the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 15 May 2025
Modeling Indian Ocean circulation to study marine debris dispersion: insights into high-resolution and Stokes drift effects with Symphonie 3.6.6
Abstract. The Indian Ocean basin faces significant anthropogenic pressure due to its connection to over 2.2 billion people through river basins. Indian Ocean dynamics are characterized by strong regional and seasonal variability driven by the monsoon system and intense eddy activity. To address the issue of land-sea transfers and marine debris dispersion in this complex ocean, we developed a new circulation modeling configuration using the hydrodynamic model SYMPHONIE. Our configuration introduces a unique telescopic grid covering the entire basin, enabling the study of sub-basin connectivity while resolving meso and submesoscale processes in the coastal region, from the Mozambique Channel to the Bay of Bengal, at a resolution of 1 to 3 km. Additionally, we integrate the recently released high-resolution GloFAS river discharge dataset to force the physical simulations with daily freshwater inputs. Three annual experiments are conducted, alternatively considering Stokes drift forcing and different grid resolutions. Comparisons of temperature, salinity and sea level with in situ and satellite data show the good performance of the simulations and the ability of the high resolution model to accurately capture the spatial and temporal variability of surface dynamics and water masses over the Indian Ocean. We further analyze energy budgets and perform Lagrangian experiments to illustrate the critical role of resolution and Stokes drift in shaping the circulation and the resulting marine debris dispersion patterns. The effect of energy levels is particularly significant on trajectory statistics such as average travel distances and preferred spread direction. Notably, Stokes drift has a significant seasonal effect in the Arabian Sea during the southwest monsoon, while current field resolution strongly influences trajectories in the Mozambique Channel. Our results provide a robust modeling framework for studying Indian Ocean dynamics and exploring their effect on marine connectivity and the transport of matter, including pollutants, larvae or organic matter.
- Preprint
(40995 KB) - Metadata XML
- BibTeX
- EndNote
Status: final response (author comments only)
-
RC1: 'Comment on egusphere-2025-1918', Anonymous Referee #1, 24 Jun 2025
Comments on Weiss et al.: “Modeling Indian Ocean circulation to study marine debris dispersion: insights into high-resolution and Stokes drift effects with Symphonie 3.6.6”
The manuscript addresses an outstanding issue in the field of marine (pollution) transport modelling: providing coherent ocean velocity output from the coast to the open ocean that resolves the dominant transport processes from the submesoscale to the basin scale. It does so by introducing a new ocean model configuration for the Indian Ocean with grid refinement towards key coastal regions, inclusion of wave forcing, and a more realistic representation of river discharge. While the grid refinement and inclusion of wave forcing are not based on novel concepts, their combination in this context yields a potentially valuable new tool for marine pollution modelling that may support both sensitivity studies and improved estimates of pollution patterns compared to standard approaches.
The overall structure and presentation of the results is clear, including useful visualizations. However, the writing includes imprecise terminology, and several methodological concepts are insufficiently or inaccurately described. This limits a thorough assessment of the approach and specifically concerns the description and implementation of wave effects (see General Comment 1) and model-observation comparisons (General Comment 2). Additionally, the stated goal of addressing land-sea transfers is not clearly supported by the Lagrangian analyses presented (General Comment 3).
To conclude, while I see the potential of the manuscript to become a relevant addition to the field, I recommend major revisions to clarify key concepts, improve the terminology, and better align overall content and objectives.
General Comments:
1. Representation of wave effects
The description of wave-related processes is vague, and terminology is inconsistently used. In theory, waves affect Lagrangian transport both (i) directly via Stokes drift and (ii) indirectly through wave-induced modifications of Eulerian currents, including (anti-)Stokes forces such as the Stokes–Coriolis force. After rereading the methods, I was left with the impression that:
- The only wave-related effect considered is the Stokes–Coriolis force, yet, this is also loosely referred to throughout as "Stokes drift forcing" or simply "Stokes drift".
- Lagrangian particles are advected by Eulerian currents only (eventually including the effect of the Stokes-Coriolis force), and not by a combination of Eulerian currents and Stokes drift.
If this is correct, the implementation is basic and not fully aligned with the current state of the art (e.g., Couvelard et al., 2020). It also contrasts with recent findings suggesting that both Stokes drift and wave-induced modifications of Eulerian currents are important for Lagrangian transport (e.g., Röhrs et al., 2022; Cunningham et al., 2022; Rühs et al., 2025). I recommend that the authors:
- Clearly distinguish between Stokes drift and Stokes–Coriolis forcing throughout the text.
- Clarify whether and how Stokes drift is included in the Lagrangian advection scheme.
- Include relevant formulas and refer to established methods, or detail any deviations from them.
2. Model-observation comparisons
The approach to model validation needs clarification:
- Are all datasets interpolated to a common grid prior to computing correlations?
- How might the applied nudging influence the agreement with observations?
Further detail on the nudging (location, depth, timescale) is needed as well, see also specific comment below.
3. Land-sea transfers
The abstract sets the goal to address land–sea transfers, but no direct analysis of this is presented. Lagrangian experiments are based on offshore releases, and sensitivity tests focus on grid resolution and wave forcing. The influence of more realistic river discharge, while implemented, is not tested. This feels like a missed opportunity. For example, exploring how coastal retention changes with the new configuration could strengthen the manuscript’s relevance considerably. If the authors choose not to pursue additional analyses, I suggest reformulating the manuscript’s goals to avoid overstating its scope.
Specific comments:
- L 9-10: “Three annual experiments are conducted, alternatively considering Stokes drift forcing and different grid resolutions”. This sentence is ambiguous about the exact combinations of Stokes drift forcing and grid resolution used in the three experiments. Please clarify.
- L 39: “Lagrangian and Eulerian dispersion modeling studies can help to fill this knowledge gap”. Why are Eulerian dispersion modeling studies mentioned here? Consider omitting “Eulerian” dispersion modelling here, as all examples relate to Lagrangian methods.
- L 57 ff: “Stokes forcing related to waves can also have a significant impact on the dispersion of floating material […], including indirectly the effect of windage on surface particle drift”. This sentence is inaccurate. In principal, Stokes drift, Stokes forces, and windage are individual processes affecting marine matter transport. If Stokes drift is parameterized via an additional transport component in wind direction, the corresponding tuning factor can also be chosen to include windage. But if Stokes drift obtained from a wave model is included in the transport simulations, that does not necessarily include any windage component. Please correct.
- L 65 ff: “One of the main challenges in modeling the dispersion of marine debris is the necessity to study the continuity of dispersion patterns from […] coastal scales […] to large scale ocean currents […]”. I completely agree! However, as summarized above, I think the current set of analyses unfortunately does not convincingly demonstrate how the new model configuration tackles this.
- L 70: “stranding issues” It is not clear what is meant by issues here, please clarify.
- L 81: “floating particle” This terminology is recurrently used within the manuscript and implies particles remain at the surface. However, the method part introduces the trajectories as 3D, including vertical displacements (cf. l. 155 ff.). “Buoyant particles” may be more appropriate to use.
- L 103: “Bathymetry is built from […] with manual verification and modification of the coastline […]. Without further explanations, these kind of manual modifications compromise reproducibility. Publishing the modified bathymetry along with a documentation of the performed changes is strongly recommended.
- L 116 ff: “A temporal nudging layer is configured wherever the resolution of the telescopic grid is lower than the 1/12° GLORYS forcing (to the southeast).” This sentence and the corresponding paragraph remain unclear about whether nudging is only applied at the lateral open boundaries or also within the domain. This should be stated explicitely, as it also impacts the interpretability of the validation of modeled SST and SSS (which naturally would be expected to be very good for regions where nudging is applied).
- Figure 1:“blue boxes in the best-resolved regions […] are used to calculate spatial averages […] and Lagrangian trajectories presented in the following”. Not clear whether this refers to the big or small blue boxes (there are two frames in each mentioned region); “presented in the following” reads weird in a figure caption (as nothing follows directly)
- L 136: “This parameterization uses the surface Stokes drift (extrapolated below the surface using the peak period)” a formula and reference would be helpful.
- L 140 ff: “In practice, Stokes drift is considered through the transport calculation in the model’s Eulerian equations. In parallel, the momentum equations take into account the anti-Stokes term, […].” What Eulerian transport equations exactly, are you referring to tracer transport here? Adding the respective equations would be helpful.
- L 157: What type of interpolation scheme is used?
- L 164 ff: Please specify the velocity fields that are used for the advection, at best via formula. Specifically, for IndOc.HR-Sto: Did you use only the modelled Eulerian current fields (that include the effect of the Stokes-Coriolis force), or did you use these Eulerian current fields plus Stokes drift?
- L 344 ff: “SLA correlations are slightly lower than those for SST and SSS, due to the complexity of mesoscale dynamics and intrinsic variability.“ Please explain why SLA is stronger impacted by intrinsic variability than SST and SSS. Could this to some degree also be related to the nudging?
- L 358: “This might be attributed to wave-induced surface momentum fluxes” This sentence surprised me, as no information about wave-induced surface momentum fluxes is given in the model description. Please clarify how these type of wave effects were included in the model (see general comment above).
- L 363 ff: “For SSS, IndOc.12 clearly outperforms the HR configurations in terms of NRMSE and bias, […] This counterintuitive result may be related to the coarser resolution [of the observational reference]” Please rephrase the “outperforms” statement. If the observational product is too coarse, then better agreement may not indicate better performance.
- L 396: Consider replacing the word “performance” for reason listed in previous comment
- Section 3.3 “Description of the regional surface circulation” The title is misleading, as mostly not circulation itself but impact on SST and SSS is discussed. Please also clarify the relevance of this section. Is that still part of the model validation?
- L 439 ff: I assume the Eulerian mean kinetic energy is calculated for comparison between the different model runs. However, as the Lagrangian mean kinetic energy would potentially be even more relevant for transport studies, this should be specified at the beginning of this section
- Section 5: The “mean spread distance” is introduced and analyzed. Yet, this seems to refer to the commonly used “displacement”. If so, please adopt standard terminology.
- L 541 ff: “Similarly, 122 to 143 km separate the median trajectories between IndOc.12 and both HR simulations.” What is the median trajectory? The median distance travelled by the trajectories differs by 122 to 134 km? Or the final position of the centroids (average position of all particles) lay 122 to 134 apart?
- L 600 ff: “The effect of the Stokes Drift is less significant either on the statistics of the distances traveled (Fig. 12) or on the direction of the trajectories (Fig. A11).” This sentence is not clear. What is the less referring to?
- Figure 13: Nice visualization!
- Appendix A: please specify all variables and parameters used in the equation. Specifically, define n, the velocity components, and averaging periods for mean velocities.
Technical corrections:
- L 32: “submesoscale are ephermal (…)” -> submesoscale features are ephermal
- L 36: “modeled hypotheses” -> model-based hypotheses
- L 59: configurations -> ocean model configurations for the Indian ocean
- L 63: “none of them […] and use ” -> “they do not […] and do not use”
- L 98: The mentioning of the number of cells after the clause about limiting open boundaries is ambiguous. If the total grid size is meant, consider rewriting, e.g., "The grid is a curvilinear Arakawa C-grid (3200 x 2800 cells) covering the entire Indian Ocean […], thereby limiting the number of open boundaries."
- L 676: There appears to be some redundant text after the acknowledgment section, please remove.
References:
- Couvelard, X., Lemarié, F., Samson, G., Redelsperger, J.-L., Ardhuin, F., Benshila, R., and Madec, G.: Development of a two-way-coupled ocean–wave model: assessment on a global NEMO(v3.6)–WW3(v6.02) coupled configuration, Geosci. Model Dev., 13, 3067–3090, https://doi.org/10.5194/gmd-13-3067-2020, 2020.
- Cunningham, H. J., Higgins, C., & van den Bremer, T. S. (2022). The role of the unsteady surface wave-driven Ekman-Stokes flow in the accumulation of floating marine litter. Journal of Geophysical Research: Oceans, 127, e2021JC018106. https://doi.org/10.1029/2021JC018106
- Röhrs, J., Christensen, K.H., Hole, L.R. et al. Observation-based evaluation of surface wave effects on currents and trajectory forecasts. Ocean Dynamics 62, 1519-1533 (2012). https://doi.org/10.1007/s10236-012-0576-y
- Rühs, S., van den Bremer, T., Clementi, E., Denes, M. C., Moulin, A., and van Sebille, E.: Non-negligible impact of Stokes drift and wave-driven Eulerian currents on simulated surface particle dispersal in the Mediterranean Sea, Ocean Sci., 21, 217-240, https://doi.org/10.5194/os-21-217-2025, 2025.
Citation: https://doi.org/10.5194/egusphere-2025-1918-RC1 -
AC1: 'Reply on RC1', Lisa Weiss, 02 Sep 2025
We sincerely thank the editor and both reviewers for their time and careful consideration of our manuscript. We have significantly improved the overall quality of the manuscript by responding to their helpful and constructive comments. In the following, we address the comments raised and propose some modifications, including mainly :
- the description and implementation of wave effects,
- the model-observation comparisons,
- addressing land-sea transfers not supported by the Lagrangian analyses of the study.
Our response to each point made by the reviewer 1 is presented attached in the pdf file (in blue). Our corrections and additions to the manuscript text are underlined here in the responses.
Best regards,
Lisa Weiss and co-authors
-
RC2: 'Comment on egusphere-2025-1918', Anonymous Referee #2, 11 Jul 2025
Comments on Weiss et al.: “Modeling Indian Ocean circulation to study marine debris dispersion: insights into high-resolution and Stokes drift effects with Symphonie 3.6.6”
To gain a better understanding of the pathways and accumulation of marine debris, significant developments have occurred over the past decade. The main three components which govern the quality of a dispersal model are the marine debris sources (initial conditions), the marine debris transport mechanisms (in other words the influence of the different forcing components namely circulation, wind and waves on the marine debris displacement) and the quality and reliability of the forcing models themselves. In that context, this paper aims to cover the latter two by studying the impact of wave-induced transport and higher resolution, which could presumably give a better description of the circulation, especially close to the coastline.
The paper is well structured and written with a good quality of English. The results are well presented with precise and good-quality figures. It, however, suffers from some unclarities in the modeling used to describe the wave-induced transport. As highlighted by the first reviewer, it is hard to understand whether we are looking at a one-way coupled circulation - wave model, or that the Stokes drift (which is a Lagrangian « thing ») is « simply » added to the circulation. These unclarities lead to troubling the appraisal of the interpretation made by the authors of such a phenomenon. Also, as stressed by the first reviewer, looking at the dispersal of marine debris only from a « synthetic » offshore release scenario feels a bit frustrating from a reviewer’s standpoint where all the effort put into having a more precise representation of river discharges and using higher resolution close to the coastline becomes, in turn, irrelevant. By solving those two points (clarifying the narrative and explanations around wave-induced transport and managing expectations on the marine debris dispersal relevance in the objectives) this work should become suitable for publication.
Comments (in addition to those given by Reviewer 1):
l.44: HYCOM has a meridional resolution of 0.08° and 0.04° at higher latitudes from the equator
l. 46: maybe worth mentionning the existence of a global LLC4320 (see Forget et al. 2015 and Rocha et al. 2016 e.g.) at a much higher resolution (1/48°) for 2011
l. 57-60: the explanation on the roles in the impact of wave-induced drift on Eulerian currents, wave-induced drift on Lagrangian particles, and possibly windage which could be added to the latter processes is pretty unclear. An interesting paper on the influence of the differents processes at a global scale (including a split between Ekman currents / geostrophy / tides etc…) is Onink et al. 2019.
l. 88: from the context it is clear that « tracers » correspond to temperature and salinity but why not explicitly mention them, in a dispersal / Lagrangian paper this can be confusing.
Figure 1: every 50 gridlines instead of meshes, maybe worth adding a red circle to materialize the release of Lagrangian particles
l. 161: rising velocity (singular) - because only one rising velocity is considered for all particles, the value of 1mm/s seems pretty low compared to experienced rising velocities for e.g. mesoscale plastics (see Lebreton et al. 2018, Supplementary Material)
l. 201: 47 rivers vs 46 rivers on line 193, am I missing something - see also Figure A4
paragraph starting l. 215: the discussion there « contradicts » the objective of having a better representation of river discharges given than GLOFAS seems to systematically overestimate the measurements.
Figure 7: the superimposition of daily tracer profiles is hard to read (dark blue and light blue)
Figure 8: it would be interesting to split between the different regions and not only focus on the whole domain - in relation with the analysis made after (figure 9) and before…
Figure 12: consider bigger fonts in the blue / red / yellow boxes
l. 622: «its implications for marine debris dispersion in the region », in light of what was said before, marine debris dispersion cannot be viewed independently from release locations. Demonstrating the impact of an improved circulation and transport modeling based on a (very) hypothetical release scenario (which cannot be validated by design) weakens the demonstration. Especially given that so much effort has been put into the river discharges (which could, combined with other socio-economic information, provide useful proxies for marine debris release, see e.g. Meijer et al. 2022)
l. 667: it is now understandable why the coastal releases were not considered so far as it is meant for future studies, so why not state clearly that the implications of such improved modeling on the marine debris dispersal in the region will only be tangible once realistic debris sources will be considered.
Figure A4: back to 46 rivers?
Figure A9/A10/A11: a slight increase of the font size in the roses could be beneficial
References:
Forget, G., Campin, J.-M., Heimbach, P., Hill, C. N., Ponte, R. M., and Wunsch, C.: ECCO version 4: an integrated framework for non-linear inverse modeling and global ocean state estimation, Geosci. Model Dev., 8, 3071–3104, https://doi.org/10.5194/gmd-8-3071-2015.
Rocha, C. B., Chereskin, T. K., Gille, S. T., and Menemenlis, D.: Mesoscale to Submesoscale Wavenumber Spectra in Drake Passage, J. Phys. Oceanogr., 46, 601–620, https://doi.org/10.1175/JPO-D-15-0087.1, 2016.
Onink, V., Wichmann, D., Delandmeter, P., van Sebille, E., 2019. The role of Ekman currents, geostrophy and stokes drift in the accumulation of floating microplastics. J. Geophys. Res. Oceans 124. https://doi.org/10.1029/2018JC014547.
Lebreton, L., et al. (2018). Evidence that the Great Pacific Garbage Patch is rapidly accumulating plastic. Sci. Rep., 8. https://doi.org/10.1038/s41598-018-22939-w
Citation: https://doi.org/10.5194/egusphere-2025-1918-RC2 -
AC2: 'Reply on RC2', Lisa Weiss, 02 Sep 2025
We sincerely thank the editor and both reviewers for their time and careful consideration of our manuscript. We have significantly improved the overall quality of the manuscript by responding to their helpful and constructive comments. In the following, we address the comments raised and propose some modifications, including mainly :
- the description and implementation of wave effects,
- the model-observation comparisons,
- addressing land-sea transfers not supported by the Lagrangian analyses of the study.
Our response to each point made by the reviewer 2 is presented attached in the pdf file (in blue). Our corrections and additions to the manuscript text are underlined here in the responses.
Best regards,
Lisa Weiss and co-authors
-
AC2: 'Reply on RC2', Lisa Weiss, 02 Sep 2025
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 505 | 93 | 13 | 611 | 18 | 28 |
- HTML: 505
- PDF: 93
- XML: 13
- Total: 611
- BibTeX: 18
- EndNote: 28
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1