Physical drivers and parameter sensitivities of  pearl river-derived sediment dispersal on the  Northern South China Sea Shelf: a modeling study

Zhang, Guang; Hu, Suan; Yu, Xiaolong; Zhang, Heng; Gong, Wenping

doi:https://doi.org/10.5194/os-21-2041-2025

Articles | Volume 21, issue 5

https://doi.org/10.5194/os-21-2041-2025

© Author(s) 2025. This work is distributed under
the Creative Commons Attribution 4.0 License.

https://doi.org/10.5194/os-21-2041-2025

© Author(s) 2025. This work is distributed under
the Creative Commons Attribution 4.0 License.

Articles | Volume 21, issue 5

Research article

|

18 Sep 2025

Research article |

| 18 Sep 2025

Physical drivers and parameter sensitivities of pearl river-derived sediment dispersal on the Northern South China Sea Shelf: a modeling study

Guang Zhang, Suan Hu, Xiaolong Yu, Heng Zhang, and Wenping Gong

Download

Final revised paper (published on 18 Sep 2025)
Supplement to the final revised paper
Preprint (discussion started on 04 Nov 2024)
Supplement to the preprint

Interactive discussion

Status: closed

RC1:
'Comment on egusphere-2024-3289', Anonymous Referee #1, 01 Dec 2024

This study investigated the wind and wave effects on the dispersal of the Pearl River-derived sediment using a coupled numerical model. The results demonstrated the crucial role of wind in the westward along-shelf transport of the riverine sediments and the formation of the mud belt along the coasts. The diagnostics of cross-shelf current and suspended sediment concentration revealed the mechanism of winds in the formation of along-shelf currents, vertical mixing, and resuspension of sediment. This work is valuable for understanding the transport processes of the Pearl River sediment and the manuscript is well written. Therefore, I recommend publication of the manuscript after minor revision.
Were winds and waves included om the ramp simulation from Jan.1, 2016 to Apr. 1, 2017? The initial bed sediment distribution would be different for the two case simulations. Which case Figure 2 represents for?
The critical shear stress chosen for clay is large. Although clay is cohesive, the large critical shear stress would restrict the transport of clay on the shelf and the riverine clay actually represents riverine sand. Comparing the initial prototype of bottom sediment with that after realistic reworking, the changes in slit component is the most significant because of the low critical shear stress.
Line 374-376. The westward extension of Peral River plume is more significant than the eastward extension. Also see the salinity distribution in Fig. 5a.
Figure 3. The source and location of wind, air temperature and wave measurements need to be presented.
Line 614-618 and Figure 8. The eastward transport of sediment is significant when winds and waves were not imposed in the model. It should be aware that the wind-driven northeastward shelf currents can be introduced from the open boundaries that were extracted from HYCOM that contains winds. Because the riverine sediment input is large in wet season and there is no wind/wave induced resuspension in dry season, the yearly averaged sediment flux has a strong northeastward component.

Citation: https://doi.org/10.5194/egusphere-2024-3289-RC1
- AC1: 'Reply on RC1', Wenping Gong, 19 Feb 2025
  
  Dear Reviewer:
  We greatly appreciate the time you spent reviewing our manuscript and providing constructive comments and minor revisions. We have made revisions in response to your suggestions.
  Comment:
  This study investigated the wind and wave effects on the dispersal of the Pearl River-derived sediment using a coupled numerical model. The results demonstrated the crucial role of wind in the westward along-shelf transport of the riverine sediments and the formation of the mud belt along the coasts. The diagnostics of cross-shelf current and suspended sediment concentration revealed the mechanism of winds in the formation of along-shelf currents, vertical mixing, and resuspension of sediment. This work is valuable for understanding the transport processes of the Pearl River sediment and the manuscript is well written. Therefore, I recommend publication of the manuscript after minor revision.
  Response:
  We very appreciate your comments for our manuscript. These comments and insights are helpful for improving the quality of our paper. We address your comments point-by-point as follows. We thank for the constructive comments from the reviewer.
  Comment:
  Were winds and waves included om the ramp simulation from Jan.1, 2016 to Apr. 1, 2017? The initial bed sediment distribution would be different for the two case simulations. Which case Figure 2 represents for?
  Response:
  Thank you for your comment. Yes, winds and waves were incorporated into the ramp simulation from January 1, 2016, to April 1, 2017. This period was selected to ensure that the model reached dynamic equilibrium before the main simulation phase. In total, we conducted seven cases: one Control case and six Sensitivity cases. Figure 2 presents the spatial distribution of seabed sediment fractions at the end of the spin-up phase in the Control run on April 1, 2017. This simulation accounts for winds, tides, waves, ambient shelf currents, and seasonal variations in critical erosion stress (i.e., increased critical erosion stress during winter). The same initial bed sediment distribution was applied to all seven cases. We have clarified this in the revised manuscript to ensure transparency and eliminate any ambiguity. Please let us know if any further clarification is needed.
  Comment:
  The critical shear stress chosen for clay is large. Although clay is cohesive, the large critical shear stress would restrict the transport of clay on the shelf and the riverine clay actually represents riverine sand. Comparing the initial prototype of bottom sediment with that after realistic reworking, the changes in slit component is the most significant because of the low critical shear stress.
  Response:
  Thank you for your valuable suggestion regarding the critical shear stress for sediment. In our initial setup, the critical shear stress was determined based on previous literature and calibration validation. However, we acknowledge that this approach may have limitations, particularly in capturing seasonal variations in sediment properties. Observations, laboratory experiments, and numerical sensitivity analyses from prior studies have demonstrated that the critical shear stress of sediments in the Pearl River Estuary is higher in winter than in summer.
  In response to your insightful comment, we have conducted additional numerical experiments incorporating seasonal variations in critical shear stress. The results reveal that the bed sediment grain size distribution after realistic reworking is now significantly closer to the initial prototype compared to our previous simulations, which did not account for such temporal variability. This improvement underscores the importance of considering seasonal changes in sediment properties for accurately modeling sediment dynamics. Specifically, the revised results highlight that the silt component, characterized by a lower critical shear stress, undergoes the most significant changes, aligning more closely with observed patterns.
  To further address this issue, we have revised our study and designed seven numerical cases. Six of these cases incorporate seasonal variations in critical erosion shear stress, while one case excludes this variation to analyze its impact on the ultimate fate of riverine sediments in the Pearl River Estuary. These updates have been integrated into the manuscript, along with a detailed discussion of their implications. Your feedback has greatly enhanced the robustness of our study, and we sincerely appreciate your contribution.
  Comment:
  Line 374-376. The westward extension of Peral River plume is more significant than the eastward extension. Also see the salinity distribution in Fig. 5a.
  Response:
  Thank you for your comment. The primary reason for this is that our previous analysis focused on the vertically averaged salinity field. In the revised manuscript, we now present both surface and bottom salinity fields. The surface salinity field more clearly highlights the significant eastward expansion of the plume. We have made the necessary modifications to the manuscript accordingly.
  Comment:
  Figure 3. The source and location of wind, air temperature and wave measurements need to be presented.
  Response:
  Thank you for your suggestion. The wind and air temperature data are sourced from NCEP, while the wave measurements are not based on observations but are derived from model results. We have clarified this in the manuscript.
  Comment:
  Line 614-618 and Figure 8. The eastward transport of sediment is significant when winds and waves were not imposed in the model. It should be aware that the wind-driven northeastward shelf currents can be introduced from the open boundaries that were extracted from HYCOM that contains winds. Because the riverine sediment input is large in wet season and there is no wind/wave induced resuspension in dry season, the yearly averaged sediment flux has a strong northeastward component.
  Response:
  Thank you for your invaluable feedback, which has significantly strengthened our study. In response to your comments, we have conducted seven additional numerical simulations to systematically evaluate the relative contributions of tides, waves, ambient shelf currents, seasonal variations in critical erosion stress, settling velocity, and the initial bed grain size distribution to the transport and dispersal of Pearl River-derived sediments. These simulations include one Control case and six Sensitivity cases. In our initial simulations, we inadvertently overlooked that the HYCOM-derived flow fields already account for the influence of monsoon-driven currents and Kuroshio intrusions on shelf circulation. This includes the strong southwestward currents in summer and the weaker southwestward or northeastward currents in winter. To address this, we have refined our approach in the NAS case (i.e., no ambient shelf currents) by excluding HYCOM-derived water levels and velocities. This adjustment allows for a clearer assessment of the influence of South China Sea (SCS) circulation while eliminating potential inaccuracies arising from monsoon-driven currents and Kuroshio intrusions. As a result, our model now provides a more accurate representation of the underlying physical processes. We believe these revisions enhance the robustness of our findings, and we sincerely appreciate your insightful suggestions. Please let us know if any further clarifications or refinements are needed.
  
  Citation: https://doi.org/10.5194/egusphere-2024-3289-AC1
RC2:
'Comment on egusphere-2024-3289', Anonymous Referee #2, 22 Jan 2025
Summary and General Comments:
I do not feel that the paper in its current form meets the standard expected for publication in Ocean Science. My general reasons are (1) some of the analysis seem poorly justified and detract from the main findings; (2) the organization should be improved to better communicate major findings; (3) the skill assessment has geographic and temporal limitations; and (3) the findings are not generalized to other systems. Despite these shortcomings, the paper has potential. The model quantifies suspended sediment dispersal on the shelf off the Pearl River Estuary (PRE) and compares dispersal during summer (wet season) and winter (energetic season). The authors explore the roles of winds and waves by comparing “Case 1” (includes winds and waves) to “Case 2” (neglects winds and waves); but the presentation of this could be improved.
Below are general points that if addressed would strengthen the paper.
Most sections of the Results include both “Case 1” and “Case 2”, and consideration of the wind/wave effects are interspersed with the seasonal variations and annual averages. For example, Figure 4g,h (section 4.1) has the impact of winds and waves. Then, Section 4.3 shows the annual averages for both Cases 1 and 2. The analyses would be clearer if the comparisons between Cases 1 and 2 were in one section.

Discussion Section 5.1 does not seem to contribute to the main conclusions, and does not seem to yield general knowledge for river-influenced shelves. My suggestion is to remove or shorten this section unless it is better justified.

Hydrodynamics: This section estimates idealized geostrophic velocities (Figure 11) and thermal wind shears (Figure 12). These analyses were unclear, and they complicate the paper and do not seem well justified. The numerical model solved the full momentum equations which would include the geostrophic and the thermal wind dynamics. What is gained from these analyses that are useful for the PRE and extendable to other locations?

Sediment Transport Analysis: Equation 2 provides the transport equation for suspended sediment. The annual averages of three of these terms are shown in Figure 13, but it is unclear how these were calculated. For example, HADV, units of mg/L/s, does this represent both the u- and v-component? Is it vertically averaged? Similar for the settling and resuspension terms: are these vertical averages? Near-bed values?

Lines 714 – 717 are somewhat general statements instead of building on the analyses presented in this section.

Discussion Section 5.2 should do a better job of extending this study within the literature. The authors cite papers but often do not provide the context. For example, Line 723 states that most of the Pearl River sediment stays near the estuary. Walsh’s and Hanebuth’s papers are cited, but this study is not related to the conceptual frameworks in these papers. Line 741 mentions a “12 km threshold” without any context. Lines 747 – 752 make an interesting point, that Walsh’ framework may not be directly suitable for monsoonal systems, particularly when the sediment delivery does not coincide with energetic conditions. This might be expanded and explained.

The conceptual model (Figure 14) presents schematics of dispersal dynamics under “normal” conditions (the annual-average?), and of a “no wind no waves” scenario. Figure 14 shows the idealized “no wind no waves” scenario, and the annual average that may not represent “normal conditions” because of the monsoonal nature. A conceptual model that compared the summer (high discharge, low wind/waves) to winter (low discharge, energetic winds/waves) would be interesting.

The model grid extending from Taiwan to the Beibu Gulf. The authors demonstrate model skill in the “Near” region of the model, and no observations from the “Western” or “Beibu Gulf” regions are shown (see Figure 1 for locations). Observations are limited to two months for water level and waves (Figure S1), and one week for S/T/SSC (Figure S2). However, the paper’s most interesting findings rely on the seasonal differences in hydrodynamics and sediment dispersal, and on delivery of sediment to distal regions such as Beibu Gulf. The model results would be more definitive if stations further from PRE were included, and if annual cycles were considered for at least some values (such as water level and wave height).

Figure S3 shows model skill for several stations that are near the PRE. These observations from August 2017 were used to justify model parameters, but sediment characteristics may vary seasonally. The paper would benefit from consideration of how uncertainties in sediment characteristics (settling velocities, critical shear stresses) propagate into uncertainties in their conclusions.

In many places the text is overly vague; including Section 5.3. Some examples are given below,

Specific Comments (by line number in the preprint)
The Introduction should better motivate the study and analyses. Many of the references are cited with no summary or context, making it difficult to see how this builds on past studies. The introduction should especially provide stronger motivation for the focus on winds and waves (Lines 78 – 79); and better place the regional setting of the PRE Shelf within the context of past research of the behavior of freshwater and sediment plumes under different wind and wave conditions.

Sections 3.1 – 3.3: The authors describe the model implementation, but it is impossible to include all of details. Many modeling papers instead provide a complete archive of the model input files (for example see the Data Availability statement in Moulton et al. JGR 2024 https://doi.org/10.1029/2023JC019685). Some issues that I see with the Methods description include:
Lines 142 – 144 provide model parameters theta_s and theta_b, but this is jargon of limited use even to ROMS modelers.

Line 171: provide a citation for the wave-current interaction.

Line 176: suggest being more specific, “The freshwater and sediment discharges for the Pearl River …”

Line 180: What is the evidence that one year of spin-up is sufficient?

Lines 188 – 190: Riverine sediment inputs was based on a rating curve and assumptions from earlier papers. The authors do not explain how they divide the total load from Zhang et al. (2012) into the distributary mouths. A 40% / 60% fraction was assumed for silt / clay based on earlier works (Table 1). Assumptions like these are necessary, but the paper does not address how uncertainties in these types of factors might impact the overall results.

Lines 190 – 233. The paper could benefit from consideration of the sensitivity of the results to uncertainties in the initial bed grain size distribution, and sediment properties (settling velocity, critical shear stress). Some specific examples:
The lack of grain size distributions in part of the model domain (i.e. the Beibu Gulf especially) lead to large uncertainties in transport there.

The text states that the “spin-up greatly reduced the irregularities” (line 231). But that is not evident in the figure.

Panels 2g-i show that spatial grain size variability develops for example at lon/lat 114E,19N. These features are also seen in the bed shear stress (Figures 4F, 4H). Are they a model artifact or instability? Are they aligned with bathymetry (bathymetric contour lines in Figure 1b might help).

Panel 2i shows that large parts of the grid become much sandier than the initial grain sizes.

Figures 9 and 10: The important points could be made with fewer figure panels.

Lines 564 – 566 and elsewhere. The paper mentions thermal winds, upwelling behavior, and a strong jet. The introduction does not provide the context for these processes. The paper does not identify these features within the model results. This adds complexity without reinforcing the main conclusions. I suggest removing these remarks here and elsewhere to focus on the main findings.

Conclusions should be strengthened to better emphasize novel findings of your paper. For example, lines 878 – 883 discuss the overall sediment dispersal patterns modeled, but do not emphasize the disconnect between sediment delivery (summer high discharge) and redistribution (winter high energy). Paragraph lines 884 – 890 also do not provide very much that is novel.

Technical Corrections (by line number in the preprint)
Figure 1: Bathymetry unclear. It is hard to see Taiwan Banks and offshore bathymetry that may impact shear stress (Figure 4f) and grain size (Figure 2g-i).

Lines 132 – 134 (etc.). Cite sources for component numerical models (ROMS, SWAN, CSTM etc). Provide citation for GEBCO and all similarly used data products.

In some places the text uses “Case 1” and “Case 2”, and in other places it uses “WW” and “NWW”. Use either “Case 1 and 2” OR “WW and NWW” in all places.

Perhaps add “Boundary Forcing” to Table 2 for ROMS (i.e. for Case 2: do you still use the same open boundaries for water levels, ubar/vbar, T/S, etc.?).

Line 372: It is vague to say “widespread expansion of the river plume”. Figure 5A does not look like the plume spreads far into the sea; it seems to hug the coast.

Line 389: vague to say “pathway is wide but magnitude is weak”. You could be less vague by giving quantitative scales here (and elsewhere).

Line 384: Confusing and vague to say “predominantly transported westward and eastward”. Suggest something like “there are both westward and eastward fluxes of riverine sediment (Figure 5c).”

Line 386: refers to the diversion around Taiwan Bank. However, this feature is not obvious in Figure 5c. Adding bathymetric contours that show Taiwan Bank may help. The time- and depth-averaged currents (figure 5a) show cross-shore variability; but the reported upwelling feature is not obvious.

Line 449 – 450: is it wind and wave “mixing” that impact the plume behavior; or wind and wave forcing which would include net momentum terms as well as turbulent mixing? Also, you are vague here saying “wider” instead of length scale for the width.

Lines 521 – 528 and associated text. Phrases like ‘Near region” and “Far off region” are awkward. Suggest “proximal” and “distal”.

Lines 555, 558, 561: the authors use the word “transported” but, then provide the percentages that are “retained”. Suggest using the words “deposited” or “retained” if you are talking about the sediment deposition.

Supplement Line 63: the 43 stations are not identified; text should refer to Figure 1.
Citation: https://doi.org/10.5194/egusphere-2024-3289-RC2
- AC2: 'Reply on RC2', Wenping Gong, 19 Feb 2025
  
  Dear Reviewer:
  We greatly appreciate the time you spent reviewing our manuscript and providing constructive comments. We have made revisions in response to your suggestions. Now we have conducted seven numerical simulations to systematically evaluate the relative contributions of tides, waves, ambient shelf currents, seasonal variations in critical erosion stress, settling velocity, and the initial bed grain size distribution to the transport and dispersal of Pearl River-derived sediments. These simulations include one Control case and six Sensitivity cases.
  Comment:
  I do not feel that the paper in its current form meets the standard expected for publication in Ocean Science. My general reasons are (1) some of the analysis seem poorly justified and detract from the main findings; (2) the organization should be improved to better communicate major findings; (3) the skill assessment has geographic and temporal limitations; and (3) the findings are not generalized to other systems. Despite these shortcomings, the paper has potential. The model quantifies suspended sediment dispersal on the shelf off the Pearl River Estuary (PRE) and compares dispersal during summer (wet season) and winter (energetic season). The authors explore the roles of winds and waves by comparing “Case 1” (includes winds and waves) to “Case 2” (neglects winds and waves); but the presentation of this could be improved.
  Response:
  We sincerely thank you for dedicating your valuable time and sharing your professional insights during the review of our manuscript. Your thoughtful comments have been instrumental in improving the quality of our paper and advancing its contribution to the academic community. Below, we provide a detailed, point-by-point response to your suggestions. We are deeply grateful for the constructive and insightful feedback provided by the reviewer.
  Comment:
  Below are general points that if addressed would strengthen the paper.
  Most sections of the Results include both “Case 1” and “Case 2”, and consideration of the wind/wave effects are interspersed with the seasonal variations and annual averages. For example, Figure 4g,h (section 4.1) has the impact of winds and waves. Then, Section 4.3 shows the annual averages for both Cases 1 and 2. The analyses would be clearer if the comparisons between Cases 1 and 2 were in one section.
  Response:
  Thank you for your insightful suggestion. We agree that organizing the comparisons between cases within a single section will enhance clarity and readability. In response, we have restructured the Results section to more clearly distinguish between seasonal variations, annual averages, and the Sensitivity cases.
  Comment:
  Discussion Section 5.1 does not seem to contribute to the main conclusions, and does not seem to yield general knowledge for river-influenced shelves. My suggestion is to remove or shorten this section unless it is better justified.
  Response:
  Thank you for your valuable suggestion. We have removed Section 5.1, as its content was not closely aligned with the main focus of this study and did not directly support our key findings. We agree that this section lacked significant insights for river-influenced shelves, and its removal enhances the clarity and focus of our discussion.
  Comment:
  Hydrodynamics: This section estimates idealized geostrophic velocities (Figure 11) and thermal wind shears (Figure 12). These analyses were unclear, and they complicate the paper and do not seem well justified. The numerical model solved the full momentum equations which would include the geostrophic and the thermal wind dynamics. What is gained from these analyses that are useful for the PRE and extendable to other locations?
  Response:
  Thank you for your valuable feedback. We acknowledge that the analyses of idealized geostrophic velocities and thermal wind shears may have complicated the manuscript and lacked clear justification. In response to your suggestion, we have removed this section from the revised manuscript to improve clarity and focus on the key findings. We appreciate your insightful comments, which have helped refine our study.
  Comment:
  Sediment Transport Analysis: Equation 2 provides the transport equation for suspended sediment. The annual averages of three of these terms are shown in Figure 13, but it is unclear how these were calculated. For example, HADV, units of mg/L/s, does this represent both the u- and v-component? Is it vertically averaged? Similar for the settling and resuspension terms: are these vertical averages? Near-bed values?
  Response:
  Thank you for your insightful comments. HADV represents the total horizontal advection of suspended sediment, incorporating both the u- and v-components. All the terms previously presented in Figure 13, including horizontal advection (HADV), settling, and resuspension, are vertically averaged values. However, we now realize that this section is not closely aligned with the main focus of our study. We have therefore removed this section to ensure that the manuscript emphasizes our primary findings more clearly.
  Comment:
  Lines 714 – 717 are somewhat general statements instead of building on the analyses presented in this section.
  Response:
  Thank you for your suggestion. We have now removed this content.
  Comment:
  Discussion Section 5.2 should do a better job of extending this study within the literature. The authors cite papers but often do not provide the context. For example, Line 723 states that most of the Pearl River sediment stays near the estuary. Walsh’s and Hanebuth’s papers are cited, but this study is not related to the conceptual frameworks in these papers. Line 741 mentions a “12 km threshold” without any context. Lines 747 – 752 make an interesting point, that Walsh’ framework may not be directly suitable for monsoonal systems, particularly when the sediment delivery does not coincide with energetic conditions. This might be expanded and explained.
  Response:
  Thank you for your constructive feedback. We have made appropriate revisions to Section 5.2 to better contextualize our citations. We believe these modifications enhance the clarity and depth of our discussion.
  Comment:
  The conceptual model (Figure 14) presents schematics of dispersal dynamics under “normal” conditions (the annual-average?), and of a “no wind no waves” scenario. Figure 14 shows the idealized “no wind no waves” scenario, and the annual average that may not represent “normal conditions” because of the monsoonal nature. A conceptual model that compared the summer (high discharge, low wind/waves) to winter (low discharge, energetic winds/waves) would be interesting.
  Response:
  Thank you for your insightful comment. We acknowledge your point and agree that our previous conceptual model may not have adequately captured the monsoonal variability. In response, we have removed the original conceptual diagram to better align with the study’s focus.
  Comment:
  The model grid extending from Taiwan to the Beibu Gulf. The authors demonstrate model skill in the “Near” region of the model, and no observations from the “Western” or “Beibu Gulf” regions are shown (see Figure 1 for locations). Observations are limited to two months for water level and waves (Figure S1), and one week for S/T/SSC (Figure S2). However, the paper’s most interesting findings rely on the seasonal differences in hydrodynamics and sediment dispersal, and on delivery of sediment to distal regions such as Beibu Gulf. The model results would be more definitive if stations further from PRE were included, and if annual cycles were considered for at least some values (such as water level and wave height).
  Response:
  Thank you for your valuable comments. We sincerely appreciate your suggestions and have taken steps to enhance the validation of our model results. First, we would like to clarify that the water level validation in the Pearl River Estuary (PRE) region we previously provided already covered a full year (12 months). In response to your suggestion, we have now included additional validation at stations farther from the PRE. Specifically, we have added water level validation (12 months) at two stations: Zhapo in the "Western" region and Qinglan in the "Distal" region. Additionally, we have incorporated a 40-day mooring-based current velocity validation in the "Southwestern" region and a one-week wave validation in the "Beibu Gulf" region. We hope these additional validations strengthen the reliability of our study’s findings and provide a more comprehensive assessment of the model’s performance. Thank you again for your insightful feedback.
  Comment:
  Figure S3 shows model skill for several stations that are near the PRE. These observations from August 2017 were used to justify model parameters, but sediment characteristics may vary seasonally. The paper would benefit from consideration of how uncertainties in sediment characteristics (settling velocities, critical shear stresses) propagate into uncertainties in their conclusions.
  Response:
  Thank you for your valuable suggestion. Now we have conducted seven numerical simulations to systematically evaluate the relative contributions of tides, waves, ambient shelf currents, seasonal variations in critical erosion stress, settling velocity, and the initial bed grain size distribution to the transport and dispersal of Pearl River-derived sediments. These simulations include one Control case and six Sensitivity cases. We have now incorporated the seasonal variation of sediment characteristics in six simulation cases, specifically by increasing the critical shear stress for sediment erosion during winter, based on findings from previous studies. Exp 5 (NVS) was identical to Control case, except that it did not account for the seasonal variation in critical erosion stress, keeping the winter and summer values the same. Exp 6 (DSV) was identical to Control case, except that it set double sediment settling velocity of the Control case. Additionally, we have analyzed these results and included a discussion in the main text to assess their impact on our conclusions. We appreciate your insightful feedback, which has helped improve the robustness of our study.
  Comment:
  In many places the text is overly vague; including Section 5.3. Some examples are given below,
  Specific Comments (by line number in the preprint)
  The Introduction should better motivate the study and analyses. Many of the references are cited with no summary or context, making it difficult to see how this builds on past studies. The introduction should especially provide stronger motivation for the focus on winds and waves (Lines 78 – 79); and better place the regional setting of the PRE Shelf within the context of past research of the behavior of freshwater and sediment plumes under different wind and wave conditions.
  Response:
  Thank you for your helpful suggestion. We have made the corresponding revisions.
  Comment:
  Sections 3.1 – 3.3: The authors describe the model implementation, but it is impossible to include all of details. Many modeling papers instead provide a complete archive of the model input files (for example see the Data Availability statement in Moulton et al. JGR 2024 https://doi.org/10.1029/2023JC019685). Some issues that I see with the Methods description include:
  Lines 142 – 144 provide model parameters theta_s and theta_b, but this is jargon of limited use even to ROMS modelers.
  Response:
  Thank you for your valuable suggestion. We understand that certain model parameters, such as theta_s and theta_b, may not be immediately familiar to all readers. Additionally, we recognize the difficulty of including every detail of the model implementation within the manuscript. In response to your comment, we will provide key configuration files (such as "ocean.in" and "sediment.in") as part of the supplementary materials and explicitly reference this in the Data Availability statement. This will enable readers to directly access and review the essential model parameters. We greatly appreciate your feedback and believe these additions will enhance the transparency, reproducibility, and accessibility of our study.
  Comment:
  Line 171: provide a citation for the wave-current interaction.
  Response: Thank you for pointing that out. We have made the necessary revision and added the appropriate citation for the wave-current interaction.
  Comment:
  Line 176: suggest being more specific, “The freshwater and sediment discharges for the Pearl River …”
  Response:
  Thank you for your helpful suggestion. We have made the corresponding revisions.
  Comment:
  Line 180: What is the evidence that one year of spin-up is sufficient?
  Response:
  Thank you for your comment. In our initial setup, the critical shear stress was determined based on previous literature and calibration validation. However, we acknowledge that this approach may have limitations, particularly in capturing seasonal variations in sediment properties. Observations, laboratory experiments, and numerical sensitivity analyses from prior studies have demonstrated that the critical shear stress of sediments in the Pearl River Estuary is higher in winter than in summer. In response to your insightful comment, we have conducted additional numerical experiments incorporating seasonal variations in critical shear stress. The results reveal that the bed sediment grain size distribution after realistic reworking is now significantly closer to the initial prototype compared to our previous simulations, which did not account for such temporal variability. This improvement underscores the importance of considering seasonal changes in sediment properties for accurately modeling sediment dynamics.
  Comment:
  Lines 188 – 190: Riverine sediment inputs was based on a rating curve and assumptions from earlier papers. The authors do not explain how they divide the total load from Zhang et al. (2012) into the distributary mouths. A 40% / 60% fraction was assumed for silt / clay based on earlier works (Table 1). Assumptions like these are necessary, but the paper does not address how uncertainties in these types of factors might impact the overall results.
  Response:
  Thank you for your feedback. In the revised manuscript, we have provided the calculation formula for the rating curve, which is used to estimate the riverine sediment input. Based on this relationship, the total amount of Pearl River sediment input over our 12-month study period was calculated to be 34.52 million tons. This value aligns closely with the annual load reported in 2017 by the Pearl River Water Resources Commission. The riverine sediment input, derived from the river discharge, was allocated across the eight outlets located on the northern boundary (Figure 1b) based on the distribution approach of Hu et al. (2011). Additionally, in the discussion section, we have addressed the potential impact of uncertainties associated with the 40% / 60% fraction of silt and clay. We acknowledge that while such assumptions are necessary, they may introduce uncertainties, and we have discussed how these uncertainties could influence the overall results.
  References
  Hu, J., Li, S., and Geng, B.: Modeling the mass flux budgets of water and suspended sediments for the river network and estuary in the Pearl River Delta, China, Journal of Marine Systems, 88, 252-266, 10.1016/j.jmarsys.2011.05.002, 2011.
  Comment:
  Lines 190 – 233. The paper could benefit from consideration of the sensitivity of the results to uncertainties in the initial bed grain size distribution, and sediment properties (settling velocity, critical shear stress). Some specific examples:
  The lack of grain size distributions in part of the model domain (i.e. the Beibu Gulf especially) lead to large uncertainties in transport there.
  Response:
  Thank you for your valuable feedback regarding the sensitivity of the results to uncertainties in the initial bed grain size distribution and sediment properties. In response, we conducted seven different cases to systematically evaluate the influence of these factors, including variations in the initial bed grain size distribution, settling velocity, and critical shear stress. Our findings indicate that the grain size distribution has a relatively minor impact on the transport of Pearl River sediment on the continental shelf. Instead, the sediment transport dynamics are predominantly governed by physical factors such as hydrodynamic conditions, tidal forces, and regional circulation patterns. This suggests that while uncertainties in grain size distribution exist, particularly in areas like the Beibu Gulf, their effect on the overall sediment transport is limited compared to the dominant physical drivers.
  Comment:
  The text states that the “spin-up greatly reduced the irregularities” (line 231). But that is not evident in the figure.
  Response:
  Thank you for your comment. We have incorporated seasonal variations in critical erosion stress into our model and re-simulated the results. The revised findings now demonstrate a closer alignment with the initial grain size distribution, effectively addressing the discrepancy identified in the previous analysis. This refinement offers a more precise representation of sediment dynamics in the study area.
  Comment:
  Panels 2g-i show that spatial grain size variability develops for example at lon/lat 114E,19N. These features are also seen in the bed shear stress (Figures 4F, 4H). Are they a model artifact or instability? Are they aligned with bathymetry (bathymetric contour lines in Figure 1b might help).
  Response:
  The features observed at 114E,19N are located near the model boundary where water depths are relatively large and boundary effects may be significant. However, this location is quite far from our main study region of interest. Based on our analysis, these boundary features have minimal impact on the key conclusions of this study. To provide more context, we have now added bathymetric contour lines to the relevant figures (e.g. Figure 1b). The bathymetry shows that 114E,19N is in a deep area near the model boundary. While some model artifacts or instabilities may occur in this boundary region, they do not affect the primary results and interpretations for our main study area. The spatial grain size variability and bed shear stress patterns in our core study region are robust and not influenced by these distant boundary effects.
  Comment:
  Panel 2i shows that large parts of the grid become much sandier than the initial grain sizes.
  Response:
  Thank you for highlighting the observation in Panel 2i. We have incorporated seasonal variations in critical erosion stress into our model and re-simulated the results. The revised findings now demonstrate a closer alignment with the initial grain size distribution, effectively addressing the discrepancy identified in the previous analysis. This refinement offers a more precise representation of sediment dynamics in the study area.
  Comment:
  Figures 9 and 10: The important points could be made with fewer figure panels.
  Response: Thank you for your suggestion. We have now removed the original Figures 9 and 10, as we have significantly revised the paper by conducting seven cases. We found that the original Figures 9 and 10 were not closely related to the content of the revised manuscript and appeared redundant.
  Comment:
  Lines 564 – 566 and elsewhere. The paper mentions thermal winds, upwelling behavior, and a strong jet. The introduction does not provide the context for these processes. The paper does not identify these features within the model results. This adds complexity without reinforcing the main conclusions. I suggest removing these remarks here and elsewhere to focus on the main findings.
  Response: Thank you for your comment. We agree with your suggestion and have removed this context, as it was not directly related to our main findings. This revision streamlines the paper and ensures a clearer focus on the key conclusions.
  Comment:
  Conclusions should be strengthened to better emphasize novel findings of your paper. For example, lines 878 – 883 discuss the overall sediment dispersal patterns modeled, but do not emphasize the disconnect between sediment delivery (summer high discharge) and redistribution (winter high energy). Paragraph lines 884 – 890 also do not provide very much that is novel.
  Response:
  Thank you for your valuable feedback. In response to your suggestion, we have revised the conclusion to more clearly highlight the novel findings of our paper, particularly the disconnect between sediment delivery during the summer high-discharge period and redistribution during the winter high-energy period. Additionally, we have removed paragraph lines 884–890, as they did not provide significant new insights. We hope these revisions enhance the clarity and focus of our conclusions.
  Comment:
  Technical Corrections (by line number in the preprint)
  Figure 1: Bathymetry unclear. It is hard to see Taiwan Banks and offshore bathymetry that may impact shear stress (Figure 4f) and grain size (Figure 2g-i).
  Response:
  Thank you for your feedback regarding Figure 1. We have taken your comments into consideration and have revised the figure to improve the clarity of the bathymetry.
  Comment:
  Lines 132 – 134 (etc.). Cite sources for component numerical models (ROMS, SWAN, CSTM etc). Provide citation for GEBCO and all similarly used data products.
  Response:
  Thank you for your feedback. We have made the necessary revisions as per your suggestions.
  Comment:
  In some places the text uses “Case 1” and “Case 2”, and in other places it uses “WW” and “NWW”. Use either “Case 1 and 2” OR “WW and NWW” in all places.
  Response:
  Thank you for your feedback. We have made the necessary revisions as per your suggestions.
  Comment:
  Perhaps add “Boundary Forcing” to Table 2 for ROMS (i.e. for Case 2: do you still use the same open boundaries for water levels, ubar/vbar, T/S, etc.?).
  Response:
  Thank you for your invaluable feedback, which has significantly strengthened our study. In response to your comments, we have conducted seven additional numerical simulations to systematically evaluate the relative contributions of tides, waves, ambient shelf currents, seasonal variations in critical erosion stress, settling velocity, and the initial bed grain size distribution to the transport and dispersal of Pearl River-derived sediments. These simulations include one Control case and six Sensitivity cases. In our initial simulations, we inadvertently overlooked that the HYCOM-derived flow fields already account for the influence of monsoon-driven currents and Kuroshio intrusions on shelf circulation. This includes the strong southwestward currents in summer and the weaker southwestward or northeastward currents in winter. To address this, we have refined our approach in the NAS case (i.e., no ambient shelf currents) by excluding HYCOM-derived water levels and velocities. This adjustment allows for a clearer assessment of the influence of South China Sea (SCS) circulation while eliminating potential inaccuracies arising from monsoon-driven currents and Kuroshio intrusions. As a result, our model now provides a more accurate representation of the underlying physical processes. We believe these revisions enhance the robustness of our findings, and we sincerely appreciate your insightful suggestions.
  Comment:
  Line 372: It is vague to say “widespread expansion of the river plume”. Figure 5A does not look like the plume spreads far into the sea; it seems to hug the coast.
  Response:
  Thank you for your comment. The primary reason for this is that our previous analysis focused on the vertically averaged salinity field. In the revised manuscript, we now present both surface and bottom salinity fields. The surface salinity field more clearly highlights the significant eastward expansion of the plume. We have made the necessary modifications to the manuscript accordingly.
  Comment:
  Line 389: vague to say “pathway is wide but magnitude is weak”. You could be less vague by giving quantitative scales here (and elsewhere).
  Response:
  Thank you for your feedback. We have made the necessary revisions as per your suggestions.
  Comment:
  Line 384: Confusing and vague to say “predominantly transported westward and eastward”. Suggest something like “there are both westward and eastward fluxes of riverine sediment (Figure 5c).”
  Response:
  Thank you for your feedback. We have made the necessary revisions as per your suggestions.
  Comment:
  Line 386: refers to the diversion around Taiwan Bank. However, this feature is not obvious in Figure 5c. Adding bathymetric contours that show Taiwan Bank may help. The time- and depth-averaged currents (figure 5a) show cross-shore variability; but the reported upwelling feature is not obvious.
  Response:
  Thank you for your suggestion. We have now added bathymetric contours to Figure 5 and have redrawn the figure to illustrate both surface and bottom currents. The bottom currents clearly show strong cross-shore upwelling.
  Comment:
  Line 449 – 450: is it wind and wave “mixing” that impact the plume behavior; or wind and wave forcing which would include net momentum terms as well as turbulent mixing? Also, you are vague here saying “wider” instead of length scale for the width.
  Response:
  Thank you for your suggestion. We have now removed this section. Regarding the role of wind, we have chosen not to consider it in this study. The results from the NWS-Control case only represent the estimation bias associated with neglecting wave effects. We have ensured that wind stress is consistent across all our simulations. In this study, wave forcing includes both the net momentum terms generated by the waves and turbulent mixing.
  Comment:
  Lines 521 – 528 and associated text. Phrases like ‘Near region” and “Far off region” are awkward. Suggest “proximal” and “distal”.
  Response:
  Thank you for your suggestion. We have now updated the text by replacing "Near region" and "Far off region" with "proximal" and "distal" to ensure consistency and clarity.
  Comment:
  Lines 555, 558, 561: the authors use the word “transported” but, then provide the percentages that are “retained”. Suggest using the words “deposited” or “retained” if you are talking about the sediment deposition.
  Response:
  Thank you for your suggestion. We have now adopted the term "retained" in our text.
  Comment:
  Supplement Line 63: the 43 stations are not identified; text should refer to Figure 1.
  Response:
  Thank you for your feedback. We have made the necessary revisions as suggested, and the text now refers to Figure 1 for the identification of the 43 stations.
  
  Citation: https://doi.org/10.5194/egusphere-2024-3289-AC2

Peer review completion

AR: Author's response | RR: Referee report | ED: Editor decision | EF: Editorial file upload

AR by Wenping Gong on behalf of the Authors (12 Mar 2025) Author's response Author's tracked changes Manuscript

ED: Referee Nomination & Report Request started (14 Mar 2025) by Mario Hoppema

RR by Anonymous Referee #2 (27 Mar 2025)

Suggestions for revision or reasons for rejection

General comments and summary:

I have read the responses to both reviews of the initial submission, and the revision. The authors have done a conscientious job of revising the manuscript to address concerns. Note that this is a very different paper than the initial submission and has a different focus than the original (e.g. “winds” was removed from the title). The initial submission compared two model runs (“with winds and waves” and “without winds and waves”). The revision compares SEVEN model runs, only one of which was in the first submission. The revision removed two discussion sections.

This submission seems much closer to the standard expected for your journal, but the paper will benefit from some additional revisions, as outlined below.

My review of the first submission raised four major concerns: “(1) some of the analysis seem poorly justified … ; (2) the organization should be improved …; (3) the skill assessment has geographic and temporal limitations; and (4) the findings are not generalized to other systems.” The authors have addressed most of these; they removed the analyses that were poorly justified, improved organization, and included a more comprehensive skill assessment in the Supplement.

Below are general comments that will improve the paper:

1. The introduction is improved upon the original submission. It is better organized and includes more context. However, most of the citations are based on fairly localized, often small-spatial scale, studies of estuaries. This seems inappropriate, given that the model domain is a ~1700 km scale continental shelf. The introduction should rely on continental shelf papers moreso than estuarine. Some suggestions (Geyer et al. 2004; McKee et al 2004; etc.); continental shelf modeling papers (Warner et al. 2017; Dalyander et al, 2013; Harris et al. 2008; Wang et al. 2020; Zang et al. 2019; Xu et al. 2016).

2. The paper models the dispersal of the Pearl River sediment on the continental shelf. However, in many cases, the paper presents the study as if it were a Pearl River Estuary (PRE) study rather than a continental shelf study. Most of the analyses is done for the region of the shelf outside of the PRE (i.e. Regions 2 – 8 on Figure 1). In these cases, the paper should refer to the shelf region rather than the estuary.

3. The abstract and motivation should be more clear about what will be analyzed. The analyses focuses on the dispersal of Pearl River discharged sediment (classes 4 and 5); and not the seabed sediment (seabed and riverine, classes 1 – 3)?

Specific Comments (by line number in the preprint)

4. Lines 297 – 328: This section is confusing and should be revised for clarity. For example, better clarify whether Pearl River sediment (classes 4 and 5) was delivered during the 15-month spinup.

5. Section 2.6 is confusing. Perhaps a table would be helpful. Some specific points:
5a. Lines 413 – 414: Run 5 uses the same critical shear stress for winter and summer, but the paper does not say what value is used?
5b. Lines 416 – 423; model run 7 is described as considering the impact of the initial sediment bed. But, it actually seems to evaluate whether the original spinup time was sufficient for supplying riverine sediment (classes 4 -5) to the seabed. It does not seem to consider uncertainties in the distribution of size classes 1- 3.
5c. Lines 423 – 427 seem out of place. It provides the wind forcing used for all model runs. Perhaps move this to the beginning of the section.

6. Section 3.3 is difficult to follow, suggest revising it. It is long (12 pages). Each of the 5 figures contains several panels that each compare a different model to the Control run. Each paragraph of the text, however, discusses one model run, so the reader needs to look at several figures to understand each paragraph.

Suggest you redo the figures to be organized like the text. If a paragraph describes one model run (e.g. NTS), it can refer to 1 figure that shows the comparison between the model and control for multiple variables (bed stress, SSC, currents, flux, etc).

7. Section 4.1 should be edited for clarity, and should be shortened.
7a. One example, lines 829 – 837 describe using a critical shear stress that is seasonally dependent, but it is confusing whether the authors feel that this was important. Line 837: “it has a more significant effect on the bed sediment grain size distribution”: not sure what this means and it does not reference a figure.
7b. Suggest removing the “Cycle 2” description or moving it to the supplement (Lines 847 – 857).
7c. Lines 802 – 902 could provide more comparisons with the results from this study and previously published papers, especially for continental shelves. Papers have shown evidence for seasonal dependence in critical shear stress (Xu et al. 2014; Briggs et al. 2015); and sensitivities to settling velocity (Harris et al. 2008).

8. Section 4.2 the writing could be more clear. Could probably be shortened.

9. Lines 999 – 1019: these paragraphs do not add much to the paper. They summarize the findings from the results but do not compare this to other systems, and do not put the results into context.

10. Conclusion section 5. This seems like a summary and a restating of the results section. Suggest revising this so that the major messages are reinforced without repeating other parts of the paper. I would avoid the “underestimation” and “overestimation” phrases (e.g. lines 1085 – 1107). The lessons learned by the reduced physics processes are that tides, waves, and remote forcing impact sediment dispersal and are important drivers of sediment transport.

Also, I do not think that the “initial sediment bed” experiments (Cycle and Cycle2) are truly tests of the initial bed so much as a test of the length of the spinup time for riverine sediment dispersal.

Technical Corrections (by line number in the preprint)

11. Line 236: Confusing and vague wording: “More than one year of hydrodynamic and sediment spin-up is sufficient…”.

12. Lines 502, 504, etc: the units of sediment flux seem incorrect. g-1 m s-1?

13. Figures 5 and 6: it is difficult to compare the sediment panels (c-f) because they use very different color scales.

14. Line 613: are the “red and blue values” on Figure 7?

15. Line 698: These represent the concentrations of classes 4 and 5 and do not include classes 1 – 3? Suggest you clarify this in the figure caption and perhaps in the text.

16. Figure 12 is complicated. Can it be simplified by removing some of the numbers?

17. Line 827: it seems inappropriate to refer to the original manuscript (preprint).

18. Table 1, Line 863, and elsewhere: it seems a bit misleading to cite the Ralston paper for the settling velocities used for the riverine sediment because Ralston used a salinity-dependent settling velocity, whereas this paper partitioned sediment into fast- and slow-settling flocs.

19. Line 918: suggest you say “finding is consistent with the earlier…” instead of “confirm.

20. Line 934: suggest saying “Within this framework, riverine sediment deposition is characterized using …”.

21. Lines 971 – 975: this is confusing because you cite several studies of estuaries. However, your numerical model focuses on the continental shelf dispersal. Should you instead say “The PRE exhibits distinctive geomorphological features, yet dispersal of its fine-grained sediment on the continental shelf conforms to general patterns observed offshore of other monsoon-influenced estuarine systems. … documented offshore of various major …”.

22. Line 980 and elsewhere: Instead of saying “Like the PRE”, suggest you say “Like the PRE Shelf”. This paper focuses on the continental shelf offshore of the PRE, not on the PRE itself. The paper should be more clear about this.

23. Lines 1026 – 1035: many of these citations are focused on the estuary itself or on global drivers, rather than continental shelf processes. There are no citations for line 1034 (episodicity of shelf transport for “many systems”).

24. Line 1040: can you be more specific? How much “less” freshwater and sediment do these small rivers contribute? Otherwise, this paragraph is a bit vague, and it does not put this sytem into the context of other systems where the relative contributions of small and large rivers have been considered.

25. Line 1084: should this say “in the PRE” or “within and offshore of the PRE”? The model domain extends very far beyond the estuary.

26. Figure S1: use different types of markers (circles, squares, triangles, etc.) for the different types of observation points (tide gages, wave stations, survey stations, etc).

27. All figures should have the units (for example, figure S7C does not identify the units of SSC mg/L?; Figures S9 and S10 do not provide the units of velocity)

28. The font sizes in many figures is very small and hard to read, especially in the supplement figures.

Hide

RR by Anonymous Referee #3 (28 Mar 2025)

Suggestions for revision or reasons for rejection

1. Introduction
This is my first time reviewing this manuscript, and my assessment has been conducted independently without taking earlier versions and previous referee comments into account.

Using a numerical case study, this paper investigates the dispersal of river-derived sediment over a continental shelf. It does well to provide insights into the spreading of sediment from the river plume over the coastal shelf, taking several processes, such as tides, waves and wind into account. While I think the paper does provide new insights into these main dispersal mechanisms, I think the attempt to form a generalised interpretation is currently not attained, and it might be more appropriate to shift the focus more to the specific case that is investigated. As I do see the value of the developed model and the on/off sensitivity analysis of several model aspects, I think a more careful presentation of the results is required for suitable embedding. Therefore, I recommend publication of the manuscript after major revision.

2. General Comments
Clarification of “Overestimation” and “Underestimation”
The manuscript frequently refers to the “overestimation” or “underestimation” of sediment-related quantities (e.g., deposition, resuspension, retention, and bed shear stress) in several instances (e.g., lines 24, 27, 631, 667, 830, 1004, 1011, 1085, among others). These terms imply comparison with real-world field data, which, to my knowledge, is not available for validation (as also noted in lines 805–808). Instead, these quantities appear to be evaluated relative to the “Control” model run. To avoid potential misinterpretation, I recommend revising the phrasing to clarify that differences are being assessed within the model framework rather than against observational data.

Evaluation of Research Objective (2)
It is unclear whether research objective (2) (line 156) has been fully achieved. My interpretation is that the model is calibrated based on hydrodynamics to establish the “Control” run, but it remains uncertain whether the model continues to behave as expected when certain processes (NWS, NTS, NAS) are omitted. Additionally, it would be beneficial to clarify whether the chosen methodology is capable of addressing objective (2).

There is potential for deeper insights into the contributions and interactions of different processes in sediment transport. Since sediment transport exhibits nonlinear responses to environmental forcing, an analysis of how tides and waves interact to influence overall sediment dynamics could strengthen the study. Currently, the approach primarily isolates individual processes, but a more integrated interpretation of process interactions could add value.

Some of the study’s findings may already be anticipated based on fundamental process-based reasoning. For instance, the following statements confirm well-known expectations:
• Lines 525–527: “With sediment load during winter nearly negligible, the suspended concentration of riverine sediment is significantly lower compared to the wet summer.”
• Lines 657–658: “For the NTS case versus the Control case, tides significantly affected bottom stress.”
• Lines 667–668: “However, NWS underestimated the nearshore bottom stress.”
• Lines 713–714: “This enhanced settling velocity resulted in an increased deposition.”

A more nuanced discussion of these results—particularly emphasizing interactions between multiple processes rather than confirming expected trends—would enhance the manuscript.

Numerical Model Performance and Limitations
The manuscript could benefit from a more detailed discussion of numerical model performance, particularly regarding:
• Grid sensitivity: While the hydrodynamic validation appears strong, has numerical behaviour and model convergence been tested under grid refinement?
• Model limitations: Section 4.3 primarily discusses the exclusion of certain physical processes but does not address intrinsic model limitations. Consider including a discussion on potential numerical constraints.
• Sigma-coordinates (Line 189): The sigma-coordinate system is known to introduce challenges in accurately modelling salt transport in regions with steep gradients, where Cartesian coordinates may perform better (Bijvelds, 2001). Given the importance of salinity in estuarine turbidity maxima (ETM) formation, it would be helpful to clarify whether ETM development is well captured in the model and whether any limitations arise from the chosen coordinate system.

The study presents interesting insights into sediment budgets and overall sediment dynamics. In particular, the introduction of the “Cycle” model run is a compelling aspect. It may be beneficial to highlight its implications more prominently, especially regarding the conclusions drawn in lines 730–731.

3. Technical comments
The level of significance reported for measured percentages is inconsistent across the manuscript. Given the model’s inherent (in)accuracy, the precision of percentage values should be adjusted accordingly. This applies particularly to the Abstract, Section 3.2, and Section 5 (Conclusions).

Line 192: Song and Haidvogel (1994) introduce a new s-coordinate system, not the general sigma-coordinate system introduced in the manuscript. It is unclear which system is actually used and what the two θ-values are, as these are not explicitly described in Song and Haidvogel (1994). Please clarify and revise accordingly.

Lines 303-305. The statement “The realistic spin-up greatly reduced the irregularities and prepared a more suitable seabed sediment particle size distribution field for subsequent simulations than the initial prototype (Figures 2d-f vs. 2g-i).” needs further clarification. Could you elaborate on why this “realistic reworking” is necessary and why it results in a more suitable initial sediment distribution? The figures suggest the presence of spurious oscillations in the model interior and near the open southern boundary, which raises concerns.

Line 483-487. The precise definition of a “river plume” is unclear. How is it quantitatively defined, and what correlation does it have with regions of high suspended sediment concentration (SSC)? Additionally, a more in-depth analysis of the freshwater plume’s role in transporting riverine sediment—relative to the overall sediment transport—would strengthen this section. Could you expand on this aspect?

Line 493: Please explain how the sediment flux is determined. Is it computed purely as advective horizontal transport, or are other processes included? Please specify.

Line 541. Figure 6 shows significant salinity gradients near the estuary, which suggests the potential formation of an estuarine turbidity maximum (ETM). Is such an ETM captured in the model? If so, could you comment on its role in sediment transport within the system?

Line 639: the term “probably being artefacts” is vague. Could you clarify what kind of numerical artifacts these are? While they appear as local deviations in deeper regions, they span several kilometres and multiple grid cells. Given their scale, could they significantly influence hydrodynamics across the domain? Furthermore, how does this uncertainty impact confidence in the NAS model results? A more detailed explanation would be helpful.

Line 694+733. Figures 9 and 10 include arrows that are somewhat unclear. While the colours indicate differences between model runs, it appears that the arrows represent only the non-Control run. Could you clarify their specific purpose and whether they provide a direct comparison with the Control case?

Lines 728-731 describe the deposition seen in Figure 11f but provide no reason for it by the driving hydrodynamics. To me, it is unclear why the sediment distribution would change. What is the process behind this? Does this not provide us with the conclusion that a more complete sediment spin-up, as performed in Cycle, is necessary for robust results?

Line 793. The caption of Figure 12 explains the significance of magenta values, but it is unclear what the other values (black, blue, red) represent. Could you specify their meaning?

4. Textual remarks
Line 25: The term “distal retention” may not be appropriate in this context. If sediment is never transported to distal regions in the first place, it does not necessarily imply a lower distal retention. Consider rephrasing for clarity.

Line 219: The abbreviation Hsig appears without an introduction. If it refers to significant wave height, please define it properly when first mentioned.

Line 298: I am not sure what “realistic reworking” means, could you explain or rephrase?

Line 495: The sediment flux is not strictly westward and eastward. Please consider rephrasing for accuracy.

Line 502 and 504. The units used for sediment flux are not correct -> g/m/s

Line 524: The term “salinity front” is somewhat vague. Does this refer to a specific isohaline? Please clarify.

Line 542. Figure 6e: “2D riverine Flux” is unclear phrasing to me. Perhaps “Horizontal sediment flux” would be a more precise alternative, as “riverine” may be ambiguous without additional context..

Line 542. Figure 6f. The term “riverine Deposition reworking” is unclear. Consider rephrasing to include “difference,” “change,” or another term that better captures the intended meaning.

There are some inconsistencies in terminology throughout the manuscript. For example, “Distal” regions is sometimes written in quotation marks, while in other places it appears as distal regions without emphasis. Similarly, “Gulf” regions and gulf regions are used interchangeably. Please ensure consistency in terminology throughout the text.

References
Bijvelds, M. D. J. P. (2001). Numerical modelling of estuarine flow over steep topography. Doctoral dissertation, TU Delft, Delft University of Technology. Ipskamp. Retrieved from http://resolver.tudelft.nl/uuid:39c0c858-579a-47fa-a911-4a4114949a11

Referee Report: PDF

Hide

ED: Reconsider after major revisions (31 Mar 2025) by Mario Hoppema

AR by Wenping Gong on behalf of the Authors (07 May 2025) Author's response Author's tracked changes Manuscript

ED: Referee Nomination & Report Request started (14 May 2025) by Mario Hoppema

RR by Anonymous Referee #3 (26 May 2025)

Suggestions for revision or reasons for rejection

I have read the revision and the responses to both reviews of the second submission. I think the authors have done a satisfying job in responding to all comments and have provided a new manuscript that presents its results much more successfully. The submission communicates the dominant processes that lead to the spreading of river-derived sediment over the continental shelf well.

I advise accepting this paper for publication, after some technical adjustments:

1. Line 99: Is it possible to make the URL a reference?

2. Lines 160-163 and Line 171. In my opinion, manuscript-specific jargon should not appear in the introduction. This especially refers to the references to sediment classes and Table 1 in lines 156 – 175. Please consider rephrasing to retain generality in the introduction and specify further details in the methodology section 2.

3. Line 278, please consider more commonly used variables such as c or C_s for concentration and Q for discharge, similar to Zhang et al., (2012).

4. Line 299, Figure 2: Please rephrase “After realistic reworking” to “After spin-up”

5. Line 651, Figure 7: Suggestion, is it worth it to add red and blue arrows as well to indicate the flux direction of each sediment class?

6. Line 703-706. Please rephrase this sentence. The words “while” and “but” together seem to be used incorrectly?

7. Line 876: Please consider removing the term “realistic reworking”.

8. Line 1067 and Line 1072, Table 3: While I appreciate the efforts of adding this information, I do not think it is a suitable addition to this manuscript. Highlighting that the PRE is dominant, with 83% of the sediment load, is sufficient to attain the goal of the paper. Advise is to remove Table 3.

9. Lines 1167-170. To me, this was a confusing addition for me at the end of the paper. Please avoid introducing a new term and explaining its meaning (Cycle2) in the last sentence of the manuscript.

Hide

RR by Anonymous Referee #2 (30 Jun 2025)

Suggestions for revision or reasons for rejection

General comments and summary:

I have read the responses to reviews of the resubmission, and the revision. The authors have done a conscientious job of revising the manuscript to address concerns and the paper is improved. This submission seems closer to the standard expected for your journal, but in my opinion is not there. While the paper is improved, it suffers from being almost entirely focused on the numerical model with less attention paid to the process-based implications of the results. The organization of the paper should be improved to increase clarity and reduce redundancy.

General Comments:
1. The paper would be more impactful if it emphasized more the lessons learned about dispersal processes, rather than emphasizing the work of modeling dispersal.

1a. This is reflected in the current title of the paper which focuses on the act of implementing a model; not the process-based results derived from the model, or the timescales covered. Is that appropriate for this journal?

1b. In comparing the cases studies to the control run: the paper presents these in terms of the act of modeling dispersal instead of what is learned about processes via the comparison. For example, in the abstract (lines 27 – 29) “the absence of tidal forcing reduces bottom shear stress…”; (lines 32 - 33) when “ambient currents” are “omitted". These describe what happens in the model when forcings are removed from the model. The paper would have more impact if it emphasized what these results show about the importance of tides and large-scale currents in dispersing sediment. These are two examples from the abstract, but the paper takes this approach for all the comparisons of case studies to the control run.

1c. For the “cycle experiment”, the results show that riverine sediment continues to be dispersed at timescales that exceed 1 year. This is not surprising for such a large-domain model. The analysis does not provide insight into the actual residence times expected.

1d. Some model experiments focus on forcing (tides, waves, large-scale currents), and others on sediment properties (critical shear stress, settling velocity, seabed grain sizes). The analysis is presented in terms of model inputs (if tides are omitted, then sediment is retained in PRE, etc.); rather than what can be learned from the experiments. The paper would be more impactful if it phrased results in terms of the relative impacts of different forcings (for example: which has more effect, the tides or the waves?); and relative impacts of uncertainty in sediment properties.

1e. Finally, the conclusions (Section 5) are very focused on modeling rather than lessons learned regarding the system as a whole. This makes the paper less impactful than it could be.

2. Comparison of 1-year model to Holocene records (Pages 51, etc). It is interesting to compare patterns calculated by the model for 1-year to the longer-term geologic record; but the paper would be more impactful if the differences in the timescales were acknowledged. Also, the paper needs to be clear about whether (or how) the geologic data were used to calibrate sensitive model parameters (settling velocity).

2a. Text starting line 929 should begin a new paragraph. This compares the 1-year model results to Holocene sediment thicknesses (6,500 years). Were the sediment settling velocities “calibrated” to match this result? If so this should not be cited as confirmation that the partitioning is valid.

2b. It is not clear how the values in lines 947 – 950 were scaled to bridge the large differences in timescales.

3. The parametrization of the sediment model remains confusing, even after revision. Part of the confusion comes from the fact that the paper is not well organized and not concise. Topics are repeated in different parts of the paper and sometimes seem to contradict each other.

3a. For example: the setting of riverine sediment properties includes the choices of 60%/40% split and settling velocities (0.005 and 0.6 mm/s). These are said to follow Bever and MacWilliams (line 350); Zhang et al 2019 (line 359, 388). Then, line 366 lists 6 source papers plus “model calibration” but does not refer to the Supplement. Line 897 attributes the choice of settling velocities to Xia et al. and calibrations and cites the Supplement. The Supplement shows modeled and observed SSC but does not mention calibration of ws.

Looking through the source papers, the 60/40 split is from Zhang et al. 2019. The choice of settling velocities seem to be calibrated but is not clearly described. The authors need to be concise and clear about selection of these important values.

3b. Another example is treatment of riverine sediment for the spin-up and cycling cases. The authors revised this but not achieved clarity and conciseness. At Line 351, the paper states that after the spin-up class 4 and 5 sediment were “added as class 1 and 2” (do the authors mean “added to class 1 and 2”?). Then in line 464 the “Cycle experiment” is described in a confusing way. After several readings: I think that the difference between the Control Run and the Cycle experiment is whether the Class 4&5 sediment from the “spin-up” period was added to Class 1&2 in the initial bed before the model run.

4. Section 4.2 is improved but the writing is still a bit rough. The paper does not provide the context of Walsh and Nittrouer’s classification scheme. Lines 976 and 986 seem contradictory (line 976 says your results are “unlike” Walsh & Nittrouer’s framework, whereas Line 9787 says your results are “consistent”).

5. Page 58 and 59 use a lot of text to argue for more sensitivity tests to better constrain sediment dispersal. This can probably be shortened. But more importantly, the argument should be that more DATA is needed. Without additional data streams, there seems little argument for doing more model runs. The paper might be more impactful if the authors provided insight into what type of data would be most useful.

Specific Comments (by line number in the preprint)

6. Line 32, Lines 452, 453, 471: use of the term “Ambient circulation” is not clear. You need to define what you mean by this (large, non-local scale forcing?).

7. Line 44 (abstract): use of the word “dispersal mechanisms” seems a mis-statement. “Dispersal mechanisms” are things like bedload, suspended load, gravity flows. This paper looks at the sensitivity of suspended transport and dispersal to forcings and parameters.

8. Introduction seems a bit long and could be shortened. Some of the text seems too specific and is redundant with later methods (for example lines 159 – 163).

9. Paragraph starting at line 328: Revise for clarity. Suggest (line 333) “initially estimated sediment distribution to evolve…”; (line 335) “introduced by Kriging, sparse or problematic data”; (line 338) “after the spin-up period is thought to exhibit spatial patterns that are better aligned…”.

10. Lines 406 – 419: this paragraph provides information about the climatology of summer conditions. It is unclear what these values represent. Are these values climatological averages over several years? (if so, data sources should be cited and the time-period noted). Or are they the values from the 1-year period modeled?

11. Section 3.2, Page 33: I found this description confusing.

12. Figures 7 and 12: These figures are still confusing and there is probably a more effective way to present this data. The numbers are hard to see and hard to compare. The percentages are relative to the “total riverine load” which was 40% for class 4 and 60% for class 5. The blue numbers therefore show that *all* of class 5 is retained in the PRE; except Figure 12(f) where 59.99% remains.

13. Table 3: this is interesting but seems tangential to the paper and adds ½ a page to an already long paper.

Text suggestions:

14. The writing could be improved. I will not try to provide an exhaustive list, but here are some specific examples.

15. Line 104: “SCS shaped by the East Asian Monsoon”. That is confusing, the sea is not “shaped” by the monsoon.

16. Line 143: “Holocene sedimentary processes”: do you mean “Holocene sedimentary deposits”?

17. Line 156: suggest the word “factors” or “drivers” instead of “processes”.

18. Line 241: suggest “Water level and current velocity”.

19. Line 296, 300 and elsewhere: suggest “1,981” instead of “1981”.

20. Line 743: “Vectors show the seasonal mean surface current fields in each experiment”. This is unclear because the figure panels are difference plots; not plots for a single experiment.

21. Line 901: suggest “all flocs in the model are retained…”. Can the authors use this model result to conclude something about the processes that are important?

22. Avoid terms in the conclusion that are very specific to this paper; e.g. “regions 3-4” (Line 1150).

23. Data availability: The authors’ data products should be listed at the beginning instead of at the end of this section. Lines 1193 – 1194 provide the link to the authors’ own data products. Lines 1181 – 1192 provide links to data that were not generated by the authors.

Hide

ED: Publish subject to minor revisions (review by editor) (08 Jul 2025) by Mario Hoppema

AR by Wenping Gong on behalf of the Authors (14 Jul 2025) Author's response Author's tracked changes Manuscript

ED: Publish subject to technical corrections (28 Jul 2025) by Mario Hoppema

Dear Dr. Gong and co-authors,

Your manuscript is now accepted for publication in Ocean Science. I am satisfied with your response to the referee's comments and changes made in the manuscript. Before submitting the final manuscript files, please consider the below listing of final technical comments from my side:

L25 Please do not use PRE here (it was not defined), but write out full. Moreover, it is not used further on in the abstract anymore.
L143 Beibu Gulf is mentioned, but it is not clear from the map in Fig. 1 where that is located. Please add the Beibu Gulf to the map.

L259-260 Please write data format like: 31 March 2018 (also at other places in the text)
L428-429 In the caption of Figure 3 at (c) it is written: “daily (black) and monthly (red) averaged” This apparently should occur at (d)

L480 delete respectively (not needed here)
Table 2: It would be helpful for reminding to have the meaning of the abbreviation letters written down in the table

L502, 507, 509 Is this a real increase or do you mean it is just higher in winter? I think the latter, so please use exact wording.
L521 delete clear

L527-533 This sentence is very long and therefore hard to understand. Please rephrase.
L536 larger instead of higher

L650 Format 1 April 2017 and elsewhere
Section 3.3 It would be helpful to the reader to mention (for example, in brackets) now and again what the different experiments stand for, as this is not easy to remind.

L845 is instead of: may be The real-world situation is extremely complex …
L849 we discuss the fidelity (delete: intend to)

L871 Not sure if prototype is correct here
L963 delete (Walsh and Nittrouer, 2009) (because already cited at the beginning)

L1069 Entrapment of what is meant here? Sediment entrapment?
L1198 doctoral dissertation (add dissertation)

L1413 delete Volume, add article number
L1417 delete Volume, add article number

L1491 This needs more info, web link?
L1492-1493 Please change capitals to lower case letters

L1510-1511 Any more info? Web link, institute/university?

With best wishes
Mario Hoppema

Hide

AR by Wenping Gong on behalf of the Authors (29 Jul 2025) Manuscript

Short summary

The study reveals strong seasonal patterns in Pearl River-derived sediment transport: calm summers favor deposition via buoyant plumes, while winter winds and waves enhance resuspension and southwestward transport into the Beibu Gulf. Tides drive estuarine export, waves dominate coastal resuspension, and shelf circulation promotes summer eastward spread. Settling velocity, spin-up duration, and seasonal changes in critical shear stress also affect sediment retention and transport pathways.