the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 24 Apr 2024
On the Cause of Large Daily River Flow Fluctuations in the Mekong River
Abstract. Natural fluctuations in river flow are central to the ecosystem productivity of basins, yet significant alterations in daily flows pose threats to the integrity of the hydrological, ecological, and agricultural systems. In the dammed Mekong River, the attribution of these large daily flow changes to upstream regions remains mechanistically unexamined, a factor blamed on challenges in estimating the time required for large daily shifts in upstream river flow to impact the downstream regions. Here, we address this by integrating a newly developed sub-basin modeling framework that incorporates 3D hydrodynamic, response time, and hydrological models. This integration allows us to estimate the time required between two hydrological stations and to distinguish the contribution of sub-basins and upstream regions to large daily river flow alterations. Findings revealed a power correlation between river discharge and the required time to reach downstream stations. Significant fluctuations in the river's daily flow were evident before the advent of the era of human activities, i.e., before 1992. This phenomenon persisted throughout subsequent periods, including the growth period from 1992 to 2009 and the mega-dam period spanning from 2010 to 2020, with minimal variation in the frequency of events. Sub-basins were found to significantly contribute to mainstream discharge- a contribution which led to a significant contribution of sub-basins into mainstream daily large river flow shifts. The outcomes and model derived from the sub-basin approach hold significant potential for managing river fluctuations and have broader applicability beyond the specific basin studied.
- Preprint
(1750 KB) - Metadata XML
-
Supplement
(1253 KB) - BibTeX
- EndNote
Status: closed
-
RC1: 'Comment on hess-2024-96', Anonymous Referee #1, 18 Jun 2024
Natural fluctuations in the river are essential to the ecosystem productivity of basins. Which has less been investigated in the dammed Mekong River Basin. In view of this, this manuscript integrated a framework consisting of hydrological model, 3D hydrodynamic model, response time to address this issue. Results show that significant fluctuations in the river's daily flow were evident before the advent of the era of human activities. Further, the sub-basins were found to significantly contribute to mainstream discharge fluctuation. Overall, this manuscript is interesting, which can attract a lot of attention from readers. However, there were still some drawbacks before it is published on this journal and were listed below for references.
Major comments:
(1) The author stated that “research on the daily assessment of large river flow alterations is limited, with most researchers focusing on monthly, seasonal, and annual scale studies.” (lines 59-60), which could be hard to make readers convinced. Many studies related to the discharge or floods (especially for floods) in the Mekong River Basin focused on daily scale, such as Wang et al., 2017, Wang et al., 2021 (listed by authors as references in the manuscript), Try et al. (2020), Yun et al. (2024). The word “most” could be not proper. More importantly, there should be an overview of researches on daily assessment of river flow before stating the lack of daily assessment of large river flow alterations.
Try, S., Tanaka, S., Tanaka, K., Sayama, T., Oeurng, C., Uk, S., ... & Han, D. (2020). Comparison of gridded precipitation datasets for rainfall-runoff and inundation modeling in the Mekong River Basin. PLoS One, 15(1), e0226814.
Yun, X., Song, J., Wang, J., & Bao, H. (2024). Modelling to assess the suitability of hydrological-hydrodynamic model under the hydropower development impact in the Lancang-Mekong river basin. Journal of Hydrology, 131393.
(2) The description for data was simple. Data from seven stations extending from Chiang Saen (CS) to Kratie (KR) stations were collected, however, these stations were not clearly marked in Figure 2 (only with red circles, no name was shown). This could have an impact on readers who were not familiar with this basin. In addition, the authors said that they collect many meteorological and precipitation data, but no spatial map for these sites was shown or information for these sites was revealed. It was worthy to note that the description for meteorological and precipitation data should be placed in 2.2.1, instead in 2.2.2.
(3) The methods were described relatively simple, with many confusions left, though the supplement information also contained some basic information. Firstly, readers did not know how authors calibrated the THREW model and the Delft-3D flow model, who also did not know what the parameters and inputs for these models were. Secondly, people also did not know how authors inputted the outputs of THREW model to Delft-3D flow model. I guessed that the authors used the simulated discharge near the mainstream to input to the hydrodynamic model. More importantly, how author used the meteorological data to prepare the inputs of THREW model remained uncleared (e.g., interpolating the meteorological data from in-situ scale to gridded scale).
(4) I noticed that the author used discharge to calculated the contribution to discharge, then why the hydrodynamic model was used in this manuscript. Many studies have shown that the hydrological model can well produce the discharge upstream Kratie. The author can only used hydrological model to make analyses. By the way, I am not sure why the author analyzed the velocity, which could be not important as discharge.
(5) Delft-3D flow model is a small-scale hydrodynamic model, how could author apply this model to the large basin (i.e., Mekong River Basin).
(6) The authors used “sub-basin” and “upstream station” terms many times in Section 3. For a given station, what did “sub-basin” and “upstream station” refer to. For example, in Figure 7, what did “upstream station” and “sub-basin” refer to for “PA”. Could I think the “upstream station” was the nearest upstream station for a given station.
(7) The legends in Figures 9, 10 were missing. In Figure 9, what did red line, grey and blue bars represent. In Figure 10, what did the x-axis represent, For CS, why did eight bars occur. and then what did the red line represent.
Minor comments:
(1) Line 63: Usually，the trend of discharge change is similar to that of water level. Here, I am not sure why discharge increased by 98% while the water level decreased by -1.55m.
(2) Line 88: The length of the Mekong River needs further confirmation. It seems that 4500km is not a commonly used result. According to MRC (2006), the correct value is 4800km. Further, “Mekong River constitutes the third most diverse aquatic ecosystem”, what were the first and second most diverse aquatic ecosystem. Mekong River should not be the second most diverse aquatic ecosystem (just followed by Amazon River Basin)?
MRC, 2006. Annual Flood Report 2005. Mekong River Commission, Vientiane, Lao PDR,
- 82.
(3) Lines 96-97: The authors took June-December as the wet season, while took November-May as the dry season. This was not consistent with the facts. Actually, the flood season is from June to December for Mekong River Basin, while wet season is from May to October (see Räsänen and Kummu, 2013, Wang et al., 2022 for reference)
Räsänen, T. A., & Kummu, M. (2013). Spatiotemporal influences of ENSO on precipitation and flood pulse in the Mekong River Basin. Journal of Hydrology, 476, 154-168.
Wang, J., Tang, Q., Yun, X., Chen, A., Sun, S., & Yamazaki, D. (2022). Flood inundation in the Lancang-Mekong River Basin: Assessing the role of summer monsoon. Journal of Hydrology, 612, 128075.
(4) Line 226: how authors defined the daily river flow alteration, whether the authors using the water level in the next day minus that in the current day.
- AC1: 'Reply on RC1', Khosro Morovati, 21 Jun 2024
-
RC2: 'Reply on AC1', Anonymous Referee #2, 03 Jul 2024
This study investigated the cause of the large daily flow fluctuations in the Mekong River. After reading the manuscript, I have a strong feeling that the manuscript needs to be carefully revised and reviewed. Precise and clear writing is important and sufficient to report the new findings to our scientific community. Especially for the figures, some irrelevant paragraphs and unclear descriptions would confuse the readers. The authors have done a lot of work to support their findings. But the current version still needs to be improved.
Main comments：
1) I concur with the previous reviewer's assessment that the author's literature review of this paper requires substantial supplementation with recent content, particularly the modelling and simulation of a series of hydrological and hydrodynamic models conducted around the Lancang-Mekong River Basin. Given that 2024 has already commenced, the modelling conducted around the LMRB has been refined to the day or even the hour. In light of the above, it is imperative that the author conducts a comprehensive synthesis and refinement of existing research, elucidating the pivotal contributions of this study. It should be noted that these works should not only be carried out in the discussion, but also require substantial supplementation and modification of the introduction.
2) The authors hope to estimate the time it takes for large daily changes in upstream rivers to affect downstream rivers, but with the large-scale construction of reservoirs and changes in river dynamics, the results of this study may not provide the expected reference value. Similarly, the authors' claim that "three aspects extend previous research" is difficult to achieve:
- a) "Quantitative assessment of the regional contribution to abnormal downstream water level/flow changes". Given that there are about 500 reservoirs in the basin, I doubt the feasibility of this vision;
- b) "Quantifying the propagation of upstream river flow changes to downstream sub-basins", as above, the presence of many reservoirs has significantly altered the river propagation process. Although the impact of reservoirs on mainstream flooding during the wet season is small, it should be noted that reservoir operations dominate mainstream water level changes during the dry season in the basin, and large-scale water conservation and diversion projects on tributaries have permanently altered river dynamics in these areas.
- c) Due to the lack of consideration of the reservoir impact in the model, this study may only be applicable to the LMRB before 2009, and it is difficult to provide an in-depth understanding of climate impacts. Figures 3 and 4 confirm this view. The author can only show the time series verification results before 2000, and lacks the evaluation of the model effect on the tributaries and mainstream in the middle and upper reaches after the large-scale reservoir development after 2008.
3) This study may not be applicable to current LMRB. Given that this manuscript submitted to HESS, I am a little unsure what new insights this paper can give us regarding the LMRB, especially considering that the basin has been undergoing large-scale dam construction for 20 years. Could the authors consider looking at other areas where dam construction has not yet begun, to increase the the validity of the study?
4) It should be pointed out that the author's model can obtain such a high NSE coefficient, which is mainly due to the input of the actual streamflow of the JH station. In fact, if the JH flow data is used directly to evaluate the CS flow data without considering the confluence runoff in the JH-CS sub-basin, its NSE will reach more than 0.85. However, I can't find any description of the JH station flow in the article. Considering that the streamflow data of JH station has been publicly released by the Chinese government, it is necessary for the author to make a detailed explanation.
5) As far as I know, THREW is not a gridded distributed model, but a model for lumped confluence in small catchments. How could this driven the Delft-3D model? I can't imagine flattening the confluence generated by the lumped model on an uneven DEM and expecting it to produce adequate confluence results.
6) I was unable to open the website https://portal.mrcmekong.org/home successfully, whether using the network service from German, Japan or China. Perhaps the author could consider uploading the data to such as https://zenodo.org/ for safekeeping.
Other comments：
7) At line88, Firstly, the official name of this basin is the Lancang-Mekong River Basin, with upstream Lancang River and downstream Mekong River. Secondly, the length of the river claimed by the author is questionable. Finally, the number of Chinese reservoirs is more than 11 and needs further verification. Considering that the collaborators include a large number of senior Chinese experts in hydraulic research, it is unacceptable to make mistakes in these details and data.
8) In Figure 1, what is “the Tonle Sap Lak”? It is recommended that the author carefully checks for the spelling and grammatical errors in the paper, as similar situations occur frequently.
9) In Figure 2, I don't think it's a good idea to use both red circles and triangles for labeling. I can't distinguish the tributary station and the mainstream station at all. Besides, I think there should be a space separating the “Delft3D”.
10) Line 220, "Comparable levels of accuracy are achieved for the years 2019 and 2020, as detailed in the SM, Section 3". My understanding is that you cannot prove the overall model usability by showing only a part. ST is located downstream and has a large main stream flow, making it less affected by reservoir operation. Therefore, using flow velocity assessment at a monthly scale during the rainy season can give better results, but this cannot prove the applicability of the model for basin-wide flow assessment after 2010.
11) The results in Figure 8 seem to be based on the comparison between the actual observed flow and the natural flow simulated by the model, or did the authors include a simulation of reservoir operation in the model? I am not sure if I missed the part about the reservoir being set up in the model. It is recommended that the authors explain how the results were obtained.
12) In Figure 9, I don't think it's a good idea to remove the year labels on the x-axis of the time series graph, as this only makes the figure harder to understand. Also, what is "recieved rainfall"? It is recommended to avoid the use of rainfall and to use precipitation uniformly. I suggest that the author consider further detailed checks on the grammar, fonts, font size, etc. of the full text and images. The current version has too many errors.
13) In Figure 10, Same as above, what is “contrinution”? I can understand that the author has a few singular and plural errors or tense problems in the manuscript. However, repeated typing errors in important figures are unacceptable.
14) In Figure 11, Please add dots of corresponding colors on the basis of the lines in the legend, which can make the image more readable.
15) In Figures 6 and 9, I am not sure how R2 is calculated, what data are used? Please explain in detail.
16) Sources of meteorological soil and vegetation DEM data used in modelling must be listed in the main text in a clear and detailed manner. Layered citations are unacceptable.
-
AC2: 'Reply on RC2', Khosro Morovati, 12 Jul 2024
The comment was uploaded in the form of a supplement: https://hess.copernicus.org/preprints/hess-2024-96/hess-2024-96-AC2-supplement.pdf
Status: closed
-
RC1: 'Comment on hess-2024-96', Anonymous Referee #1, 18 Jun 2024
Natural fluctuations in the river are essential to the ecosystem productivity of basins. Which has less been investigated in the dammed Mekong River Basin. In view of this, this manuscript integrated a framework consisting of hydrological model, 3D hydrodynamic model, response time to address this issue. Results show that significant fluctuations in the river's daily flow were evident before the advent of the era of human activities. Further, the sub-basins were found to significantly contribute to mainstream discharge fluctuation. Overall, this manuscript is interesting, which can attract a lot of attention from readers. However, there were still some drawbacks before it is published on this journal and were listed below for references.
Major comments:
(1) The author stated that “research on the daily assessment of large river flow alterations is limited, with most researchers focusing on monthly, seasonal, and annual scale studies.” (lines 59-60), which could be hard to make readers convinced. Many studies related to the discharge or floods (especially for floods) in the Mekong River Basin focused on daily scale, such as Wang et al., 2017, Wang et al., 2021 (listed by authors as references in the manuscript), Try et al. (2020), Yun et al. (2024). The word “most” could be not proper. More importantly, there should be an overview of researches on daily assessment of river flow before stating the lack of daily assessment of large river flow alterations.
Try, S., Tanaka, S., Tanaka, K., Sayama, T., Oeurng, C., Uk, S., ... & Han, D. (2020). Comparison of gridded precipitation datasets for rainfall-runoff and inundation modeling in the Mekong River Basin. PLoS One, 15(1), e0226814.
Yun, X., Song, J., Wang, J., & Bao, H. (2024). Modelling to assess the suitability of hydrological-hydrodynamic model under the hydropower development impact in the Lancang-Mekong river basin. Journal of Hydrology, 131393.
(2) The description for data was simple. Data from seven stations extending from Chiang Saen (CS) to Kratie (KR) stations were collected, however, these stations were not clearly marked in Figure 2 (only with red circles, no name was shown). This could have an impact on readers who were not familiar with this basin. In addition, the authors said that they collect many meteorological and precipitation data, but no spatial map for these sites was shown or information for these sites was revealed. It was worthy to note that the description for meteorological and precipitation data should be placed in 2.2.1, instead in 2.2.2.
(3) The methods were described relatively simple, with many confusions left, though the supplement information also contained some basic information. Firstly, readers did not know how authors calibrated the THREW model and the Delft-3D flow model, who also did not know what the parameters and inputs for these models were. Secondly, people also did not know how authors inputted the outputs of THREW model to Delft-3D flow model. I guessed that the authors used the simulated discharge near the mainstream to input to the hydrodynamic model. More importantly, how author used the meteorological data to prepare the inputs of THREW model remained uncleared (e.g., interpolating the meteorological data from in-situ scale to gridded scale).
(4) I noticed that the author used discharge to calculated the contribution to discharge, then why the hydrodynamic model was used in this manuscript. Many studies have shown that the hydrological model can well produce the discharge upstream Kratie. The author can only used hydrological model to make analyses. By the way, I am not sure why the author analyzed the velocity, which could be not important as discharge.
(5) Delft-3D flow model is a small-scale hydrodynamic model, how could author apply this model to the large basin (i.e., Mekong River Basin).
(6) The authors used “sub-basin” and “upstream station” terms many times in Section 3. For a given station, what did “sub-basin” and “upstream station” refer to. For example, in Figure 7, what did “upstream station” and “sub-basin” refer to for “PA”. Could I think the “upstream station” was the nearest upstream station for a given station.
(7) The legends in Figures 9, 10 were missing. In Figure 9, what did red line, grey and blue bars represent. In Figure 10, what did the x-axis represent, For CS, why did eight bars occur. and then what did the red line represent.
Minor comments:
(1) Line 63: Usually，the trend of discharge change is similar to that of water level. Here, I am not sure why discharge increased by 98% while the water level decreased by -1.55m.
(2) Line 88: The length of the Mekong River needs further confirmation. It seems that 4500km is not a commonly used result. According to MRC (2006), the correct value is 4800km. Further, “Mekong River constitutes the third most diverse aquatic ecosystem”, what were the first and second most diverse aquatic ecosystem. Mekong River should not be the second most diverse aquatic ecosystem (just followed by Amazon River Basin)?
MRC, 2006. Annual Flood Report 2005. Mekong River Commission, Vientiane, Lao PDR,
- 82.
(3) Lines 96-97: The authors took June-December as the wet season, while took November-May as the dry season. This was not consistent with the facts. Actually, the flood season is from June to December for Mekong River Basin, while wet season is from May to October (see Räsänen and Kummu, 2013, Wang et al., 2022 for reference)
Räsänen, T. A., & Kummu, M. (2013). Spatiotemporal influences of ENSO on precipitation and flood pulse in the Mekong River Basin. Journal of Hydrology, 476, 154-168.
Wang, J., Tang, Q., Yun, X., Chen, A., Sun, S., & Yamazaki, D. (2022). Flood inundation in the Lancang-Mekong River Basin: Assessing the role of summer monsoon. Journal of Hydrology, 612, 128075.
(4) Line 226: how authors defined the daily river flow alteration, whether the authors using the water level in the next day minus that in the current day.
- AC1: 'Reply on RC1', Khosro Morovati, 21 Jun 2024
-
RC2: 'Reply on AC1', Anonymous Referee #2, 03 Jul 2024
This study investigated the cause of the large daily flow fluctuations in the Mekong River. After reading the manuscript, I have a strong feeling that the manuscript needs to be carefully revised and reviewed. Precise and clear writing is important and sufficient to report the new findings to our scientific community. Especially for the figures, some irrelevant paragraphs and unclear descriptions would confuse the readers. The authors have done a lot of work to support their findings. But the current version still needs to be improved.
Main comments：
1) I concur with the previous reviewer's assessment that the author's literature review of this paper requires substantial supplementation with recent content, particularly the modelling and simulation of a series of hydrological and hydrodynamic models conducted around the Lancang-Mekong River Basin. Given that 2024 has already commenced, the modelling conducted around the LMRB has been refined to the day or even the hour. In light of the above, it is imperative that the author conducts a comprehensive synthesis and refinement of existing research, elucidating the pivotal contributions of this study. It should be noted that these works should not only be carried out in the discussion, but also require substantial supplementation and modification of the introduction.
2) The authors hope to estimate the time it takes for large daily changes in upstream rivers to affect downstream rivers, but with the large-scale construction of reservoirs and changes in river dynamics, the results of this study may not provide the expected reference value. Similarly, the authors' claim that "three aspects extend previous research" is difficult to achieve:
- a) "Quantitative assessment of the regional contribution to abnormal downstream water level/flow changes". Given that there are about 500 reservoirs in the basin, I doubt the feasibility of this vision;
- b) "Quantifying the propagation of upstream river flow changes to downstream sub-basins", as above, the presence of many reservoirs has significantly altered the river propagation process. Although the impact of reservoirs on mainstream flooding during the wet season is small, it should be noted that reservoir operations dominate mainstream water level changes during the dry season in the basin, and large-scale water conservation and diversion projects on tributaries have permanently altered river dynamics in these areas.
- c) Due to the lack of consideration of the reservoir impact in the model, this study may only be applicable to the LMRB before 2009, and it is difficult to provide an in-depth understanding of climate impacts. Figures 3 and 4 confirm this view. The author can only show the time series verification results before 2000, and lacks the evaluation of the model effect on the tributaries and mainstream in the middle and upper reaches after the large-scale reservoir development after 2008.
3) This study may not be applicable to current LMRB. Given that this manuscript submitted to HESS, I am a little unsure what new insights this paper can give us regarding the LMRB, especially considering that the basin has been undergoing large-scale dam construction for 20 years. Could the authors consider looking at other areas where dam construction has not yet begun, to increase the the validity of the study?
4) It should be pointed out that the author's model can obtain such a high NSE coefficient, which is mainly due to the input of the actual streamflow of the JH station. In fact, if the JH flow data is used directly to evaluate the CS flow data without considering the confluence runoff in the JH-CS sub-basin, its NSE will reach more than 0.85. However, I can't find any description of the JH station flow in the article. Considering that the streamflow data of JH station has been publicly released by the Chinese government, it is necessary for the author to make a detailed explanation.
5) As far as I know, THREW is not a gridded distributed model, but a model for lumped confluence in small catchments. How could this driven the Delft-3D model? I can't imagine flattening the confluence generated by the lumped model on an uneven DEM and expecting it to produce adequate confluence results.
6) I was unable to open the website https://portal.mrcmekong.org/home successfully, whether using the network service from German, Japan or China. Perhaps the author could consider uploading the data to such as https://zenodo.org/ for safekeeping.
Other comments：
7) At line88, Firstly, the official name of this basin is the Lancang-Mekong River Basin, with upstream Lancang River and downstream Mekong River. Secondly, the length of the river claimed by the author is questionable. Finally, the number of Chinese reservoirs is more than 11 and needs further verification. Considering that the collaborators include a large number of senior Chinese experts in hydraulic research, it is unacceptable to make mistakes in these details and data.
8) In Figure 1, what is “the Tonle Sap Lak”? It is recommended that the author carefully checks for the spelling and grammatical errors in the paper, as similar situations occur frequently.
9) In Figure 2, I don't think it's a good idea to use both red circles and triangles for labeling. I can't distinguish the tributary station and the mainstream station at all. Besides, I think there should be a space separating the “Delft3D”.
10) Line 220, "Comparable levels of accuracy are achieved for the years 2019 and 2020, as detailed in the SM, Section 3". My understanding is that you cannot prove the overall model usability by showing only a part. ST is located downstream and has a large main stream flow, making it less affected by reservoir operation. Therefore, using flow velocity assessment at a monthly scale during the rainy season can give better results, but this cannot prove the applicability of the model for basin-wide flow assessment after 2010.
11) The results in Figure 8 seem to be based on the comparison between the actual observed flow and the natural flow simulated by the model, or did the authors include a simulation of reservoir operation in the model? I am not sure if I missed the part about the reservoir being set up in the model. It is recommended that the authors explain how the results were obtained.
12) In Figure 9, I don't think it's a good idea to remove the year labels on the x-axis of the time series graph, as this only makes the figure harder to understand. Also, what is "recieved rainfall"? It is recommended to avoid the use of rainfall and to use precipitation uniformly. I suggest that the author consider further detailed checks on the grammar, fonts, font size, etc. of the full text and images. The current version has too many errors.
13) In Figure 10, Same as above, what is “contrinution”? I can understand that the author has a few singular and plural errors or tense problems in the manuscript. However, repeated typing errors in important figures are unacceptable.
14) In Figure 11, Please add dots of corresponding colors on the basis of the lines in the legend, which can make the image more readable.
15) In Figures 6 and 9, I am not sure how R2 is calculated, what data are used? Please explain in detail.
16) Sources of meteorological soil and vegetation DEM data used in modelling must be listed in the main text in a clear and detailed manner. Layered citations are unacceptable.
-
AC2: 'Reply on RC2', Khosro Morovati, 12 Jul 2024
The comment was uploaded in the form of a supplement: https://hess.copernicus.org/preprints/hess-2024-96/hess-2024-96-AC2-supplement.pdf
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 472 | 103 | 24 | 599 | 49 | 17 | 24 |
- HTML: 472
- PDF: 103
- XML: 24
- Total: 599
- Supplement: 49
- BibTeX: 17
- EndNote: 24
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1