the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 17 Jul 2023
River flow in the near future: a global perspective in the context of a high-emission climate change scenario
Abstract. There is high confidence that global warming intensifies all components of the global water cycle. Our goal is to investigate the possible effects of the global warming on river flows worldwide in the coming decades. We conducted 18 global hydrological simulations to assess how the river flows are expected to change in the near future (2015–2050) compared to the recent past (1950–2014). The simulations are forced by runoff from HighResMIP-CMIP6 GCMs, which assume a high-emission scenario for the projections. The assessment includes estimating the signal-to-noise (S/N) ratio and the time of emergence (ToE) of all the rivers in the world, with further evaluation of those presenting significant departures from their historic mean flow. Consistent with the water cycle intensification, the hydrological simulations project a clear positive global river discharge trend from ~2000, that emerges beyond the levels of natural variability and becomes 'unfamiliar' by 2017 and 'unusual' by 2033. This climate change signal is dominated by strong increases in flows of rivers originating in central Africa, east Russia, Alaska and Greenland. African rivers project most future annual cycles above the climatological annual cycle, with the largest differences occurring during peak flows. Recent unprecedent floods in the Republic of Congo, D.R.C., Nigeria, and Chad highlight the potential catastrophic consequences of these changes in metropolitan areas. However, the positive trend of Lake Chad tributaries may aid its recovery from the strong reduction observed since the 1970s. Lastly, the projected Nile streamflow rise reinforces the need for collaboration in dam management. The simulated and observed extra release of freshwater into the Arctic Ocean produces a freshening of the ocean, potentially impacting the global ocean overturning circulation. It is concerning that several important rivers are projected to exceed their natural variability. However, the hydrological predictions assume a very high baseline emission scenario and should be interpreted as an upper limit for decision-making.
- Preprint
(10526 KB) - Metadata XML
- BibTeX
- EndNote
Status: closed
-
RC1: 'Comment on egusphere-2023-1281', Anonymous Referee #1, 14 Oct 2023
General comments:
The aim of this work is to provide a global assessment of near future river flow changes within the context of a high-emission climate change scenario. The authors leverage the TRIP model to simulate river flow dynamics, which is forced by runoff from HighResMIP simulations. They then quantify the changes by using the signal to noise ratio and time of emergency metrics. I have two major concerns on this work: 1) the need for a thorough justification, validation, and demonstration of signal and noise estimations; 2) the absence of logistical and substantive content that elucidates the implications for metropolitan areas and water resource management. Please see my detailed comments below.
Major comments:
- Abstract: the flow of the abstract from L9 to L14 is hard to follow. They did not conclusively reflect what has been discussed in the main text. Please consider refine the abstract.
- L77-79 and Appendix A: additional evidence is necessary to support the hidden assumption that all models share similar internal variability as HadGEM-GC31. Do the remaining models contain ensemble members of precipitation? If so, how does the precipitation spread across the ensemble members? Do they share the similar internal variability as the precipitation spread of HadGEM-GC31? Consider exploring such variables that could influence runoff to earn the statement.
- While the authors direct readers to Müller et al. (2021a) for details on the TRIP model, it would be beneficial to provide a brief overview of TRIP’s global-scale performance and the rationale for its selection.
- L110-114: The demonstration of the estimation of signal and noise ratio needs more details: 1) the mathematical details of the low-pass filter, justification for its choice, and an explanation of how the choice of low-pass filter might impact the noise term; 2) regarding the noise term, why this could be a representation of “natural variability”? Do the variation of the river flow in the PRESENT period include impact of human water activities? If so, how would this impact be separated from the “natural variability”? 3) why the signal of a local river flow anomaly can be linearly regressed on the global signal given large spatial heterogeneity in local river flow variability. These terms need to be better demonstrated in the method section and the robustness of the estimation of signal to noise ratio need to be proved.
- In Figure 2, the difference in runoff between the two periods in Greenland and central Australia is small while in Figure 3 (a) the percentage changes in river flow for these locations are outstanding. Is this discrepancy due to the use of the percent change as a measure, rather than absolute differences? I think it would be beneficial to explain the discrepancy.
- About ToE: again, the method of signal and noise decomposition would largely determine the results of ToE. I wonder if the author could check the robustness of the ToE as well.
- Figure 7: the discussion on the results shown in Figure 7 is confusing. In my perspective, changes are described in the context of a simulation scenario so that they are not yet scalable to real world situation. However, the authors mentioned about the risk to metropolitan areas and the implications to infrastructure management. I think these discussions are relevant but somehow feel disconnected from the results of the simulation. To better connect, I think the authors should first demonstrate how the simulated river flow for PRESENT period compare to observations and form the discussion based on it. This might also help condense the content in the abstract.
Minor comments:
- L68: Clarification is needed regarding what “availability”
- L86-87: Please specify the resolution of the target grid.
- L97-98: the term “expected changes in rivers” requires clarification. Rationale is needed to connect the three steps.
Citation: https://doi.org/10.5194/egusphere-2023-1281-RC1 - AC1: 'Reply on RC1', Omar Müller, 24 Nov 2023
-
RC2: 'Comment on egusphere-2023-1281', Anonymous Referee #2, 23 Oct 2023
The manuscript titled "River flow in the near future: a global perspective in the context of a high-emission climate change scenario" investigates the potential effects of global warming on river flows worldwide from 2015-2050 with hydrological simulations. The study aims to provide insights into the potential changes in river flows and their broader socio-environmental consequences. Overall, the manuscript is well structured and clear. However, the following points raise my concerns.
1) There seems to be a lack of explicit validation of simulated river flows by observations from the same historical period. In general, it is important to ensure that the modeling framework accurately reproduces observed river flows in the historical period before trusting projections for the future.
2) section 3.2.1 is overly descriptive, providing detailed information about individual rivers, their importance to their respective regions, historical context, and model projections. While such detail is valuable to some extent, it may be more than necessary for the main message of the article. I recommend a more concise way to convey the broader implications of this study. For example, group rivers with similar trends and mention the specifics only when they significantly deviate from the general trend. The implications for human settlements, global water systems, and climate systems can be summarized in one final paragraph. In addition, explaining the "why" behind the trend or time of emergence can provide more insightful analysis and make the findings more compelling. Also, the discussion section reads repetitively, which seems to summarize the above finding again.
3) For the information in Appendix A, it's a bit of a leap to conclude that "it may be assumed that our set of simulations is adequate for the proposed objectives and that more realizations would not present substantial alter the presented results" without information about the internal variability of other models. While this may be true for the HadGEM3-GC31 model, it may be premature to state this as a broader conclusion without data from other models to support it.
Other points:
1) Figure 4, can you explain why most of the solid lines show a downward trend after 2045?
2) Figure 5, I'm curious if the order of calculating the ensemble mean and calculating the signal-to-noise ratio would have an impact, let's say calculating the signal-to-noise ratio of the ensemble mean of the simulation instead.Editorial point:
1) Figure 1, there seem to be no squares in the three main plots.Citation: https://doi.org/10.5194/egusphere-2023-1281-RC2 - AC2: 'Reply on RC2', Omar Müller, 24 Nov 2023
Status: closed
-
RC1: 'Comment on egusphere-2023-1281', Anonymous Referee #1, 14 Oct 2023
General comments:
The aim of this work is to provide a global assessment of near future river flow changes within the context of a high-emission climate change scenario. The authors leverage the TRIP model to simulate river flow dynamics, which is forced by runoff from HighResMIP simulations. They then quantify the changes by using the signal to noise ratio and time of emergency metrics. I have two major concerns on this work: 1) the need for a thorough justification, validation, and demonstration of signal and noise estimations; 2) the absence of logistical and substantive content that elucidates the implications for metropolitan areas and water resource management. Please see my detailed comments below.
Major comments:
- Abstract: the flow of the abstract from L9 to L14 is hard to follow. They did not conclusively reflect what has been discussed in the main text. Please consider refine the abstract.
- L77-79 and Appendix A: additional evidence is necessary to support the hidden assumption that all models share similar internal variability as HadGEM-GC31. Do the remaining models contain ensemble members of precipitation? If so, how does the precipitation spread across the ensemble members? Do they share the similar internal variability as the precipitation spread of HadGEM-GC31? Consider exploring such variables that could influence runoff to earn the statement.
- While the authors direct readers to Müller et al. (2021a) for details on the TRIP model, it would be beneficial to provide a brief overview of TRIP’s global-scale performance and the rationale for its selection.
- L110-114: The demonstration of the estimation of signal and noise ratio needs more details: 1) the mathematical details of the low-pass filter, justification for its choice, and an explanation of how the choice of low-pass filter might impact the noise term; 2) regarding the noise term, why this could be a representation of “natural variability”? Do the variation of the river flow in the PRESENT period include impact of human water activities? If so, how would this impact be separated from the “natural variability”? 3) why the signal of a local river flow anomaly can be linearly regressed on the global signal given large spatial heterogeneity in local river flow variability. These terms need to be better demonstrated in the method section and the robustness of the estimation of signal to noise ratio need to be proved.
- In Figure 2, the difference in runoff between the two periods in Greenland and central Australia is small while in Figure 3 (a) the percentage changes in river flow for these locations are outstanding. Is this discrepancy due to the use of the percent change as a measure, rather than absolute differences? I think it would be beneficial to explain the discrepancy.
- About ToE: again, the method of signal and noise decomposition would largely determine the results of ToE. I wonder if the author could check the robustness of the ToE as well.
- Figure 7: the discussion on the results shown in Figure 7 is confusing. In my perspective, changes are described in the context of a simulation scenario so that they are not yet scalable to real world situation. However, the authors mentioned about the risk to metropolitan areas and the implications to infrastructure management. I think these discussions are relevant but somehow feel disconnected from the results of the simulation. To better connect, I think the authors should first demonstrate how the simulated river flow for PRESENT period compare to observations and form the discussion based on it. This might also help condense the content in the abstract.
Minor comments:
- L68: Clarification is needed regarding what “availability”
- L86-87: Please specify the resolution of the target grid.
- L97-98: the term “expected changes in rivers” requires clarification. Rationale is needed to connect the three steps.
Citation: https://doi.org/10.5194/egusphere-2023-1281-RC1 - AC1: 'Reply on RC1', Omar Müller, 24 Nov 2023
-
RC2: 'Comment on egusphere-2023-1281', Anonymous Referee #2, 23 Oct 2023
The manuscript titled "River flow in the near future: a global perspective in the context of a high-emission climate change scenario" investigates the potential effects of global warming on river flows worldwide from 2015-2050 with hydrological simulations. The study aims to provide insights into the potential changes in river flows and their broader socio-environmental consequences. Overall, the manuscript is well structured and clear. However, the following points raise my concerns.
1) There seems to be a lack of explicit validation of simulated river flows by observations from the same historical period. In general, it is important to ensure that the modeling framework accurately reproduces observed river flows in the historical period before trusting projections for the future.
2) section 3.2.1 is overly descriptive, providing detailed information about individual rivers, their importance to their respective regions, historical context, and model projections. While such detail is valuable to some extent, it may be more than necessary for the main message of the article. I recommend a more concise way to convey the broader implications of this study. For example, group rivers with similar trends and mention the specifics only when they significantly deviate from the general trend. The implications for human settlements, global water systems, and climate systems can be summarized in one final paragraph. In addition, explaining the "why" behind the trend or time of emergence can provide more insightful analysis and make the findings more compelling. Also, the discussion section reads repetitively, which seems to summarize the above finding again.
3) For the information in Appendix A, it's a bit of a leap to conclude that "it may be assumed that our set of simulations is adequate for the proposed objectives and that more realizations would not present substantial alter the presented results" without information about the internal variability of other models. While this may be true for the HadGEM3-GC31 model, it may be premature to state this as a broader conclusion without data from other models to support it.
Other points:
1) Figure 4, can you explain why most of the solid lines show a downward trend after 2045?
2) Figure 5, I'm curious if the order of calculating the ensemble mean and calculating the signal-to-noise ratio would have an impact, let's say calculating the signal-to-noise ratio of the ensemble mean of the simulation instead.Editorial point:
1) Figure 1, there seem to be no squares in the three main plots.Citation: https://doi.org/10.5194/egusphere-2023-1281-RC2 - AC2: 'Reply on RC2', Omar Müller, 24 Nov 2023
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 402 | 180 | 33 | 615 | 22 | 25 |
- HTML: 402
- PDF: 180
- XML: 33
- Total: 615
- BibTeX: 22
- EndNote: 25
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1