Research of Impacts of the 2018 Hokkaido Eastern Iburi Earthquake on Sediment Transport in the Atsuma River Basin Using the SWAT Model
Round 1
Reviewer 1 Report
The paper entitled: “Research of impacts of 2018 Hokkaido Eastern Iburi Earthquake on sediment transport in the Atsuma River basin using SWAT model” by Yuechao Chen, Makoto Nakatsugawa and Hiroki Ohashi is reporting an interesting topic that is related to the dynamics of discharge and sediment dynamics after an strong triggering event. However, the paper needs profound review of the English wording and grammar. Moreover, the SWAT sediment module is based on the modified universal soil loss equation and thus, seems to me the wrong model for this study. As reported by the authors main erosion processes are related to debris flows and landslides. However, MUSLE is not simulating these processes. The MUSLE was developed to assess sheet or rill interrill erosion. Furthermore, the authors compare the MULSE results to a time series of NTU measurements. According to the study area description the turbidity due to suspended sediments might be due to debris flows and landslides that are not considered in the MUSLE model. Consequently, it remains unclear and questionable to compare the sediment discharge in SWAT with the NTU because you might get the right results for the wrong reasons. The authors should discuss at the beginning the erosion processes active in the catchment. What changed after the earthquake? Do the sediments washed into the river belong to the mass movements or to rill-interrill erosion?
In the diagrams also the precipitation should be reported otherwise the river discharges and sediment flows can not be interpreted in a proper way. Finally, it would be also interesting to know if there have been registered some precipitation events before the earthquake and if the soils were already saturated?
For further comments and corrections see the word text attached.
Comments for author File: 
Comments.docx
Author Response
Response to Reviewer 2 Comments
Point 1: The paper entitled: “Research of impacts of 2018 Hokkaido Eastern Iburi Earthquake on sediment transport in the Atsuma River basin using SWAT model” by Yuechao Chen, Makoto Nakatsugawa and Hiroki Ohashi is reporting an interesting topic that is related to the dynamics of discharge and sediment dynamics after an strong triggering event. However, the paper needs profound review of the English wording and grammar. 

Response 1:
First of all, thank you very much for your constructive comments and hard work. Regarding the English wording and grammar of this manuscript, firstly, based on your valuable revision comments, we carefully revised it verbatim; secondly, with the help of professional English editors, we carefully, conscientiously and deeply modified the full text again. We believe that the quality of English writing has been improved greatly. Please review it again.
Point 2: Moreover, the SWAT sediment module is based on the modified universal soil loss equation and thus, seems to me the wrong model for this study. As reported by the authors main erosion processes are related to debris flows and landslides. However, MUSLE is not simulating these processes. The MUSLE was developed to assess sheet or rill interrill erosion. Furthermore, the authors compare the MULSE results to a time series of NTU measurements. According to the study area description the turbidity due to suspended sediments might be due to debris flows and landslides that are not considered in the MUSLE model. Consequently, it remains unclear and questionable to compare the sediment discharge in SWAT with the NTU because you might get the right results for the wrong reasons. The authors should discuss at the beginning the erosion processes active in the catchment. What changed after the earthquake? Do the sediments washed into the river belong to the mass movements or to rill-interrill erosion?
Response 2:
Thank you very much for your comments. The impacts of landslides on the sediment in river basin are mainly divided into two aspects. On the one hand, the landslide changes the land use of the basin, making the area that was woodland, grassland or cultivated originally transform into bare land. When it is raining, the bare land will produce more sediment and the sediment will flows into the river channel. On the other hand, after the earthquake, the sediment, debris generated by the landslide flow directly into the river channel in a short time, increasing the amount of river sediment. The changes of the land use can be represented by modifying the sediment parameters of SWAT, but the SWAT model can't cover the sediment caused by landslide flowing into the river directly. This is the limitation of the SWAT model. This study uses the SWAT model to study the changes in sediment transport in the Atsuma River basin before and after the earthquake. In addition to making full use of the SWAT model’s advantages in simulating the runoff and sediment transport process in the basin, the simulated sediment transport is compared with the observation turbidity. The turbidity data is compared with calculation sediment transport to help to estimate the amount of sediment transport in the Atsuma River basin after the earthquake, and facilitate the analysis of the amount of sediment transport per unit of precipitation before and after the earthquake.
The change of turbidity has important reference significance for judging the change of sediment transport. Both the land use changes and the sediment which flows into river channel directly would influence the change of turbidity. The observation turbidity data can directly reflect the characteristics of sediment change in each stage of the basin. In view of the fact that there is no measured sediment transport data in the Atsuma River basin, this manuscript uses the observation turbidity to correct and evaluate the sediment simulation process of the SWAT model, and to make up for the fact that the SWAT model can not cover the amount of sediment that flows directly into the river channel with the landslide. In order to analyze the impact of the earthquake on the basin, the landslide information is more clearly included in Figure 3, and the slope of the Atsuma River basin is added in Figure 4. The updated content were in Line 113.
According to our analysis and observation, most of the sediments washed into the river belong to rill-interrill erosion.
Point 3: In the diagrams also the precipitation should be reported otherwise the river discharges and sediment flows can not be interpreted in a proper way. Finally, it would be also interesting to know if there have been registered some precipitation events before the earthquake and if the soils were already saturated?
Response 3:
Thank you very much for your comments. Based on your suggestions, we have added precipitation data to Figure 6, Figure 7 and Figure 8 to better explain the changes in river flow and sediment transport.
For the precipitation in the Atsuma River basin before the earthquake, firstly, the Table 4 in manuscript shows precipitation in the upstream of the Atsuma bridge station in the Atsuma River basin from 2015 to 2019. By analyzing the precipitation in the past five years, it can be found that the inter-annual precipitation was fluctuating. The precipitation in 2016 and 2018 was more than that in 2015, 2017 and 2019. The precipitation in 2019 was the least in five years. In addition, through analysis, it can be found that from 2015 to 2019, the proportion of precipitation transformed into runoff was increasing. From this, it can be roughly judged that the inter-annual soil water content is a downward trend. Secondly, by consulting the data, the precipitation before the earthquake, there were 5 major precipitation events from April 2018 to the earthquake. The nearest precipitation event to the earthquake occurred on August 13-18, 2018, 18 days before the earthquake. It can be judged that the soil water content before the earthquake in the Atsuma River basin was relatively high, but not saturated.
The updated content was in Line 272.
Point 4: Line 141: In each HRU the number of research factors?? Is calculated separately, and the total amount is calculated finally by confluence calculation.(not clear)
Response 4:
Thank you very much for your comments. As a typical distributed hydrological model, the SWAT model can simulate various hydrological processes in the basin with high precision. The main reason is that based on various geographic and meteorological data of the basin, the SWAT model can firstly divide the basin into several sub-basins with basically same climatic conditions, and then each sub-basin is divided into hydrological response units (HRU) with the same land type, soil conditions, topography and management methods. Then, when simulating the transport process of runoff, sediment, nutrients, pollutants, etc., the SWAT model first calculate and summarize in the hydrological response unit(HRU), then summarize in the sub-basin, and finally summarize at the basin scale, which greatly improves the accuracy of the simulation and also provides convenience to researchers to analyze the various hydrological process at any hydrological response unit(HRU) or sub-basin in the basin.
Point 5: Line 143: In order to predict the process of runoff and sediment, nutrients and pollutants migration reasonably, various changed process simulated by SWAT???
Response 5:
Thank you very much for your comments. The SWAT model can simulate the transport process of runoff, sediment, nutrients and pollutants in the basin. This study only uses the SWAT model to simulate the process of runoff and sediment transport in the Atsuma River basin.
Runoff plays an important role as a carrier and energy source for the transport of sediment, nutrients and pollutants. Therefore, in order to accurately simulate the transport process of sediment, nutrients and pollutants, the SWAT model should firstly accurately simulate the actual water cycle process in the basin. So, when various processes are simulated, the principle of water balance should be followed.
Point 6: Line 219: next, using the corrected parameters to reproduce the runoff process from 2018 to 2019 by SWAT model. Verify the applicability and accuracy of SWAT model to runoff simulation in the Atsuma River basin.??? Not clear!
Response 6:
Thank you very much for your comments. When using SWAT model to simulate process of runoff, the model needs to be corrected first. This study sets the calibration period from 2015 to 2017. The 2015_2017 observation meteorological data in the Atsuma River basin were input into the SWAT model to simulate the runoff process of the Atsuma River basin from 2015 to 2017. And then, the 2015_2017 observation discharge data were compared with the calculation discharge results, the SWAT-CUP software was used to correct the runoff parameters. Then, using the corrected parameters and the 2018_2019 observation meteorological data in the Atsuma River basin were inputted into the SWAT model to simulate the Atsuma River basin from 2018 to 2019. In the end, compared the calculation discharge results with 2018_2019 observation discharge data to verify the applicability and accuracy of runoff simulation of the SWAT model in the Atsuma River basin.
The updated content was in Line 219.
Point 7: Line 222: Reproduce the sediment transport process from 2018 to 2019 supposing there is no earthquake in 2018. Then, analyze changes of turbidity before and after the earthquake, study the correlation and relationship changes between sediment transport and turbidity before and after the earthquake, and prove that the sediment transport process has been changed after 2018 Hokkaido Eastern Ibur Earthquake….not clear ??
Response 7:
Thank you very much for your comments. On the basis that the data and parameters are not changed, if there is no earthquake, or the occurrence of the earthquake has no impact on the sediment production and transport in the basin, the relationship and correlation before and after the earthquake between turbidity and sediment transport will not change. By the analyzing Figure 7, it can be observed that the correlation and relationship between sediment transport and turbidity before and after the earthquake have change greatly, which proves that after the earthquake, the process of sediment transport in the Atsuma River basin has been changed.
The updated content was in Line 222.
Finally, we would like to express our heartfelt thanks again to you for your constructive comments and hard work.
Author Response File: Author Response.docx
Reviewer 2 Report
Line 32: In recent years, earthquakes appeared frequently, earthquakes and their secondary disasters have plagued human for long time [5,6].
If this sentence is related to your investigation area please mention the data sources?
Line 37: Suggestion: To improve the understanding of the impacts of earthquakes on sediment transport processes in a river basin, the use of hydrological modeling is investigated.
Figure 1: The legend within this figure should not contain the source and web address. Could you please write this in the caption? The coordinates are missing.
Line 83: Please include the depth of the main eartquake in Hokkaido and the orientation and moving direction of the faults in the subsurface. Surely, there were after-shock earthquakes after the main event? This is of importance for the understanding of the occurrence of slope failure.
Line 92: Figure 2: Please add coordinates and a scale bar into this Figure.
Figure 3 is too vague. It should contain the information of landslides more precisely. It would help to identify those rivers prone to higher sediment input due to mass movements along the slopes.
A slope degree map would support the better visualization of steep slopes. The description of the size and volume of the landslides and their composition is missing. Are there slope failures with block-gliding containing more coarse-grained material or more earthflow sites prevailing? This differentiation of the types of mass movements is important for the estimation of the intensity of sediment flow, its volume and, finally, for the dynamics of accumulation processes.
The description of the geologic background, especially the lithologic units, is missing. The weathering conditions are very different whether there are volcanic rocks or loose sediments, granitic rocks, sandstones, etc. . The sediment and soil development depends on the lithologic properties.
Line 153: Could you please clarify in Figure 4 a the values? Meter?
The results of the SWAT-model are very useful, however, cover not all parts of the knowledge required for the estimation of sedimentation and accumulation processes. The results should be embedded and integrated into geophysical and geological investigations. Orographic effects on preciptation amount distribution can cause local differences in the dynamic of sedimentation processes in the catchment areas.
The varying saisonal effects should be considered more. A stronger earthquake during a period with less precipitation will probably cause less landslides and, thus, less sedimention.
Lines 320-325: For sediment transport, since there is no observation sediment transport data in the Atsuma River basin, so based on the high-precision runoff simulation, and assuming that no earthquake occurred, simulation of the sediment transport process in the Atsuma River basin from 2018 to 2019 2 was carried out. By analyzing the change characteristics of turbidity before and after the earthquake and the correlation between turbidity and sediment transport, it can be judged that the sediment transport in the Atsuma River basin has been increased greatly after 2018 Hokkaido Eastern Iburi Earthquake.
Why did you not analyze the earthquake data of the various seismological stations to be sure? What happens when earthquakes happened during your study period before the M7 event? What would this mean for your results, then?
Line 271: …….a large amount of soil and sand enter into the river,
How do you know this? Did you make investigation in the field about the size of the sediments? What about the different particle sizes and size distribution? Composition?
Line 287: The landslide debris directly entered the river channel which produced more sediment load. ?
This manuscript presents interesting insights and relationships between a stronger earthquake event and its impact on sediment flow on a quantitative data base. The SWAT model can simulate runoff processes in the Atsuma River basin with a relatively high accuracy. It is recommended, however, to discuss and mention the limits and critical aspects of this model in the discussion.
Comments for author File: 
Comments.pdf
Author Response
Response to Reviewer 3 Comments
Point 1: Line 32: In recent years, earthquakes appeared frequently, earthquakes and their secondary disasters have plagued human for long time [5,6]. If this sentence is related to your investigation area please mention the data sources?
Response 1:
First of all, thank you very much for your constructive comments and hard work. Your comments has played a vital role in improving the quality of the paper.
From 2017 to 2019, a total of 332 earthquakes with intensity of 6 or more occurred globally during the three-year period, including 35 earthquakes with intensity of 7 or more and 2 earthquakes with intensity of 8 or more. Line 32 has been updated to:
In recent years, earthquakes appeared frequently. From 2017 to 2019, a total of 332 earthquakes with intensity of 6 or more occurred globally, including 35 earthquakes with intensity of 7 or more and 2 earthquakes with intensity of 8 or more.
Point 2: Line 37: Suggestion: To improve the understanding of the impacts of earthquakes on sediment transport processes in a river basin, the use of hydrological modeling is investigated.
Response 2:
Thank you very much for your comments. According to your suggestion, Line 37 has been revised to: To improve the understanding of the impacts of earthquakes on sediment transport processes in a river basin, the use of hydrological modeling is investigated.
This can better express the reasons for the using of hydrological models in this study.
Point 3: Figure 1: The legend within this figure should not contain the source and web address. Could you please write this in the caption? The coordinates are missing.
Response 3:
Thank you for your comments. According to your suggestion, the source and URL in the legend of Figure 1 have been removed, the source has been marked in the title, and the coordinates have been added to the figure. Please see updated Figure 1.
Point 4: Line 83: Please include the depth of the main earthquake in Hokkaido and the orientation and moving direction of the faults in the subsurface. Surely, there were after-shock earthquakes after the main event? This is of importance for the understanding of the occurrence of slope failure.
Response 4:
Thank you very much for your comments. According to your suggestion, the seismic depth, faults, aftershocks and other related content have been added to Line 101 and Line 109. The content added are as follows:
Line101: The epicenter was at 42.7°N, 142°E, depth was 37.0 km;
Line 109: Uplift of up to~7 cm is predominantly distributed in the source region, and eastward movement of up to~4 cm is widely observed on the eastern side of the source region.The fault plane extends in a north–south direction and is dipping to the east with a dip angle of 74° and the fault top is positioned at around 15 km in depth.
Point 5: Line 92: Figure 2: Please add coordinates and a scale bar into this Figure.
Response 5:
Thank you for your comments. According to your suggestion, coordinates and scale have been added to Figure 2. Please see updated Figure 2.
Point 6: Figure 3 is too vague. It should contain the information of landslides more precisely. It would help to identify those rivers prone to higher sediment input due to mass movements along the slopes. A slope degree map would support the better visualization of steep slopes. The description of the size and volume of the landslides and their composition is missing. Are there slope failures with block-gliding containing more coarse-grained material or more earth flow sites prevailing? This differentiation of the types of mass movements is important for the estimation of the intensity of sediment flow, its volume and, finally, for the dynamics of accumulation processes.
Response 6:
Thank you very much for your comments. According to your suggestion, firstly, the Figure 3 has been replaced, and the new figure contains the landslide information more clearly and accurately. Secondly, the slope map of the Atsuma River basin has been added to Figure 4. Please see updated Figures 3 and 4. Information about landslides has been added to Line 121. The content added are as follows:
Line121: The 2018 Hokkaido Eastern Iburi Earthquake caused a large number of shallow landslides, and several large-scale deep-seated landslides involving basement rocks such as shale and mudstone of Miocene, these landslides occurred over approximately 400 km2 hilly areas 200–400 m in elevation. The total landslides area is 43.8 km2 which including 33.1 km2 in the Atsuma River basin; the total accumulation area is 11.8 km2 which including 9.6 km2 in the Atsuma River basin, and the total volume of landslides is 30 million m3.
Point 7: The description of the geologic background, especially the lithologic units, is missing. The weathering conditions are very different whether there are volcanic rocks or loose sediments, granitic rocks, sandstones, etc. . The sediment and soil development depends on the lithologic properties.
Response 7:
Thank you very much for your comments. According to your suggestion, the geology and lithology of the area where the Atsuma River basin has been added to Line83, and the geology and lithology of the area around the epicenter has been added to Line105. The content added are as follows:
Line83: The basement complex of area where Atsuma River basin is located consists mainly of sedimentary rocks of the Neogene tertiary system: the Kawabata Formation and Fureoi Formation (alternate layers of sandstone and mudstone, sandstone, and conglomerate) and the Karumai Formation (mainly diatomaceous siltstone, sandstone, and conglomerate). At the western edge of this area, the Moebetsu Formation (diatomaceous siltstone), which is also of the Neocene tertiary system, or the sedimentary materials of the Middle Pleistocene of the Quaternary system (a sand gravel layer) forms the basement complex.
Line105: The basement rock around the epicenter mainly consists of Neogene sedimentary rocks that are conglomerate, sandstone and mudstone of Fureoi formation of Middle to Late Miocene, sandstone, hard shale and mudstone of Karumai formation of Late Miocene, and conglomerate, sandstone, hard shale and siltstone of Moebetsu (Nina) formation of Late Miocene to Early Pliocene.
Point 8: Line 153: Could you please clarify in Figure 4 a the values? Meter?
Response 8:
Thank you very much for your comments. The unit of the value in Figure 4 a is meter, which has been modified in the legend of Figure 4. The unit meter has been added. Please see updated Figure 4 .
Point 9: The results of the SWAT-model are very useful, however, cover not all parts of the knowledge required for the estimation of sedimentation and accumulation processes. The results should be embedded and integrated into geophysical and geological investigations. Orographic effects on preciptation amount distribution can cause local differences in the dynamic of sedimentation processes in the catchment areas. The varying saisonal effects should be considered more. A stronger earthquake during a period with less precipitation will probably cause less landslides and, thus, less sedimention.
Response 9:
Thank you very much for your comments, and thank you for thinking that the results of the SWAT model are very useful. I agree that the SWAT model can’t cover all parts of the knowledge required for the estimation of sedimentation and accumulation processes, and that the influence of topography on the distribution of precipitation will cause local differences in the dynamics of the sedimentation process in the catchment area. Therefore, during the period of this research, we conducted investigation in the Atsuma River basin for twice. The survey points covered the upper, middle and lower reaches of the basin to better understand the impact of earthquakes on land use in the Atsuma River basin and the actual situation of landslide points. In addition, the observation turbidity data is used in this study. The turbidity and sediment transport are highly correlated. The change of turbidity has important reference significance for judging the change of sediment transport. The change of turbidity can be influenced by the land use changes which produce more sediment and the impact of sediment which flows directly into the river with the landslide at the same time. The observation turbidity can directly reflect the characteristics of sediment change in each stage of the basin. In view of the fact that there is no observation sediment transport data in the Atsuma River basin, this study uses the observation turbidity data to correct and evaluate the sediment simulation process of the SWAT model, try to make up for the inability that the SWAT model can't cover the amount of sediment that flows directly into the river with the landslide.
For precipitation, this study uses radar precipitation data. The radar stations are distributed in a square of 1 km, and the radar stations are sufficient and uniform. While inputting radar rainfall data, the latitude, longitude and elevation of all radar stations are input, considering the influence of orographic on precipitation.
The impacts of landslides on the sediment in river basin are mainly divided into two aspects. On the one hand, the landslide changes the land use of the basin, making the area that was woodland, grassland or cultivated originally transform into bare land. When it is raining, the bare land will produce more sediment and the sediment will flow into the river channel. On the other hand, the sediment, debris generated by the landslide flows directly into the river channel after the earthquake in a short time, increasing the amount of river sediment. The changes of the land use can be represented by modifying the sediment parameters of SWAT, but the SWAT model can't cover the sediment caused by landslide flowing into the river directly. In order to solve this problem, and there is no observation sediment transport data in the Atsuma River basin, this study refers to the observation turbidity to modify the sediment simulation process of the SWAT model. One is because the change of turbidity has important reference significance for judging the change of sediment transport, and the other is because the observation turbidity can intuitively and accurately reflect the characteristics of sediment change in each stage of the basin, so as to make up for the SWAT model as much as possible. The sediment transport in the Atsuma River basin after the earthquake can be estimated numerically.
Point 10: Lines 320-325: For sediment transport, since there is no observation sediment transport data in the Atsuma River basin, so based on the high-precision runoff simulation, and assuming that no earthquake occurred, simulation of the sediment transport process in the Atsuma River basin from 2018 to 2019 2 was carried out. By analyzing the change characteristics of turbidity before and after the earthquake and the correlation between turbidity and sediment transport, it can be judged that the sediment transport in the Atsuma River basin has been increased greatly after 2018 Hokkaido Eastern Iburi Earthquake.
Why did you not analyze the earthquake data of the various seismological stations to be sure? What happens when earthquakes happened during your study period before the M7 event? What would this mean for your results, then?
Response 10:
Thank you very much for your comments. Analyzing the seismological data of each seismological stations can fully understand the situation of the earthquake and its geographical impacts on the Atsuma River basin. However, the impacts of the earthquake on the sediment transport in the river basin is mainly caused by the landslide caused by the earthquake. On the one hand, The landslide will change the land use of the river basin, making the area that was originally forest, grassland or cultivated land transform into bare land. When it is raining, the land will produce more sediment and flow into the river. On the other hand, some debris and other debris generated by landslides will directly flow into the river, which will directly increase the amount of river sediment after the earthquake in a short time.
To study the changes in sediment transport in the Atsuma River basin after the earthquake. Firstly, it is necessary to make sure whether the sediment transport in the Atsuma River basin after the earthquake has been increased. On the basis that there is no change in the data and parameters of SWAT model, if there is no earthquake, or the earthquake has no impact on the sediment production and transport in the basin, there is no change of the relationship and correlation between turbidity and sediment transport before and after the earthquake. Through the analysis of Figure 7, it can be clearly found that the correlation and relationship between sediment transport and turbidity before and after the earthquake have change greatly, which proves that after the earthquake, the sediment transport in the Atsuma River basin has been changed. Base on the apparent increase in turbidity after the earthquake and the correlation between turbidity and sediment transport, it can be judged that the amount of sediment transport in the Atsuma River basin increased after the earthquake. By modifying the sediment parameters of SWAT model, the sediment transport process after the earthquake can be roughly simulated, and the two simulation results can be compared to analyze the increase in sediment per unit rainfall caused by the earthquake.
Point 11: Line 271: …….a large amount of soil and sand enter into the river,
How do you know this? Did you make investigation in the field about the size of the sediments? What about the different particle sizes and size distribution? Composition?
Response 11:
Thank you very much for your comments. Firstly, through observation of satellite imagery, it can be found that there are many places in the upper reaches of the basin where sediment and other debris generated by landslides directly flowed into the river. Secondly, through field investigations, it can be confirmed that the collapse caused by the landslide caused some sediments to directly accumulate in the river. In the field investigation, the size of the landslide sediment was not investigated, but according to previous investigations that flatlands in this area including Atsuma River basin were covered with less cohesive sedimentation of many pyroclastic fall deposits (volcanic ash, pumice, and scoria). The topsoil layer is generally composed of alternate layers of pyroclastic fall deposits and buried humus, with its thickness about 2.5 to 3.5 m on the middle of the slope.
The bed of Atsuma River basin are composed of 23% gravel, 40% sand and 37% silt and clay. In gravel , the proportion of medium gravel of 19.0 mm～4.75 mm has the largest proportion, in sand, the proportion of medium sand of 0.850 mm～0.250 mm has the largest proportion.
Point 12: Line 287: The landslide debris directly entered the river channel which produced more sediment load. ?
Response 12:
Thank you very much for your comments. The impacts of landslides caused by earthquakes on sediments in river basin are mainly divided into two aspects. On the one hand, the landslides change the land use of the river basin, making some area that was woodland, grassland or cultivated originally transform into bare land, when it is raining, more sediment will produce and flow into the river channel. On the other hand, the sediment and other debris generated by the landslide flows directly into the river channel after the earthquake in a short time, directly increasing the amount of river sediment. According to field investigations and satellite imagery, there are some sediment and other debris generated by landslides flow directly into the river after the earthquake in a short time, resulting in more sediment load.
Point 13: This manuscript presents interesting insights and relationships between a stronger earthquake event and its impact on sediment flow on a quantitative data base. The SWAT model can simulate runoff processes in the Atsuma River basin with a relatively high accuracy. It is recommended, however, to discuss and mention the limits and critical aspects of this model in the discussion.
Response 13:
Thank you very much for your comments. The discussion part really lacks the discussion of the limits and critical aspects of the SWAT model. According to your suggestion, additional related content has been added to Line 374. The content added are: the SWAT model is powerful and can accurately simulate various processes in the basin. In the process of using the SWAT model, the accuracy of the data and the understanding of various parameters are very important. The SWAT model also has some limitations. For example, professional software such as GIS and ENVI are required to prepare geographic data, which increases the difficulty of use. The operation of the SWAT model also requires the user to have high hydrological knowledge, so it is difficult for beginners to get started quickly. Secondly, the SWAT model simulates various processes in the basin based on the input geographic data, meteorological data and other data. SWAT model is limited by the data and model structure that can't fully reproduce the real situation of the basin, and it is also have the same results with different parameters, which requires further optimization.
Finally, we would like to express our heartfelt thanks again to you for your constructive comments and hard work. On behalf of all authors.
Author Response File: Author Response.docx
Round 2
Reviewer 1 Report
The reviewed paper now is better readable and English wording and grammar need only small corrections. However, the authors still do not discuss in details the main critical point. The SWAT model is based on the USLE application and therefore does not consider shallow landslides or deep seated landlsides. These processes, as the authors also illustrate, are the dominant processes after the earthquake. As I understand it correctly, the authors adjust the sediment transport parameters of SWAT to "match" the turbidity values. This might be also a valuable practice if the sediments are only produced by the sheet (rill-interrill) erosion processes. The authors should clearly discuss first which erosion processes are active and why. The model should reflect these processes. As far as I understand in the present model version the sediments are only produced by sheet erosion not by landslides. However, obviously landslides as reported by the authors play a mayor role in sediment contribution. Anyway, the landslide processes generally have different frequency and magnitudes compared to sheet erosion. So I am wondering that you might get the "right" curves modifying the wrong processes.
It is also not clear why the K-factor was doubled...maybe the substrates changing? ....surely not in the whole catchment....or if so please document or prove. Does lumped model values are used or is the k-factor spatially distributed as the soil map? However, this is not explained in detail and makes it hard to evaluate the study.
Comments for author File: Comments.pdf
Author Response
Response to Reviewer 1 Comments
Point 1:The reviewed paper now is better readable and English wording and grammar need only small corrections. However, the authors still do not discuss in details the main critical point. The SWAT model is based on the USLE application and therefore does not consider shallow landslides or deep seated landlsides. These processes, as the authors also illustrate, are the dominant processes after the earthquake. As I understand it correctly, the authors adjust the sediment transport parameters of SWAT to "match" the turbidity values. This might be also a valuable practice if the sediments are only produced by the sheet (rill-interrill) erosion processes. The authors should clearly discuss first which erosion processes are active and why. The model should reflect these processes. As far as I understand in the present model version the sediments are only produced by sheet erosion not by landslides. However, obviously landslides as reported by the authors play a mayor role in sediment contribution. Anyway, the landslide processes generally have different frequency and magnitudes compared to sheet erosion. So I am wondering that you might get the "right" curves modifying the wrong processes.
Response 1:
First of all, thank you very much for your constructive comments and hard work.
It is true that the SWAT model is based on the USLE application and does not consider shallow or deep landslides directly, SWAT model can consider the impact of landslides on sediment transport in the basin. Whether it is shallow or deep landslide, the impacts of landslides on the sediment in river basin are mainly divided into two aspects. On the one hand, the landslide changes the land use of the basin, making the area that was woodland, grassland or cultivated originally transform into bare land. When it is raining, the bare land will produce more sediment and the sediment will flows into the river channel. This aspect mainly affects the erosion process of the watershed. On the other hand, after the earthquake, the sediment, debris generated by the landslide flow directly into the river channel in a short time, increasing the amount of river sediment and erodibility of the river channel.
By analyzing the landslide data, it can be seen that after the earthquake, the total landslides area is 43.8 km2 which including 33.1 km2 in the Atsuma River basin; the total accumulation area is 11.8 km2 which including 9.6 km2 in the Atsuma River basin. Therefore, the area where the landslide occurs is much larger than the accumulation area, and by observing the accumulation area, it can be found that most of the accumulation area are at the foot of the mountain far away from the river channel. After the earthquake, the sediment and debris generated by the landslide directly flowed into the river in a short period of time was small proportion. What needs attention is that, based on the research results of others, after the earthquake and large-scale landslides, the safety factor under natural conditions in the Atsuma River basin is very large, which shows that there will be no slope landslides without rainfall (The updated content was in Line 119). The sediments after the earthquake are most produced by the erosion processes.
This manuscript believes that by changing the parameters of sediment production and transport in the watershed, the sediment transport process in the Atsuma River basin after the earthquake can be evaluated to the greatest extent numerically. For example, the parameter "CH_COV" means the channel erodibility factor is conceptually similar to the soil erodibility factor used in the USLE equation. Channel erodibility is a function of properties of the bed or bank materials. By adjusting this parameter, you can simulate a landslide to make the river bank Plants are reduced, the land is exposed, the erodibility of the river bank is increased, and more sediment will be produced after rain; the parameter "USLE_K" represents Soil erodibility, by increasing this parameter, the reduction of plants in the landslide/accumulation area caused by the landslide can be simulated, The erodibility of the accumulation area caused by exposed land and landslides increases; CH_EROD represents the erodibility of the river channel. The larger the parameter, the more sediment the river itself can produce. The sediment generated by the landslide directly entering the river channel after the earthquake can be increased by increasing this parameter to simulate the sediment directly entering the river channel by the landslide; the parameters SPCON and SPEXP can affect the sediment transport capacity of the river channel. By increasing these two parameters, The sand transport capacity of the river is coordinated with the simulation. This not only takes into account the changes in the erosion process of the drainage basin after the earthquake, but also considers the impact of sediments and debris generated by the landslide into the river channel in a short period of time on the sediment transport in the drainage basin.
This manuscript uses the SWAT model to study the changes in sediment transport in the Atsuma River Basin before and after the earthquake. In addition to making full use of the advantages of the SWAT model in simulating watershed runoff and sediment transportation, the simulated sediment transportation was compared with the observation of turbidity. Comparing the turbidity data with the calculated sediment transport can help estimate the sediment transport in the Atsuma River Basin after the earthquake and analyze the sediment transport per unit of precipitation before and after the earthquake. The change of turbidity has important reference significance for judging the change of sediment transport. The change of turbidity has an important reference significance for judging the change of sediment transport. Changes in land use and sediment directly flowing into the river channel will affect the change of turbidity. Observed turbidity data can directly reflect the characteristics of sediment changes in each stage of the basin. By analyzing the changes in turbidity, it can be intuitively judged that the amount of sediment transport in the basin after the earthquake has increased. In view of the absence of measured sediment transport data in the thick water basin, this manuscript uses observed turbidity to correct and evaluate the sediment simulation process of the SWAT model, and try to make up for the fact that the SWAT model cannot directly cover the sediment that flows directly into the river channel after the landslide Quantity. By adjusting the SWAT sediment transport parameters, the numerical results of the sediment transport simulation are matched with the characteristics of turbidity performance, so as to maximize the situation of sediment transport in the basin after the earthquake, and this is also carried out in the article. Description: Line 340: Although there is no observation sediment transport data for calibration, the sediment transport increased can be roughly assessed. The predictions presented here should be viewed more as qualitative trends, rather than as accurate absolute numerical predictions.
Point 2: It is also not clear why the K-factor was doubled...maybe the substrates changing? ....surely not in the whole catchment....or if so please document or prove. Does lumped model values are used or is the k-factor spatially distributed as the soil map? However, this is not explained in detail and makes it hard to evaluate the study.
Response 2:
Thank you very much for your comments. About sediment parameter. Because this manuscript does not have measured sediment transport data, SWAT-CUP can not be used for sensitivity analysis and correction of sediment parameters. Therefore, this study relies on experience and multiple simulation attempts to select and adjust the sediment parameters. According to previous research, field investigation, the correlation between sediment transport with turbidity and multiple simulation attempts, choose and correct the sediment transport parameters after earthquake manually.
The parameter "USLE_K" represents Soil erodibility. The impact of earthquakes and landslides caused by earthquakes on the erodibility of the watershed will not affect the entire watershed, but the specific impact area and the impact distribution in the impact area are difficult to judge. For the convenience of research, a lumped value was used for the parameter "USLE_K". The lumped model values change the soil erodibility of the entire watershed to achieve a rough assessment of the sediment transport after the earthquake.
The value of K-factor is manually determined through multiple experiments. By observing the landslide area of the watershed, it also can be found that a large area of landslide has changed the soil structure and the vegetation on the soil surface increasing the soil erodibility in the Atsuma River basin.
Point 3: Line 14: Please mention here the full name with the Model Acronym in brackets.
Response 3:
Thank you very much for your comments. The full name of the model has been added, and the model acronym has been marked in brackets. The updated content was in Line 14.
Point 4: Line 35: what do you mean with gravity erosion? Graviational processes or water erosion? Please specify.
Response 4:
Thank you very much for your comments. In this manuscript, the gravity erosion, referred to as gravitational erosion or mass erosion, is the mass failure on steep slope triggered by self-weigh. Gravity erosion usually occurs randomly and it combines with hydraulic erosion. Gravity erosions including landslide, earth flow and creep. In an event of rainfall, various types of gravity erosion might emerge in the same period. Climate-driven factors and topography triggers had prominent influences on gravity erosion. The updated content was in Line 36.
Point 5: Line 198: Please explain in detail the setting of the values....How you defined them based on lierature....or own measurements???
Line 328: how you set the correction parameters on which criteria?
Response 5:
Thank you very much for your comments. SWAT model itself had the function of generating default parameters according to the soil data, land use data, and elevation data of the study area. Researchers could also adjust the parameters by historical hydrological data or the SWAT-CUP software.
About runoff parameter. This manuscript uses the SWAT-CUP software to analyze the runoff parameter sensitivity and to modify parameters accordingly. Sensitivity analysis was performed to identify the most sensitive runoff parameters for the model calibration using Global Sensitivity part of SWAT-CUP. The sensitive runoff parameters were automatically calibrated using the Sequential Uncertainty Fitting (SUFI-2) algorithm. The SWAT model was run daily for 11 years, the warm up period is from 2009 to 2014, the period from 2015 to 2017 was used for the calibration and the period from 2018 to 2019 was used for the validation. The parameters obtained through the calibration period are verified in the verification period.
About sediment parameter. Because this manuscript does not have measured sediment transport data, SWAT-CUP can not be used for sensitivity analysis and correction of sediment parameters. Therefore, this manuscript relies on experience and multiple simulation attempts to select and adjust the sediment parameters. According to previous research, field investigation and the correlation between sediment transport with turbidity, choose and correct the sediment transport parameters after earthquake manually. The updated content was in Line 193.
Point 6: Line 200: Do I understand it correctly that the K-Factor was set to a lumped value? Why? You have a soil map so distributed k-Factor values should be used maybe according to the texture values. Why the K-factor is changing and where? It is not plausible to change the factor for the whole catchment..... Please explain.
Response 6:
Thank you very much for your comments. Yes, the K-Factor was set to a lumped value. Although this is not reasonable enough, it is difficult to judge the precise impact area of the earthquake and the landslide caused by the earthquake in the Atsuma River basin, and the different case of the distribution of the degree of regional influence. In order to facilitate the study, and without the pursuit of absolute accuracy in the simulation of sediment transport, a lumped value was used for the K-Factor parameter. Just like using a lumped hydrological model to simulate the hydrological process of a watershed, although the model can not accurately and truly reflect the watershed conditions, the simulation results also have research significance.
Point 7: Line 288: ok that is true but SWAT simulates only sheet erosion (rill-interrill erosion) whereas the turbidity is reflecting sheet erosion adn all other processes that contribute to the sediment yield.
Response 7:
Thank you very much for your comments. Both the land use changes and the sediment which flows into river channel directly would influence the change of turbidity and sediment transport. By analyzing Figure 3, it can be found that only a few landslide and accumulation caused by the earthquake are located in the river channel. And based on the research results of others, after the earthquake and large-scale landslides, the safety factor under natural conditions in the Atsuma River basin is very large, which shows that there will be no slope landslides without rainfall. Therefore, it can be judged that the impact of sediment produced and transported in the basin after the earthquake is mainly erosion. The use of turbidity data to help correct the sediment transport process is also to numerically minimize the impact of processes other than erosion on the sediment transport.
Point 8: Line 301: ok, but is it due to sheet erosion or due to higher landslide frequency due to earthquake's destabilization of the slopes. landslide procresses are not modelled by SWAT!!!
Response 8:
Thank you very much for your comments. The SWAT model is based on the USLE application and does not directly consider shallow or deep landslides, but can consider the impact of landslides on sediment transport in the basin. The impact of landslides on the basin is mainly divided into two aspects. On the one hand, landslides have changed the land use of the watershed, turned areas that were originally forest, grassland or cultivated land into bare land. When it rains, the bare land will produce more sediment, and the sediment will flow into the river. This aspect mainly affects the erosion process of the watershed. On the other hand, after the earthquake, the sediment and debris produced by the landslide flowed directly into the river channel in a short time, directly increasing the amount of sediment and erodibility of the river channel. What needs attention is that, based on the research results of others, after the earthquake and large-scale landslides, the safety factor under natural conditions in the Atsuma River basin is very large, which shows that there will be no slope landslides without rainfall. The sediment transport increased is due to erosion increased after the earthquake. This study believes that by changing the parameters of sediment produced and transported in the Atsuma River basin, the sediment transport process in the Atsuma River basin after the earthquake can be evaluated to the greatest extent numerically. Both the land use changes and the sediment which flows into river channel directly would influence the change of turbidity. Observed turbidity data can directly reflect the characteristics of sediment change at each stage of the basin. By analyzing the changes of turbidity, it can intuitively judge the changes in the amount of sediment transport in the basin after the earthquake.
Point 9: Line 323: Please show these "correction" values you used and motivate why you changed to this values based on what ?
Response 9:
Thank you very much for your comments. These "corrected" values are the sediment parameters used in the SWAT model after the earthquake. These sediment parameters are summarized in the "sediment" section of Table 2. Regarding sediment parameters, because this manuscript does not have measured sediment transport data, SWAT-CUP can not be used for sensitivity analysis and correction of sediment parameters. Therefore, this manuscript relies on experience and multiple simulation attempts to select sediment parameters to adjust. According to previous studies, field investigations, the relationship between sediment transport and turbidity and multiple manual adjustments to adjust sediment parameters, and the values of sediment parameters were finally determined.
Point 10: Line 362: ok but why ? due to increased sheet erosion? due to landslides and debris flows?
This is not revealed by the authors. Moreover, the increase could also be due to higher precipitation or more intense precipitation in 2019 in respect to 2018....e.g. more shallow landslides ?
Response 10:
Thank you very much for your comments. Based on the research results of others, after the earthquake and large-scale landslides, the safety factor under natural conditions in the Atsuma River basin is very large, which shows that there will be no slope landslides without rainfall. This manuscript believes that most of the sediment transport increased is due to increased erosion after the earthquake. Landslides increase the erodibility of landslide, accumulation and adjacent areas. It can also be seen from Table 4 that the precipitation and runoff in 2019 were less than that in 2018. In the case of less precipitation and runoff, the increase of turbidity in 2019 can prove the increase of sediment transport in 2019.
Finally, we would like to express our heartfelt thanks again to you for your constructive comments and hard work.
Author Response File: Author Response.docx
Reviewer 2 Report
Dear Authors,
thank you for your efforts to consider the review suggestions.
Best wishes
Author Response
Thank you very much.
