley, B. A., and Mix, A. C.: Evolution of the Global Overturning Circulation since the Last Glacial Maximum based on marine authigenic neodymium isotopes, Quat. Sci. Rev., 241, 106396, https://doi.org/10.1016/j.quascirev.2020.106396, 2020.
Du, J., Haley, B. A., Mix, A. C., Abbott, A. N., McManus, J., and Vance, D.: Reactive-transport modeling of neodymium and its radiogenic isotope in deep-sea sediments: The roles of authigenesis, marine silicate weathering and reverse weathering, Earth Planet. Sci. Lett., 596, 117792, https://doi.org/10.1016/j.epsl.2022.117792, 2022.
Haley, B. A., Wu, Y., Muratli, J. M., Basak, C., Pena, L. D., and Goldstein, S. L.: Rare earth element and neodymium isotopes of the eastern US GEOTRACES Equatorial Pacific Zonal Transect (GP16), Earth Planet. Sci. Lett., 576, 117233, https://doi.org/10.1016/j.epsl.2021.117233, 2021.
Hayes, C. T., Anderson, R. F., Cheng, H., Conway, T. M., Edwards, R. L., Fleisher, M. Q., Ho, P., Huang, K.-F., John, S. G., Landing, W. M., Little, S. H., Lu, Y., Morton, P. L., Moran, S. B., Robinson, L. F., Shelley, R. U., Shiller, A. M., and Zheng, X.-Y.: Replacement Times of a Spectrum of Elements in the North Atlantic Based on Thorium Supply, Glob. Biogeochem. Cycles, 32, 1294–1311, https://doi.org/10.1029/2017GB005839, 2018.
Jaume-Seguí, M., Kim, J., Pena, L. D., Goldstein, S. L., Knudson, K. P., Yehudai, M., Hartman, A. E., Bolge, L., and Ferretti, P.: Distinguishing Glacial AMOC and Interglacial Non-AMOC Nd Isotopic Signals in the Deep Western Atlantic Over the Last 1 Myr, Paleoceanogr. Paleoclimatology, 36, e2020PA003877, https://doi.org/10.1029/2020PA003877, 2021.
Jones, K. M., Khatiwala, S. P., Goldstein, S. L., Hemming, S. R., and van de Flierdt, T.: Modeling the distribution of Nd isotopes in the oceans using an ocean general circulation model, Earth Planet. Sci. Lett., 272, 610–619, https://doi.org/10.1016/j.epsl.2008.05.027, 2008.
Lacan, F. and Jeandel, C.: Acquisition of the neodymium isotopic composition of the North Atlantic Deep Water, Geochem. Geophys. Geosystems, 6, Q12008, https://doi.org/10.1029/2005GC000956, 2005.
Lam, P. J., Lee, J.-M., Heller, M. I., Mehic, S., Xiang, Y., and Bates, N. R.: Size-fractionated distributions of suspended particle concentration and major phase composition from the U.S. GEOTRACES Eastern Pacific Zonal Transect (GP16), Mar. Chem., 201, 90–107, https://doi.org/10.1016/j.marchem.2017.08.013, 2018.
Pasquier, B., Hines, S. K. V., Liang, H., Wu, Y., Goldstein, S. L., and John, S. G.: GNOM v1.0: an optimized steady-state model of the modern marine neodymium cycle, Geosci. Model Dev., 15, 4625–4656, https://doi.org/10.5194/gmd-15-4625-2022, 2022.
Sholkovitz, E. R., Landing, W. M., and Lewis, B. L.: Ocean particle chemistry: The fractionation of rare earth elements between suspended particles and seawater, Geochim. Cosmochim. Acta, 58, 1567–1579, https://doi.org/10.1016/0016-7037(94)90559-2, 1994.
Siddall, M., Khatiwala, S., van de Flierdt, T., Jones, K., Goldstein, S. L., Hemming, S., and Anderson, R. F.: Towards explaining the Nd paradox using reversible scavenging in an ocean general circulation model, Earth Planet. Sci. Lett., 274, 448–461, https://doi.org/10.1016/j.epsl.2008.07.044, 2008.
Wu, Y., Pena, L. D., Anderson, R. F., Hartman, A. E., Bolge, L. L., Basak, C., Kim, J., Rijkenberg, M. J. A., de Baar, H. J. W., and Goldstein, S. L.: Assessing neodymium isotopes as an ocean circulation tracer in the Southwest Atlantic, Earth Planet. Sci. Lett., 599, 117846, https://doi.org/10.1016/j.epsl.2022.117846, 2022.
Citation: https://doi.org/10.5194/egusphere-2022-937-RC2
RC3:
'Comment on egusphere-2022-937', Anonymous Referee #3, 17 Feb 2023
Review of "Optimization of the marine Nd isotope scheme in the ocean component of the FAMOUS general circulation model" by Robinson et al., submitted to EGUsphere, Feb. 2023.
This work presents the incorporation of Nd isotopes into the FAMOUS GCM, including optimization and validation. They authors evaluate the model and do some sensitivity tests, particularly on the dichotomy whether a benthic Nd source flux is ubiquitous or confined to the margins. The authors conclude that the model is largely robust, but that outstanding questions regarding a few key boundary conditions and processes remain.
Overall, this is a well-written ms. I think some further editing (culling) would make it stronger, but it was not challenging to follow the discourse. The figures were well made, but I had a difficult time with them: the 3D plot (Fig.1) is without droplines and is very difficult to evaluate. Similar, but less so, for the heat data on the 2D plots (Figs.3 & 4). To the point where these figures do not add anything to help understand the text. And I have a personal dislike of model section/point data figures (Figs. 6, 8, 9, 10). I just can’t properly evaluate the data:model fit from these figures. Why not show the data as a scatterplot?
Overall, I thought the ms. was well-done and offers a contribution to the field, largely in terms of description/validation of a model (for future use) and directions suggested for future research. I recommend publication, perhaps with consideration of the following questions/comments:
L380 (and discussion): Does the evaluation of Nd(I) need to be done on a full-model, or could it be done in the simple box model the authors use to steer their GCM runs?

Why are margins defined by sediment thickness?

I am unsatisfied with the solution that the Pacific is influenced by benthic sources and the Atlantic is not. This seems like a case of too many degrees of freedom to me. Adding a dial to the model (benthic processes) simply to fix problems in the Pacific is not a real solution. If the Pacific is effected by benthic processes, so too should the Atlantic. The "ring of fire" explanation seems too thin; large areas of the Pacific are floored in carbonates and clays (the latter with known non-radiogenic signals). Likewise, there are significant sources of radiogenic Nd to the Atlantic (Iceland!). Is it possible, for instance, that the preformed end-members in the Atlantic (the Pacific has only one - Circum Polar Water) are not characterized properly? Can benthic fluxes similar to the Pacific be imposed in the Atlantic, and then testing of boundary conditions of the Atlantic be done to find a reasonable solution? Or is this not possible? I am concerned that the authors found a model solution in the Atlantic that fits, and thus stopped looking at other possibilities.

Finally, this ms. should be evaluated by an expert in modeling. I cannot and have not evaluated the mechanics of the model presented.
Citation: https://doi.org/10.5194/egusphere-2022-937-RC3

Suzanne Robinson, Ruza Ivanovic, Lauren Gregoire, Lachlan Astfalck, Tina van de Flierdt, Yves Plancherel, Frerk Pöppelmeier, and Kazuyo Tachikawa

Supplement

https://doi.org/10.5194/egusphere-2022-937-supplement

Suzanne Robinson, Ruza Ivanovic, Lauren Gregoire, Lachlan Astfalck, Tina van de Flierdt, Yves Plancherel, Frerk Pöppelmeier, and Kazuyo Tachikawa

Viewed

Total article views: 819 (including HTML, PDF, and XML)

HTML	PDF	XML	Total	Supplement	BibTeX	EndNote
552	223	44	819	77	34	44

HTML: 552
PDF: 223
XML: 44
Total: 819
Supplement: 77
BibTeX: 34
EndNote: 44

Views and downloads (calculated since 07 Oct 2022)

Month	HTML	PDF	XML	Total
Oct 2022	135	35	5	175
Nov 2022	25	13	0	38
Dec 2022	31	10	3	44
Jan 2023	36	20	0	56
Feb 2023	81	17	4	102
Mar 2023	33	10	0	43
Apr 2023	19	10	0	29
May 2023	14	2	0	16
Jun 2023	8	5	1	14
Jul 2023	12	9	1	22
Aug 2023	8	5	0	13
Sep 2023	14	10	0	24
Oct 2023	5	9	3	17
Nov 2023	5	0	5
Dec 2023	13	9	4	26
Jan 2024	11	1	12
Feb 2024	4	7	0	11
Mar 2024	11	9	1	21
Apr 2024	11	5	3	19
May 2024	10	9	1	20
Jun 2024	12	6	4	22
Jul 2024	7	8	2	17
Aug 2024	13	2	4	19
Sep 2024	9	4	1	14
Oct 2024	10	4	2	16
Nov 2024	14	5	4	23
Dec 2024	1	0	1

Cumulative views and downloads (calculated since 07 Oct 2022)

Month	HTML	PDF	XML	Total
Oct 2022	135	35	5	175
Nov 2022	25	13	0	38
Dec 2022	31	10	3	44
Jan 2023	36	20	0	56
Feb 2023	81	17	4	102
Mar 2023	33	10	0	43
Apr 2023	19	10	0	29
May 2023	14	2	0	16
Jun 2023	8	5	1	14
Jul 2023	12	9	1	22
Aug 2023	8	5	0	13
Sep 2023	14	10	0	24
Oct 2023	5	9	3	17
Nov 2023	5	0	5
Dec 2023	13	9	4	26
Jan 2024	11	1	12
Feb 2024	4	7	0	11
Mar 2024	11	9	1	21
Apr 2024	11	5	3	19
May 2024	10	9	1	20
Jun 2024	12	6	4	22
Jul 2024	7	8	2	17
Aug 2024	13	2	4	19
Sep 2024	9	4	1	14
Oct 2024	10	4	2	16
Nov 2024	14	5	4	23
Dec 2024	1	0	1

Viewed (geographical distribution)

Total article views: 811 (including HTML, PDF, and XML) Thereof 811 with geography defined and 0 with unknown origin.

Country	#	Views	%

Latest update: 13 Dec 2024

Short summary

The neodymium (Nd) isotope (ε_Nd) scheme in the ocean model of FAMOUS is used to explore a benthic Nd flux to seawater. Our results demonstrate that sluggish modern Pacific waters are sensitive to benthic flux alterations, whereas the well-ventilated North Atlantic displays a much weaker response. In closing, there are distinct regional differences in how seawater acquires its ε_Nd signal, in part relating to the complex interactions of Nd addition and water advection.


Total:	0
HTML:	0
PDF:	0
XML:	0