Three-Dimensional Hydrodynamic and Sediment-Transport Modeling of a Shallow Urban Lake in the Brazilian Amazon
Round 1
Reviewer 1 Report
Comments and Suggestions for AuthorsThe manuscript presents a three‑dimensional (3‑D) hydrodynamic and sediment‑transport simulation of Água Preta Lake (APL), a shallow (max. 4 m depth) urban lake located in Belém, Brazil. The authors use the Delft3D‑FLOW model to examine how wind forcing and the artificial inflow from the Guamá River influence circulation, temperature stratification, and sediment transport during November 2023 – September 2024. The current manuscript extends this by employing 3‑D hydrodynamics coupled with a cohesive sediment module, which allows examination of vertical stratification, reverse flow, and sediment resuspension. I recommend publication of the manuscript with major revisions. Here are my comments:
Meteorological forcing is derived from a local station (air temperature, humidity, wind speed/direction) and a national station for solar radiation and precipitation. The inflow discharge series provided by COSANPA is modified by a factor of 5 “to better reflect actual data”. The manuscript does not justify this large adjustment or provide an independent measurement to support it. Similarly, inflow water temperature (29 °C) and sediment concentration (0.2 kg m⁻³) are “empirically assumed”. These assumptions directly influence sediment deposition patterns and should either be measured or sensitivity‑analysed. Air temperature, humidity and wind are obtained from a local station; solar radiation and precipitation from a station 3 km away. Solar radiation is multiplied by 1.8 to compensate for high cloudiness and wind speed is multiplied by 1.5 to adjust from land to lake. These corrections lack supporting data. Since heat‑flux calculations depend strongly on solar radiation and wind, the authors should justify or calibrate these factors (e.g., by comparing simulated and observed net radiation or by using onsite meteorological measurements at 10 m height).
The bottom roughness is determined using Manning coefficients based on grain size and vegetation, which is appropriate. However, the sediment module uses default cohesive sediment parameters and a morphological scaling factor (MORFAC) without specifying its value. Morphological acceleration can distort deposition predictions; the authors should state and justify MORFAC and perform sensitivity analysis.
Temperature at one point (A17) is used for calibration, yielding RMSE 0.27 °C and MAE 0.87 °C. Calibrating only one variable (temperature) may be insufficient because the model also simulates currents and sediments. Calibration against measured velocities, sediment concentrations or turbidity would improve confidence.
The study identifies low velocities with higher speeds near inflow/outflow, reverse flow at depth, and stronger currents during the dry season due to wind. These findings are plausible and agree with previous understanding of wind‑driven dynamics in shallow lakes. However, the authors sometimes draw broad ecological conclusions (e.g., enhanced macrophyte growth, eutrophication) without supporting measurements (nutrient concentrations, macrophyte density). Discussions of potential ecological impacts should be framed cautiously as hypotheses.
Suspended sediment originates mainly from the Guamá inflow and moves toward central/western sectors, causing deposition near the inflow and an average bed elevation increase of 0.05 m. The authors argue that this promotes turbidity and resuspension, potentially triggering eutrophication. While the model predicts deposition patterns, it does not simulate turbidity or nutrients, so statements about eutrophication should be couched as expectations rather than conclusions. Comparing the predicted sediment accumulation rate with previous estimates (23 000–29 000 m³ yr⁻¹) would strengthen the analysis.
Overall, the manuscript is understandable, but the English needs improvement. Some sentences are long or grammatically awkward (e.g., “The meteorological parameters of the region have particular features in relation to the urban environment in its surroundings”). It would benefit from professional editing or review by a native speaker.
Figures are not visible in the text provided to the reviewer; however, descriptions suggest that maps and zonation diagrams are included (Figures 1–10). Ensure that figures have scale bars, north arrows, labelled axes, and colour‑blind‑friendly palettes. Table 3 summarises validation metrics at 19 stations; it is extensive and could be moved to supplementary material with a condensed version in the main text to improve readability.
Cite and discuss earlier modelling studies of APL (e.g., Holanda et al., 2011), highlighting how the current 3‑D model improves upon the 2‑D simulation and how the predicted sedimentation rates compare. Also reference other 3‑D lake modelling studies (e.g., Lake Taihu, Okeechobee, Lake Biwa) to situate the research within broader literature.
Author Response
We sincerely thank the reviewer for taking the time to carefully evaluate our manuscript and provide constructive feedback. Please find our detailed point-by-point responses below. All corresponding revisions and corrections have been incorporated into the manuscript and are highlighted using track changes in the re-submitted files.
Meteorological forcing is derived from a local station (air temperature, humidity, wind speed/direction) and a national station for solar radiation and precipitation. The inflow discharge series provided by COSANPA is modified by a factor of 5 “to better reflect actual data”. The manuscript does not justify this large adjustment or provide an independent measurement to support it. Similarly, inflow water temperature (29 °C) and sediment concentration (0.2 kg m⁻³) are “empirically assumed”. These assumptions directly influence sediment deposition patterns and should either be measured or sensitivity‑analysed. Air temperature, humidity and wind are obtained from a local station; solar radiation and precipitation from a station 3 km away. Solar radiation is multiplied by 1.8 to compensate for high cloudiness and wind speed is multiplied by 1.5 to adjust from land to lake. These corrections lack supporting data. Since heat‑flux calculations depend strongly on solar radiation and wind, the authors should justify or calibrate these factors (e.g., by comparing simulated and observed net radiation or by using onsite meteorological measurements at 10 m height).
A: Thank you for your detailed and constructive feedback. In response, we have added a paragraph explaining that a one-at-a-time (OAT) sensitivity analysis was performed to assess the influence of key parameters on model performance. This analysis involved systematically adjusting individual input parameters—such as solar radiation, wind speed, and inflow conditions—and evaluating their effect on simulated temperature profiles. A supporting citation has been included to reference similar sensitivity approaches applied in lake modeling studies.
The scaling factors applied to wind speed (×1.5) and solar radiation (×1.8) were derived from this sensitivity analysis to better match observed thermal profiles and compensate for local atmospheric conditions, such as persistent cloud cover and the overland-to-water transition. The suspended sediment concentration (0.2 kg m⁻³) was empirically assumed and tested during this analysis, once there is a lack of data in sediment that enters through the lake artificial channel, we could not support this data. However, for the inflow temperature (29 °C) it was supported by regional studies and observed seasonal conditions (Pires et al. 2024). These updates provide a stronger empirical and analytical basis for the parameter choices that influence sediment dynamics.
The bottom roughness is determined using Manning coefficients based on grain size and vegetation, which is appropriate. However, the sediment module uses default cohesive sediment parameters and a morphological scaling factor (MORFAC) without specifying its value. Morphological acceleration can distort deposition predictions; the authors should state and justify MORFAC and perform sensitivity analysis.
A: We used the default value of MORFAC = 1. This ensures that morphological changes occur on the same time scale as the hydrodynamic simulation, avoiding distortions in deposition predictions.
Temperature at one point (A17) is used for calibration, yielding RMSE 0.27 °C and MAE 0.87 °C. Calibrating only one variable (temperature) may be insufficient because the model also simulates currents and sediments. Calibration against measured velocities, sediment concentrations or turbidity would improve confidence.
A: We calibrated the model using the temperature time series at point A17, achieving an RMSE of 0.27 °C and an MAE of 0.87 °C. We then validated the model using measured flow velocities and temperature profiles. This validation step provides statistical confidence that the simulation adequately represents the lake dynamics, including hydrodynamics and sediment transport. The validation process included velocity measurements at two locations and 19 temperature profiles collected over a four-month period.
The study identifies low velocities with higher speeds near inflow/outflow, reverse flow at depth, and stronger currents during the dry season due to wind. These findings are plausible and agree with previous understanding of wind‑driven dynamics in shallow lakes. However, the authors sometimes draw broad ecological conclusions (e.g., enhanced macrophyte growth, eutrophication) without supporting measurements (nutrient concentrations, macrophyte density). Discussions of potential ecological impacts should be framed cautiously as hypotheses.
A: We acknowledge that some ecological implications were previously stated as conclusions. Therefore, we revised these sentences to present these results as hypotheses rather than definitive conclusions.
Suspended sediment originates mainly from the Guamá inflow and moves toward central/western sectors, causing deposition near the inflow and an average bed elevation increase of 0.05 m. The authors argue that this promotes turbidity and resuspension, potentially triggering eutrophication. While the model predicts deposition patterns, it does not simulate turbidity or nutrients, so statements about eutrophication should be couched as expectations rather than conclusions. Comparing the predicted sediment accumulation rate with previous estimates (23 000–29 000 m³ yr⁻¹) would strengthen the analysis.
A: We agree that the model does not explicitly simulate turbidity or nutrient dynamics, so we have revised the related statements to present them as expected implications rather than definitive conclusions.
Additionally, we have incorporated the comparison of the predicted sediment accumulation rate with previous estimates (23,000–29,000 m³ yr⁻¹) both in the Abstract and the Morphological Results sections to strengthen the analysis.
Overall, the manuscript is understandable, but the English needs improvement. Some sentences are long or grammatically awkward (e.g., “The meteorological parameters of the region have particular features in relation to the urban environment in its surroundings”). It would benefit from professional editing or review by a native speaker.
A: Our manuscript was translated by a professional service; however, as our native language is Brazilian Portuguese, some sentences initially retained a certain elaborateness and length typical of Portuguese writing style. Following your comment, I have carefully reviewed the manuscript and revised lengthy or awkward sentences to improve clarity and readability.
Figures are not visible in the text provided to the reviewer; however, descriptions suggest that maps and zonation diagrams are included (Figures 1–10). Ensure that figures have scale bars, north arrows, labelled axes, and colour‑blind‑friendly palettes. Table 3 summarises validation metrics at 19 stations; it is extensive and could be moved to supplementary material with a condensed version in the main text to improve readability.
A: While we agree that the table is extensive, I believe it is a crucial part of the manuscript to ensure that the simulations were well calibrated and validated. Therefore, we prefer to keep it in the main text to ensure the results are fully accessible to readers.
Cite and discuss earlier modelling studies of APL (e.g., Holanda et al., 2011), highlighting how the current 3‑D model improves upon the 2‑D simulation and how the predicted sedimentation rates compare. Also reference other 3‑D lake modelling studies (e.g., Lake Taihu, Okeechobee, Lake Biwa) to situate the research within broader literature.
A: I found this commentary very productive so I added in the results section, before the calibration subsection, a discussion that empathizes on what our simulations brings of new and different to study area and situates it in other literatures of 3D models.
Author Response File: 
Author Response.pdf
Reviewer 2 Report
Comments and Suggestions for AuthorsThe study presents a technically detailed and valuable investigation into the hydrodynamics and sediment transport of Á gua Preta Lake in Brazil using a 3D model (Delft3D-FLOW). The content is strong scientifically, but some improvements are needed. These improvements are:
- The abstract and introduction sections are very long-winded, use simpler language, and have shorter sentences, focusing on problems, methodology, solution and results.
- The Delft3D model is well-described, but all software has its assumptions, which are not described in the paper.
- The authors didn’t perform the sensitivity analysis to show how changes in key parameters affect model outputs.
- A proper justification is required for why only cohesive sediments are modeled, with an assumed fixed concentration of 0.2 kg/m³ at the inflow.
- Why shouldn’t authors have used a finer mesh? or a mesh justification is required.
- The model outputs focus primarily on temperature and velocity, with limited use of sediment transport results. Why?
- RMSE, MAE, R², and Pearson’s r are used to check the model accuracy, but their spatial/temporal variation is not deeply analyzed.
- The conclusions are required to be specific and well-explained.
Author Response
|
Comments 1: The abstract and introduction sections are very long-winded, use simpler language, and have shorter sentences, focusing on problems, methodology, solution and results.
|
|
Response 1: Thank you for your insightful comments regarding the Abstract and Introduction. In response, we have made several revisions to improve clarity, focus, and conciseness. The Abstract was reformulated to highlight the main results more directly, incorporating key quantitative data and using simpler, more concise language. In the Introduction, we identified and revised overly long sentences that affected readability. Paragraphs were restructured (e.g., lines 31–40 and 67–77) to improve flow and comprehension. Redundant content—such as repeated mentions of the ecological services provided by lakes—was consolidated into a single, streamlined sentence. Furthermore, in lines 84–93, we strengthened the justification for using a three-dimensional model over simpler one- or two-dimensional approaches. We also added a new paragraph emphasizing the relevance of temperature as a key variable in hydrodynamic modeling. These changes aim to make both sections more accessible and aligned with the expectations for scientific clarity.
|
|
Comments 2: The Delft3D model is well-described, but all software has its assumptions, which are not described in the paper.
|
|
Response 2: Thank you for pointing this out. We agree that it is important to acknowledge the underlying assumptions of the Delft3D model. Accordingly, we have added a paragraph to the manuscript outlining the main assumptions, including the use of the hydrostatic approximation, the Boussinesq assumption. This addition helps clarify the scope and limitations of the modeling approach.
|
|
Comments 3: The authors didn’t perform the sensitivity analysis to show how changes in key parameters affect model outputs.
|
|
Response 3: Thank you for your detailed and constructive feedback. While a full sensitivity analysis was not the primary objective of this study, we acknowledge its importance in supporting the reliability of the model results. In response, we have added a paragraph to the manuscript explaining that a one-at-a-time (OAT) sensitivity analysis was performed to evaluate the influence of key input parameters on model performance. This analysis involved systematically varying individual parameters—such as solar radiation, wind speed, and inflow conditions—and assessing their effect on the simulated temperature profiles. Additionally, we included a supporting citation referencing similar sensitivity approaches in lake modeling studies to provide context for the methodology used (Duka et al. 2024).
|
|
Comments 4: A proper justification is required for why only cohesive sediments are modeled, with an assumed fixed concentration of 0.2 kg/m³ at the inflow.
|
|
Response 4: The scaling factors applied to wind speed (×1.5) and solar radiation (×1.8) were derived from a sensitivity analysis to better align the model outputs with observed temperature profiles and to account for local atmospheric conditions, such as persistent cloud cover and the transition from land-based to over-water wind speeds. The suspended sediment concentration at the inflow (0.2 kg m⁻³) was empirically assumed due to the lack of direct measurements from the artificial channel. This value was tested through the same sensitivity analysis and was found to yield consistent sedimentation patterns within the expected range. Although we acknowledge the limitation of not having field data for this parameter, this approach provided a practical and stable input for the simulation. In contrast, the inflow water temperature (29 °C) was supported by regional studies and observed seasonal conditions (Pires et al., 2024), offering a more robust empirical basis. These clarifications and supporting information have been added to the parameterization section of the manuscript.
|
|
Comments 5: Why shouldn’t authors have used a finer mesh? or a mesh justification is required.
|
|
Response 5: Thank you for your comment. The model uses a horizontal grid resolution of 20×20 meters and includes 10 vertical sigma layers, which we consider sufficiently refined for capturing the key hydrodynamic and sediment transport processes in a shallow lake of this size. This resolution represents a balance between numerical accuracy and computational efficiency. A finer mesh was evaluated during the initial model setup but resulted in a significantly higher computational cost without a proportionate gain in model performance or output quality. The current resolution allows for accurate representation of flow features near inflow and outflow zones, as well as vertical stratification, while maintaining model stability and reasonable run times.
|
|
Comments 6: The model outputs focus primarily on temperature and velocity, with limited use of sediment transport results. Why?
|
|
Response 6: Thank you for your observation. In response, we have expanded the section on sediment transport results to include a more detailed analysis of sedimentation patterns and accumulation within the lake. This addition strengthens the connection between hydrodynamic conditions and sediment dynamics, providing a more comprehensive interpretation of the model outputs. The revised section now includes spatial distribution maps and a discussion of key depositional zones. We note that the analysis of temperature and velocity was more extensive due to the need to validate the model using spatial and temporal statistical metrics, which required detailed discussion. Nonetheless, we have improved the balance of the results section by enriching the sediment transport componente.
|
|
Comments 7: RMSE, MAE, R², and Pearson’s r are used to check the model accuracy, but their spatial/temporal variation is not deeply analyzed.
|
|
Response 7: Thank you for your comment. We applied RMSE, MAE, R², and Pearson’s metrics to validate the model across both spatial and temporal dimensions. Spatially, these statistics were computed for 19 temperature profiles collected at different locations within the lake. Temporally, the metrics were applied to time series of current velocities and temperature at key monitoring points. To address spatial and temporal variability, we carefully examined locations and time periods where the model showed poorer statistical performance. These analyses helped us identify possible reasons for discrepancies and improve model calibration.
|
|
Comments 8: The conclusions are required to be specific and well-explained. |
|
Response 8: Thank you for this valuable suggestion. In response, I revised the final paragraph of the manuscript to better highlight the significance of applying numerical modeling to this lake system, particularly in the context of urban and environmental management. Additionally, a new paragraph was added to discuss some limitations of the model, including its reduced accuracy in capturing peak temperatures. Moreover, I included a brief discussion on how the model outputs could support decision-making and improve management practices, such as identifying zones of sediment accumulation, assessing the impact of artificial inflows, and evaluating the effects of future climate or land-use changes |
Author Response File: Author Response.pdf
Reviewer 3 Report
Comments and Suggestions for AuthorsThis paper presented the outputs from a 3-D hydrodynamic and sediment transport model in an urban shallow lake in Brazil where site specific data were used to parameterise and calibrate the model. With a focus on water temperature pattern and profile of an important waterbody, the research topic fits into the journal scope and the use of a sophisticated model will attract wide range of readers. However, there are some issues which need to be addressed:
Introduction: the current research is about the application of an existing model for a specific site. More are expected on why the model was chosen and why other more simpler model can not deliver the information required. Less on why lakes are important and their broad ecological significances. Since calibration is mainly on water temperature, it is recommend to explain the significance of water temperature in active processes in the lake.
Methodology: Since there is no model development / modification, there is no need for the detailed model process description (section 2.3.1 and 2.3.2). They could be moved to a supplementary document if necessary. More should be on what options are available in the model, why certain option are chosen, what are the main model data inputs and key parameters and how site specific parameters were sourced and specified.
Conclusions: the limitations of the modelling approach has to be discussed. It will be good to discuss how the modelled outputs could be used to improve management practices and
Reference: Are all the reference necessary? Please check as they are too many for a paper of this length.
Figures and tables:
Some suggestions for figures and tables were made in the attached annotated paper.
Comments for author File: Comments.pdf
English writing need to be improved significantly. There are some long sentences and difficult to comprehend phrases which have been highlighted in the annotated version of the paper. It has to be said that no proof reading was done for the whole paper by this reviewer.
Author Response
Thank you for your valuable and constructive feedback. I have carefully considered each of your comments and revised the manuscript accordingly. Your detailed suggestions in the PDF were particularly helpful in guiding the improvements. Below, I outline the main changes made to enhance the quality of the manuscript.
|
Comments 1: The current research is about the application of an existing model for a specific site. More are expected on why the model was chosen and why other more simpler model can not deliver the information required. Less on why lakes are important and their broad ecological significances. Since calibration is mainly on water temperature, it is recommend to explain the significance of water temperature in active processes in the lake.
|
|||||
|
Response 1: Thank you for your insightful comments regarding the Introduction. I found your recommendations highly pertinent and have revised the section accordingly to enhance clarity and readability. Specifically, I divided several overly long paragraphs (e.g., lines 31–40 and 67–77) to improve the flow of information. I also consolidated redundant content—such as repeated references to the ecological services provided by lakes—into a single, more concise sentence. Additionally, in lines 84–93, I strengthened the justification for using a three-dimensional model over one- or two-dimensional approaches. A new paragraph was also added to highlight the relevance of temperature as a key variable in hydrodynamic modeling.
|
|||||
|
Comments 2: Since there is no model development / modification, there is no need for the detailed model process description (section 2.3.1 and 2.3.2). They could be moved to a supplementary document if necessary. More should be on what options are available in the model, why certain option are chosen, what are the main model data inputs and key parameters and how site specific parameters were sourced and specified.
|
|||||
|
Response 2: Regarding Sections 2.3.1 and 2.3.2: I appreciated your suggestion to move these to the supplementary material to make the manuscript more concise. However, other reviewers requested further explanation on the use of the morphological factor and the selection of Heat Flux Model 1. Therefore, I included a brief explanation within the main text to address both concerns.
|
|||||
|
Comments 3: The limitations of the modelling approach has to be discussed. It will be good to discuss how the modelled outputs could be used to improve management practices and
|
|||||
|
Response 3: Thank you for this valuable suggestion. In response, I revised the final paragraph of the manuscript to better highlight the significance of applying numerical modeling to this lake system, particularly in the context of urban and environmental management. Additionally, a new paragraph was added to discuss some limitations of the model, including its reduced accuracy in capturing peak temperatures. Moreover, I included a brief discussion on how the model outputs could support decision-making and improve management practices, such as identifying zones of sediment accumulation, assessing the impact of artificial inflows, and evaluating the effects of future climate or land-use changes.
|
|||||
|
Comments 4: Are all the reference necessary? Please check as they are too many for a paper of this length. |
|||||
|
Response 4: I reviewed the reference list to ensure all citations are relevant and appropriately used. While the number of references may appear high, this reflects the extensive research required to support the development of a 3D hydrodynamic model, especially given the limited number of publications that offer detailed guidance on this type of modeling. Thank you again for your thoughtful feedback, which greatly contributed to strengthening the manuscript.
|
|||||
|
|
Author Response File: Author Response.pdf
Round 2
Reviewer 2 Report
Comments and Suggestions for AuthorsAccepted
Author Response
Thank you for your time and thoughtful feedback throughout the review process. We are very grateful for your recommendation to accept the manuscript. Your comments helped significantly in the manuscript improvement
Best regards,
Reviewer 3 Report
Comments and Suggestions for AuthorsI have reviewed the original version of the paper. The authors have made efforts to address my concerns and the paper has been improved. There are still some minor issues but I support the paper's publication on technical and scientific grounds.
Comments for author File: Comments.pdf
I am still not quite happy with English writing, especially the introduction part. I hope it could be improved further. I have not checked the full text but highlighted some paragraphs as examples in the annotated version of the paper provided.
Author Response
|
Comments 1: I have reviewed the original version of the paper. The authors have made efforts to address my concerns and the paper has been improved. There are still some minor issues but I support the paper's publication on technical and scientific grounds.
|
|||||
|
Response 1: Thank you for your support in the paper, I read the attached PDF and made changes in the paper language, specially in introduction and methodology. This made the reading clearer, and it really improved the manuscript understandment.
|
|||||
|
Comments 2: Little surprised here. Why do you need one minute resolution to cover the dry and rainy seasons? What changes were made to the methodological data?
|
|||||
|
Response 2: Thank you for your question. This comment helped me realize that the original paragraph needed clarification. A one-minute time step was chosen to accurately simulate the dynamic processes in the lake domain, as it is used in the hydrodynamic equations. Higher temporal resolution improves the accuracy of the modeled parameters. However, if the time step is too large, it can violate the assumptions of the numerical scheme and lead to instabilities. Given the fine spatial grid used in our simulations, time steps larger than one minute caused significant errors and reduced model accuracy. We tested larger time steps to optimize computational efficiency, but a one-minute resolution provided a stable and reliable simulation. Regarding methodological adjustments, the meteorological forcing data for wind speed and solar radiation were scaled based on sensitivity analyses to better match observed temperature profiles, as described in the parameterization section. These modifications ensure that the high-resolution simulation reliably represents lake dynamics over the full seasonal cycle.
|
|||||
|
|
|||||
|
Author Response File: Author Response.pdf
