the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 29 Nov 2023
The contribution of diminishing river sand loads to beach erosion worldwide
Abstract. The erosion of sandy beaches can have a profound impact on human activities and ecosystems, especially on developed coasts. The scientific community has, to date, primarily focused on the potential impact of changes in sea level and waves on sandy beaches. While being abundantly recognized at local to regional scales in numerous studies over the last two decades, the contribution of diminishing fluvial sediment supply to sandy beach erosion at the global scale is still to be investigated. Here, we present a model of sediment budget computed from the balance between land riverine input and coastal transport. It results in a global picture of sand pathway from land to sea. Our analysis demonstrates the massive impact of the thousands of river dams on beach erosion worldwide. Sand can be mobilized with wave-induced longshore transport over large distances, in general toward the equator, within sediment cells, often at distance from the outlets.
This preprint has been withdrawn.
-
Withdrawal notice
This preprint has been withdrawn.
-
Preprint
(1975 KB)
-
Supplement
(1241 KB)
-
This preprint has been withdrawn.
- Preprint
(1975 KB) - Metadata XML
-
Supplement
(1241 KB) - BibTeX
- EndNote
Interactive discussion
Status: closed
-
RC1: 'Comment on nhess-2023-165', Anonymous Referee #1, 28 Dec 2023
The manuscript explores the contribution of diminishing sediment supply from rivers to global beach erosion. In particular, the manuscript analyzes the impact of thousands of dams on beach erosion worldwide. The topic is of great interest to NHESS readers; the text is of appropriate length and is easy to read and understand for a broad and diverse readership, including fellow scientists.
To achieve their goal, the authors develop a sediment conservation model that calculates the balance between sediment input from rivers and longitudinal transport along the coast due to wave action. This model is implemented on the world´s coast segmented into alongshore cells of 0.5º (11,161 cells of about 50 km each). The model results in sediment accumulation/erosion patterns in each cell. The authors emphasize that the conversion between these sediment accumulation/erosion patterns and coastal morphological changes (advance/recession) is not straightforward.
River contributions are estimated using two formulations: the BQART formula and the Maffre et al. model. Longshore wave-induced transport is calculated using the Kamphuis formula. An important aspect is that to obtain global-scale results, the authors make very severe simplifications. For example, they use a constant value for the "Te" variable in the BQART formula, use deep-water waves instead of propagating them to the breaking zone, or assume that dams retain 100% of sediment, among other severe simplifications. Another point to highlight is the size of the cells, 50 km, which results in physically impossible coastal cells. The set of simplifications detracts credibility from the results obtained, at least from a quantitative perspective.
Despite the topic's interest and the significant amount of work done by the authors, this reviewer believes that the article does not meet the standards for publication in NHESS in its current state, and requires major revision. Among the aspects to improve/correct are:
Title: The current title, "The contribution of diminishing river sand loads to beach erosion worldwide," does not reflect the paper's content. The reduction in sediment load from a river can be due to multiple causes (agriculture, urbanization, land use change, hydrological regime change, etc.), which the article does not explore. It only evaluates the decrease in riverine sand load due to the presence of dams. The title should be changed and confined to the impact of dams. Additionally, the word "contribution to" suggests that the manuscript provides the percentage of coastal erosion attributable to the reduced river sediment input, which is not the case.
Structure: The manuscript lacks a conclusions section that summarizes the novel contributions of the work. Sections 5 and 6 discuss interesting topics related to river sediment load but the paper needs a conclusions section.
Scientific Significance: The main problem with the manuscript in its current state, according to this reviewer, is that it does not represent a substantial contribution to the understanding or quantification of the coastal erosion problem due to reduced river inputs.
The issues addressed in sections 5 and 6 are interesting for general public dissemination but are widely known to scientists and river/coastal managers. Phrases such as "Our results show that a primary global threat to sandy coasts may arise from fluvial sediment supply deficit" or “Our study strongly suggests that efficient management strategies cannot be limited to just coasts but should consider entire river basins within a more holistic treatment of the problem” or "Our study constitutes a major breakthrough by providing the first evidence that the current trend in sandy coastal evolution is controlled by the local imbalance of sand transport in a predictable way" do not contribute to scientific knowledge and do not require the development of any global mathematical model.
In its current state, the authors fail to achieve the objectives set in the title and abstract because, as mentioned earlier, the reduction in sediment load from a river can be due to multiple causes, and the article only evaluates the impact of dams. Furthermore, it does so assuming overly severe hypotheses for the reduction values to be considered realistic, and the verisimilitude of these values is unknown.
The manuscript should focus on the results of section 4 and, in addition to providing global values, regionalize the results into homogeneous areas where other factors are analogous, and the impact of dams can be isolated for a more realistic comparison.
Citation: https://doi.org/10.5194/nhess-2023-165-RC1 -
AC1: 'Reply on RC', Vincent Regard, 30 Apr 2024
Dear EditorWe have taken note of the criticisms that led to the rejection of the manuscript. Some of these criticisms are well-founded, while others seem to us to be either erroneous or have arisen from misunderstandings. The main criticisms are discussed below.
Firstly, we thank the reviewers for their critical reading. We confess that we did not focus this work properly in time and space. In terms of time, as we are only calculating current or “instantaneous” differences in fluxes, we are not in a position to indicate the characteristic sediment transport time nor response time to a change in river sediment outflux to ocean. We obviously agree with the reviewers that the effect of a reduction in river sediment fluxes will not be immediate hundreds of kilometres away. In space, the 50 km resolution has been strongly criticised, and we are carrying out tests to assess the influence of this resolution. However, even if our model is an 'instantaneous' model, we maintain that the predictions constitute a breakthrough in the understanding of the global dynamics of coastal sediment transport. We note that, to our knowledge, such a global modelling approach has never been published. Of course, such an approach requires simplifications, some of which we have probably not sufficiently justified and discussed. This model is not intended to explain the details of all coastlines, and to reject this approach because it does not explain all local situations (ex California) seems to us to reveal a misunderstanding for which we take responsibility. One of the predictions of this model is the long-distance teleconnection - several thousand kilometres - between a river and the sediment budget on a stretch of coastline. This distance is criticised by the reviewers, and rightly so, because we have not discussed sufficiently the effect of the barriers that our model neglects. However, we have also not sufficiently justified the possibility that such teleconnections exist in natural cases. The data from Garzanti et al. (2014) along Africa, for example, show that this is possible. Of course, we need to improve our model to take account of the response times in these teleconnections in the future. On the other hand, to claim that our conclusions are already well known seems to us to contradict the a priori rejection of the existence of coastal cells several thousand km long. Another major prediction is the concentration of sediments in the inter-tropical zone. This is a testable prediction, consistent with some observations and only a global model, despite its limitations, can address these phenomena on this scale, bringing an explanation. Finally, and this is the minimum for a model, our predictions are confronted with data and we show that a significant proportion of the observations are explained by our very simple model. We believe that this should be taken into account in the evaluation of our study, at least in the same way as the model's shortcomings.
Secondly, our reasoning is based on a general consideration of the balance between inputs to and outputs from the coastal zone. For BQART we took two end-members, and the truth lies somewhere in between. We are aware that coastal erosion or canyons matter and could be taken into account; we were concentrating on demonstrating the importance of taking river inputs into account when exploring the evolution of beaches. Reviewer 1 claims that all this is known, which is not quite true; for example, the transfer of sediment from rivers to beaches is neither as straightforward as the reviewer may be presuming nor so quite well understood. Estuaries can sequester considerable amounts of sand that do not reach adjacent beaches through a variety of tidal and freshwater/seawater interactions as they infill over time, while wave-dominated deltas (overall the smallest category of world deltas, compared, for instance, to river- and tide-dominated deltas), can develop morphosedimentary mechanisms of sand sequestration within the confines of their beaches (flanking spits being a fine example).
Finally, there are many beaches on Earth that do not show a direct link with present river sourcing, including beaches dominated by terrigenous siliciclastic sand, despite the firm belief of a direct link between rivers and beaches. An important point we show is that LST has the potential to redistribute most of the sand that actually succeeds in leaving river mouths onto beaches. Again we also do not want to focus on California, which can give rise to misconceptions. For example, the fact that sediment can end up 1000 km from the mouth from which it originated has been proven (Garzanti et al., 2014). Here we also note, however, that we were wrong to evoke 'littoral cell', without a rigorous definition of this term.
Both reviewers pointed out that the paper is interesting and even important. A few errors on our part as well as some misunderstandings on the part of the reviewers led to its rejection, which is unfortunate because we really think it brings value to science.Vincent Regard, on behalf the coauthors, Apr 2024Reference:
Garzanti, E., Vermeesch, P., Andò, S., Lustrino, M., Padoan, M., and Vezzoli, G., 2014, Ultra-long distance littoral transport of Orange sand and provenance of the Skeleton Coast Erg (Namibia): Marine Geology, v. 357, p. 25–36, doi:10.1016/j.margeo.2014.07.005.Citation: https://doi.org/10.5194/nhess-2023-165-AC1
-
AC1: 'Reply on RC', Vincent Regard, 30 Apr 2024
-
RC2: 'Comment on nhess-2023-165', Anonymous Referee #2, 06 Feb 2024
This paper examines the controls on sandy beaches and particularly (1) the longshore scales of beaches and potential sand exchange, and (2) the role to terrestrial sand supply from rivers to the coast, including the reduction in supply due to dams. The authors adopt a modelling approach where they drive coastal evolution with waves including longshore transport and consider sediment supply to the coast with well-established models and databases on dams. The title is strange as it makes no reference to the first aspect on coastal cells and in many ways this feels like two papers put together. The models that are used have important limitations it is unclear why many assumptions are made, and the model results are not compared to the available field data and do not stand up to scrutiny. Hence, much further work is required to develop useful results and understand their implications.
For example, the length of coastal cells that are defined in the analysis are , very long – thousands of kms. Do these coastal cells actually exist? A large coastal cell is show on the Pacific coast of North America. The coast of California has been well studied and the cells are actually much smaller due to the presence of submarine canyons which provide a one-way route for beach sediment away from the coast to the seabed and ocean flow – this is omitted in the paper’s model and the California case is not considered. Is such submarine canyon control on open coast dimensions a general characteristic of open coasts? So the results in the paper are fine within their limitations – but they define ‘potential coastal cells’ due to longshore transport and much further work is needed to compare this to field data and map the actual sediment cells.
The paper also asserts that disruption of sediment supply to one part of a cell degrades the entire cell. This may be true, but at what timescale? If sediment supply is interrupted or removed it will take time for this loss to propagate along the coast. I hypothesise that within a long cell it may take centuries or more for these effects to be felt everywhere in the cell. However, temporal issues are not mentioned and only spatial linkage is considered in the paper.
This raises a general point that the issue of scale is poorly considered throughout the paper and the relevant scales should be thought about and defined throughout. The importance of this issue has been written about extensively by coastal geomorphologists (for more than a century) and by coastal engineers (for decades) – but in the paper this important conceptual aspect is implicit at best.
A final example query concerns river supply of sediment to the coast. There is an assumption that loss of sand supply is felt immediately by the open coast (another example of a scale assumption). This is certainly true on a steep coast like Catalonia where the sand supply has failed and erosion today is a critical issue. However, on much less steep barrier coasts, such as the US East coast in nearly all locations any river sand is delivered to the lagoons or estuaries well landward of the open coast. Hence under current conditions, the barrier island coast would not “see” the loss of river sand supply in its sediment budget which comprises reworking of the shoreface and the barrier islands themselves. These sediment systems are not coupled at a human timescale and the sediment budget of the barrier islands does not see changes in river supply in any meaningful sense. So the ‘global’ conclusions are not appropriate.
Other criticisms could be raised.
So my assessment is that these examples show that the analysis is flawed in multiple ways and key processes and field evidence have been ignored. Therefore, I must recommend rejection. As a piece of advice when taking this research forward as I assume the authors will do, I would suggest they consider splitting the two aspects – the length of coastal cells and the sediment supply from rivers. Expanding these two contributions into two separate papers might provide a more focussed output.
Below are some more detailed remarks.
- Abstract hard to read and uninformative
- Figure 1 – this is a rather simplified sediment budget – coastal erosion such as cliff or shoreface erosion is a source of sediment, losses due to submarine canyons, etc. Why are some terms considered and others ignored?
- Lines 57-81 discusses a lot of processes – better if it was structured as a table, list or figure.
- Line 115 ‘modern-era’ – define. The treatment of timescale in the paper is weaker than the treatment of space.
- Line 270-271 ‘However, sediment has the potential to travel across the entire sediment cell, which can span thousands of kilometers along the coast.’ – what is the evidence that this is the case. The classical view of sediment cells in California has links over 100-kms – not 1000-kms. In my experience that seems correct over a 100 year timescale– or is this a geological observation over thousands or more of years – and is it supported by evidence – for example provenance analysis of sediment?
- Figure 3 shows the cells – and what I consider an unrealistic length on west coast of North America compared published sediment budgets. Submarine canyons are important and not represented in the global model.
- Line 297-315 – what is the meaning of an average yield per cell – as cells are not average – some have river input and some do not. Please explain.
- Line 360-384 – key point is that we need to consider the sediment budget in addition to sea-level rise. This is a strong point and I think others have said this before – so could cite these papers and reinforce the message.
- Line 429 – Dams ‘and direct sediment mining’ which is already an industry in some locations such as the Mekong delta and may spread as the demand for sand is huge. Direct removal of beach sand is not uncommon in many locations.
Citation: https://doi.org/10.5194/nhess-2023-165-RC2 -
AC2: 'Reply on RC', Vincent Regard, 30 Apr 2024
Dear EditorWe have taken note of the criticisms that led to the rejection of the manuscript. Some of these criticisms are well-founded, while others seem to us to be either erroneous or have arisen from misunderstandings. The main criticisms are discussed below.
Firstly, we thank the reviewers for their critical reading. We confess that we did not focus this work properly in time and space. In terms of time, as we are only calculating current or “instantaneous” differences in fluxes, we are not in a position to indicate the characteristic sediment transport time nor response time to a change in river sediment outflux to ocean. We obviously agree with the reviewers that the effect of a reduction in river sediment fluxes will not be immediate hundreds of kilometres away. In space, the 50 km resolution has been strongly criticised, and we are carrying out tests to assess the influence of this resolution. However, even if our model is an 'instantaneous' model, we maintain that the predictions constitute a breakthrough in the understanding of the global dynamics of coastal sediment transport. We note that, to our knowledge, such a global modelling approach has never been published. Of course, such an approach requires simplifications, some of which we have probably not sufficiently justified and discussed. This model is not intended to explain the details of all coastlines, and to reject this approach because it does not explain all local situations (ex California) seems to us to reveal a misunderstanding for which we take responsibility. One of the predictions of this model is the long-distance teleconnection - several thousand kilometres - between a river and the sediment budget on a stretch of coastline. This distance is criticised by the reviewers, and rightly so, because we have not discussed sufficiently the effect of the barriers that our model neglects. However, we have also not sufficiently justified the possibility that such teleconnections exist in natural cases. The data from Garzanti et al. (2014) along Africa, for example, show that this is possible. Of course, we need to improve our model to take account of the response times in these teleconnections in the future. On the other hand, to claim that our conclusions are already well known seems to us to contradict the a priori rejection of the existence of coastal cells several thousand km long. Another major prediction is the concentration of sediments in the inter-tropical zone. This is a testable prediction, consistent with some observations and only a global model, despite its limitations, can address these phenomena on this scale, bringing an explanation. Finally, and this is the minimum for a model, our predictions are confronted with data and we show that a significant proportion of the observations are explained by our very simple model. We believe that this should be taken into account in the evaluation of our study, at least in the same way as the model's shortcomings.
Secondly, our reasoning is based on a general consideration of the balance between inputs to and outputs from the coastal zone. For BQART we took two end-members, and the truth lies somewhere in between. We are aware that coastal erosion or canyons matter and could be taken into account; we were concentrating on demonstrating the importance of taking river inputs into account when exploring the evolution of beaches. Reviewer 1 claims that all this is known, which is not quite true; for example, the transfer of sediment from rivers to beaches is neither as straightforward as the reviewer may be presuming nor so quite well understood. Estuaries can sequester considerable amounts of sand that do not reach adjacent beaches through a variety of tidal and freshwater/seawater interactions as they infill over time, while wave-dominated deltas (overall the smallest category of world deltas, compared, for instance, to river- and tide-dominated deltas), can develop morphosedimentary mechanisms of sand sequestration within the confines of their beaches (flanking spits being a fine example).
Finally, there are many beaches on Earth that do not show a direct link with present river sourcing, including beaches dominated by terrigenous siliciclastic sand, despite the firm belief of a direct link between rivers and beaches. An important point we show is that LST has the potential to redistribute most of the sand that actually succeeds in leaving river mouths onto beaches. Again we also do not want to focus on California, which can give rise to misconceptions. For example, the fact that sediment can end up 1000 km from the mouth from which it originated has been proven (Garzanti et al., 2014). Here we also note, however, that we were wrong to evoke 'littoral cell', without a rigorous definition of this term.
Both reviewers pointed out that the paper is interesting and even important. A few errors on our part as well as some misunderstandings on the part of the reviewers led to its rejection, which is unfortunate because we really think it brings value to science.Vincent Regard, on behalf the coauthors, Apr 2024Reference:
Garzanti, E., Vermeesch, P., Andò, S., Lustrino, M., Padoan, M., and Vezzoli, G., 2014, Ultra-long distance littoral transport of Orange sand and provenance of the Skeleton Coast Erg (Namibia): Marine Geology, v. 357, p. 25–36, doi:10.1016/j.margeo.2014.07.005.Citation: https://doi.org/10.5194/nhess-2023-165-AC2
Interactive discussion
Status: closed
-
RC1: 'Comment on nhess-2023-165', Anonymous Referee #1, 28 Dec 2023
The manuscript explores the contribution of diminishing sediment supply from rivers to global beach erosion. In particular, the manuscript analyzes the impact of thousands of dams on beach erosion worldwide. The topic is of great interest to NHESS readers; the text is of appropriate length and is easy to read and understand for a broad and diverse readership, including fellow scientists.
To achieve their goal, the authors develop a sediment conservation model that calculates the balance between sediment input from rivers and longitudinal transport along the coast due to wave action. This model is implemented on the world´s coast segmented into alongshore cells of 0.5º (11,161 cells of about 50 km each). The model results in sediment accumulation/erosion patterns in each cell. The authors emphasize that the conversion between these sediment accumulation/erosion patterns and coastal morphological changes (advance/recession) is not straightforward.
River contributions are estimated using two formulations: the BQART formula and the Maffre et al. model. Longshore wave-induced transport is calculated using the Kamphuis formula. An important aspect is that to obtain global-scale results, the authors make very severe simplifications. For example, they use a constant value for the "Te" variable in the BQART formula, use deep-water waves instead of propagating them to the breaking zone, or assume that dams retain 100% of sediment, among other severe simplifications. Another point to highlight is the size of the cells, 50 km, which results in physically impossible coastal cells. The set of simplifications detracts credibility from the results obtained, at least from a quantitative perspective.
Despite the topic's interest and the significant amount of work done by the authors, this reviewer believes that the article does not meet the standards for publication in NHESS in its current state, and requires major revision. Among the aspects to improve/correct are:
Title: The current title, "The contribution of diminishing river sand loads to beach erosion worldwide," does not reflect the paper's content. The reduction in sediment load from a river can be due to multiple causes (agriculture, urbanization, land use change, hydrological regime change, etc.), which the article does not explore. It only evaluates the decrease in riverine sand load due to the presence of dams. The title should be changed and confined to the impact of dams. Additionally, the word "contribution to" suggests that the manuscript provides the percentage of coastal erosion attributable to the reduced river sediment input, which is not the case.
Structure: The manuscript lacks a conclusions section that summarizes the novel contributions of the work. Sections 5 and 6 discuss interesting topics related to river sediment load but the paper needs a conclusions section.
Scientific Significance: The main problem with the manuscript in its current state, according to this reviewer, is that it does not represent a substantial contribution to the understanding or quantification of the coastal erosion problem due to reduced river inputs.
The issues addressed in sections 5 and 6 are interesting for general public dissemination but are widely known to scientists and river/coastal managers. Phrases such as "Our results show that a primary global threat to sandy coasts may arise from fluvial sediment supply deficit" or “Our study strongly suggests that efficient management strategies cannot be limited to just coasts but should consider entire river basins within a more holistic treatment of the problem” or "Our study constitutes a major breakthrough by providing the first evidence that the current trend in sandy coastal evolution is controlled by the local imbalance of sand transport in a predictable way" do not contribute to scientific knowledge and do not require the development of any global mathematical model.
In its current state, the authors fail to achieve the objectives set in the title and abstract because, as mentioned earlier, the reduction in sediment load from a river can be due to multiple causes, and the article only evaluates the impact of dams. Furthermore, it does so assuming overly severe hypotheses for the reduction values to be considered realistic, and the verisimilitude of these values is unknown.
The manuscript should focus on the results of section 4 and, in addition to providing global values, regionalize the results into homogeneous areas where other factors are analogous, and the impact of dams can be isolated for a more realistic comparison.
Citation: https://doi.org/10.5194/nhess-2023-165-RC1 -
AC1: 'Reply on RC', Vincent Regard, 30 Apr 2024
Dear EditorWe have taken note of the criticisms that led to the rejection of the manuscript. Some of these criticisms are well-founded, while others seem to us to be either erroneous or have arisen from misunderstandings. The main criticisms are discussed below.
Firstly, we thank the reviewers for their critical reading. We confess that we did not focus this work properly in time and space. In terms of time, as we are only calculating current or “instantaneous” differences in fluxes, we are not in a position to indicate the characteristic sediment transport time nor response time to a change in river sediment outflux to ocean. We obviously agree with the reviewers that the effect of a reduction in river sediment fluxes will not be immediate hundreds of kilometres away. In space, the 50 km resolution has been strongly criticised, and we are carrying out tests to assess the influence of this resolution. However, even if our model is an 'instantaneous' model, we maintain that the predictions constitute a breakthrough in the understanding of the global dynamics of coastal sediment transport. We note that, to our knowledge, such a global modelling approach has never been published. Of course, such an approach requires simplifications, some of which we have probably not sufficiently justified and discussed. This model is not intended to explain the details of all coastlines, and to reject this approach because it does not explain all local situations (ex California) seems to us to reveal a misunderstanding for which we take responsibility. One of the predictions of this model is the long-distance teleconnection - several thousand kilometres - between a river and the sediment budget on a stretch of coastline. This distance is criticised by the reviewers, and rightly so, because we have not discussed sufficiently the effect of the barriers that our model neglects. However, we have also not sufficiently justified the possibility that such teleconnections exist in natural cases. The data from Garzanti et al. (2014) along Africa, for example, show that this is possible. Of course, we need to improve our model to take account of the response times in these teleconnections in the future. On the other hand, to claim that our conclusions are already well known seems to us to contradict the a priori rejection of the existence of coastal cells several thousand km long. Another major prediction is the concentration of sediments in the inter-tropical zone. This is a testable prediction, consistent with some observations and only a global model, despite its limitations, can address these phenomena on this scale, bringing an explanation. Finally, and this is the minimum for a model, our predictions are confronted with data and we show that a significant proportion of the observations are explained by our very simple model. We believe that this should be taken into account in the evaluation of our study, at least in the same way as the model's shortcomings.
Secondly, our reasoning is based on a general consideration of the balance between inputs to and outputs from the coastal zone. For BQART we took two end-members, and the truth lies somewhere in between. We are aware that coastal erosion or canyons matter and could be taken into account; we were concentrating on demonstrating the importance of taking river inputs into account when exploring the evolution of beaches. Reviewer 1 claims that all this is known, which is not quite true; for example, the transfer of sediment from rivers to beaches is neither as straightforward as the reviewer may be presuming nor so quite well understood. Estuaries can sequester considerable amounts of sand that do not reach adjacent beaches through a variety of tidal and freshwater/seawater interactions as they infill over time, while wave-dominated deltas (overall the smallest category of world deltas, compared, for instance, to river- and tide-dominated deltas), can develop morphosedimentary mechanisms of sand sequestration within the confines of their beaches (flanking spits being a fine example).
Finally, there are many beaches on Earth that do not show a direct link with present river sourcing, including beaches dominated by terrigenous siliciclastic sand, despite the firm belief of a direct link between rivers and beaches. An important point we show is that LST has the potential to redistribute most of the sand that actually succeeds in leaving river mouths onto beaches. Again we also do not want to focus on California, which can give rise to misconceptions. For example, the fact that sediment can end up 1000 km from the mouth from which it originated has been proven (Garzanti et al., 2014). Here we also note, however, that we were wrong to evoke 'littoral cell', without a rigorous definition of this term.
Both reviewers pointed out that the paper is interesting and even important. A few errors on our part as well as some misunderstandings on the part of the reviewers led to its rejection, which is unfortunate because we really think it brings value to science.Vincent Regard, on behalf the coauthors, Apr 2024Reference:
Garzanti, E., Vermeesch, P., Andò, S., Lustrino, M., Padoan, M., and Vezzoli, G., 2014, Ultra-long distance littoral transport of Orange sand and provenance of the Skeleton Coast Erg (Namibia): Marine Geology, v. 357, p. 25–36, doi:10.1016/j.margeo.2014.07.005.Citation: https://doi.org/10.5194/nhess-2023-165-AC1
-
AC1: 'Reply on RC', Vincent Regard, 30 Apr 2024
-
RC2: 'Comment on nhess-2023-165', Anonymous Referee #2, 06 Feb 2024
This paper examines the controls on sandy beaches and particularly (1) the longshore scales of beaches and potential sand exchange, and (2) the role to terrestrial sand supply from rivers to the coast, including the reduction in supply due to dams. The authors adopt a modelling approach where they drive coastal evolution with waves including longshore transport and consider sediment supply to the coast with well-established models and databases on dams. The title is strange as it makes no reference to the first aspect on coastal cells and in many ways this feels like two papers put together. The models that are used have important limitations it is unclear why many assumptions are made, and the model results are not compared to the available field data and do not stand up to scrutiny. Hence, much further work is required to develop useful results and understand their implications.
For example, the length of coastal cells that are defined in the analysis are , very long – thousands of kms. Do these coastal cells actually exist? A large coastal cell is show on the Pacific coast of North America. The coast of California has been well studied and the cells are actually much smaller due to the presence of submarine canyons which provide a one-way route for beach sediment away from the coast to the seabed and ocean flow – this is omitted in the paper’s model and the California case is not considered. Is such submarine canyon control on open coast dimensions a general characteristic of open coasts? So the results in the paper are fine within their limitations – but they define ‘potential coastal cells’ due to longshore transport and much further work is needed to compare this to field data and map the actual sediment cells.
The paper also asserts that disruption of sediment supply to one part of a cell degrades the entire cell. This may be true, but at what timescale? If sediment supply is interrupted or removed it will take time for this loss to propagate along the coast. I hypothesise that within a long cell it may take centuries or more for these effects to be felt everywhere in the cell. However, temporal issues are not mentioned and only spatial linkage is considered in the paper.
This raises a general point that the issue of scale is poorly considered throughout the paper and the relevant scales should be thought about and defined throughout. The importance of this issue has been written about extensively by coastal geomorphologists (for more than a century) and by coastal engineers (for decades) – but in the paper this important conceptual aspect is implicit at best.
A final example query concerns river supply of sediment to the coast. There is an assumption that loss of sand supply is felt immediately by the open coast (another example of a scale assumption). This is certainly true on a steep coast like Catalonia where the sand supply has failed and erosion today is a critical issue. However, on much less steep barrier coasts, such as the US East coast in nearly all locations any river sand is delivered to the lagoons or estuaries well landward of the open coast. Hence under current conditions, the barrier island coast would not “see” the loss of river sand supply in its sediment budget which comprises reworking of the shoreface and the barrier islands themselves. These sediment systems are not coupled at a human timescale and the sediment budget of the barrier islands does not see changes in river supply in any meaningful sense. So the ‘global’ conclusions are not appropriate.
Other criticisms could be raised.
So my assessment is that these examples show that the analysis is flawed in multiple ways and key processes and field evidence have been ignored. Therefore, I must recommend rejection. As a piece of advice when taking this research forward as I assume the authors will do, I would suggest they consider splitting the two aspects – the length of coastal cells and the sediment supply from rivers. Expanding these two contributions into two separate papers might provide a more focussed output.
Below are some more detailed remarks.
- Abstract hard to read and uninformative
- Figure 1 – this is a rather simplified sediment budget – coastal erosion such as cliff or shoreface erosion is a source of sediment, losses due to submarine canyons, etc. Why are some terms considered and others ignored?
- Lines 57-81 discusses a lot of processes – better if it was structured as a table, list or figure.
- Line 115 ‘modern-era’ – define. The treatment of timescale in the paper is weaker than the treatment of space.
- Line 270-271 ‘However, sediment has the potential to travel across the entire sediment cell, which can span thousands of kilometers along the coast.’ – what is the evidence that this is the case. The classical view of sediment cells in California has links over 100-kms – not 1000-kms. In my experience that seems correct over a 100 year timescale– or is this a geological observation over thousands or more of years – and is it supported by evidence – for example provenance analysis of sediment?
- Figure 3 shows the cells – and what I consider an unrealistic length on west coast of North America compared published sediment budgets. Submarine canyons are important and not represented in the global model.
- Line 297-315 – what is the meaning of an average yield per cell – as cells are not average – some have river input and some do not. Please explain.
- Line 360-384 – key point is that we need to consider the sediment budget in addition to sea-level rise. This is a strong point and I think others have said this before – so could cite these papers and reinforce the message.
- Line 429 – Dams ‘and direct sediment mining’ which is already an industry in some locations such as the Mekong delta and may spread as the demand for sand is huge. Direct removal of beach sand is not uncommon in many locations.
Citation: https://doi.org/10.5194/nhess-2023-165-RC2 -
AC2: 'Reply on RC', Vincent Regard, 30 Apr 2024
Dear EditorWe have taken note of the criticisms that led to the rejection of the manuscript. Some of these criticisms are well-founded, while others seem to us to be either erroneous or have arisen from misunderstandings. The main criticisms are discussed below.
Firstly, we thank the reviewers for their critical reading. We confess that we did not focus this work properly in time and space. In terms of time, as we are only calculating current or “instantaneous” differences in fluxes, we are not in a position to indicate the characteristic sediment transport time nor response time to a change in river sediment outflux to ocean. We obviously agree with the reviewers that the effect of a reduction in river sediment fluxes will not be immediate hundreds of kilometres away. In space, the 50 km resolution has been strongly criticised, and we are carrying out tests to assess the influence of this resolution. However, even if our model is an 'instantaneous' model, we maintain that the predictions constitute a breakthrough in the understanding of the global dynamics of coastal sediment transport. We note that, to our knowledge, such a global modelling approach has never been published. Of course, such an approach requires simplifications, some of which we have probably not sufficiently justified and discussed. This model is not intended to explain the details of all coastlines, and to reject this approach because it does not explain all local situations (ex California) seems to us to reveal a misunderstanding for which we take responsibility. One of the predictions of this model is the long-distance teleconnection - several thousand kilometres - between a river and the sediment budget on a stretch of coastline. This distance is criticised by the reviewers, and rightly so, because we have not discussed sufficiently the effect of the barriers that our model neglects. However, we have also not sufficiently justified the possibility that such teleconnections exist in natural cases. The data from Garzanti et al. (2014) along Africa, for example, show that this is possible. Of course, we need to improve our model to take account of the response times in these teleconnections in the future. On the other hand, to claim that our conclusions are already well known seems to us to contradict the a priori rejection of the existence of coastal cells several thousand km long. Another major prediction is the concentration of sediments in the inter-tropical zone. This is a testable prediction, consistent with some observations and only a global model, despite its limitations, can address these phenomena on this scale, bringing an explanation. Finally, and this is the minimum for a model, our predictions are confronted with data and we show that a significant proportion of the observations are explained by our very simple model. We believe that this should be taken into account in the evaluation of our study, at least in the same way as the model's shortcomings.
Secondly, our reasoning is based on a general consideration of the balance between inputs to and outputs from the coastal zone. For BQART we took two end-members, and the truth lies somewhere in between. We are aware that coastal erosion or canyons matter and could be taken into account; we were concentrating on demonstrating the importance of taking river inputs into account when exploring the evolution of beaches. Reviewer 1 claims that all this is known, which is not quite true; for example, the transfer of sediment from rivers to beaches is neither as straightforward as the reviewer may be presuming nor so quite well understood. Estuaries can sequester considerable amounts of sand that do not reach adjacent beaches through a variety of tidal and freshwater/seawater interactions as they infill over time, while wave-dominated deltas (overall the smallest category of world deltas, compared, for instance, to river- and tide-dominated deltas), can develop morphosedimentary mechanisms of sand sequestration within the confines of their beaches (flanking spits being a fine example).
Finally, there are many beaches on Earth that do not show a direct link with present river sourcing, including beaches dominated by terrigenous siliciclastic sand, despite the firm belief of a direct link between rivers and beaches. An important point we show is that LST has the potential to redistribute most of the sand that actually succeeds in leaving river mouths onto beaches. Again we also do not want to focus on California, which can give rise to misconceptions. For example, the fact that sediment can end up 1000 km from the mouth from which it originated has been proven (Garzanti et al., 2014). Here we also note, however, that we were wrong to evoke 'littoral cell', without a rigorous definition of this term.
Both reviewers pointed out that the paper is interesting and even important. A few errors on our part as well as some misunderstandings on the part of the reviewers led to its rejection, which is unfortunate because we really think it brings value to science.Vincent Regard, on behalf the coauthors, Apr 2024Reference:
Garzanti, E., Vermeesch, P., Andò, S., Lustrino, M., Padoan, M., and Vezzoli, G., 2014, Ultra-long distance littoral transport of Orange sand and provenance of the Skeleton Coast Erg (Namibia): Marine Geology, v. 357, p. 25–36, doi:10.1016/j.margeo.2014.07.005.Citation: https://doi.org/10.5194/nhess-2023-165-AC2
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 424 | 122 | 41 | 587 | 47 | 37 | 30 |
- HTML: 424
- PDF: 122
- XML: 41
- Total: 587
- Supplement: 47
- BibTeX: 37
- EndNote: 30
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
Cited
1 citations as recorded by crossref.
Marcan Graffin
Sébastien Carretier
Edward Anthony
Roshanka Ranasinghe
Pierre Maffre
This preprint has been withdrawn.
- Preprint
(1975 KB) - Metadata XML
-
Supplement
(1241 KB) - BibTeX
- EndNote