the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 08 Aug 2025
A shifting pattern of tropical cyclone induced high river discharges in the Greater Mekong Region, 1970–2019
Abstract. On average flood events impact over 100 million people globally every year, and because of demographic changes and economic development in flood-prone areas, as well as climate change, the population exposed to flood risk is expected to double by 2050. Under anthropogenic climate change it is expected that flood events previously considered extreme will be occurring with more frequency, due to changing patterns of precipitation in a warming climate. It is, therefore, critically important to better understand how extreme weather events generate high river flows in exposed regions. Here we look specifically at the influence of precipitation from tropical cyclone (TC) activity on high river flows within one such exposed region: a 1.2 million km2 area of South-east Asia encompassing the entirety of the Mekong and Red River catchments, plus 13 smaller catchments along the coastal fringe of Vietnam (collectively referred to here as the Greater Mekong region, or GMR). We use a hydrological model (GM-HYPE) with ERA5 precipitation data to simulate streamflows over the last 50 years (1970–2019) with, and without, TC-linked precipitation. Our results demonstrate that TC-linked precipitation around the GMR generate notable increases in high (95th percentile) streamflows, and this is most notable in the steep sub-catchments draining to Vietnam’s northern coastline. These locations are more exposed to TC activity, and we determine that the elevated soil moisture levels there from monsoonal precipitation, prior to the typhoon season, are an exacerbating factor. Furthermore, trend analysis also shows that shifts in the spatial locations of TC-induced high river flows have been occurring since the 1970s: while statistically significant increases in TC-induced high river discharges are evident in localised regions of the GMR including the highlands of Laos and the Mekong’s delta region, declines in TC-induced high river discharges are much more widespread, with notable declines in the headwater and middle reaches of the Red and Mekong Rivers. Our findings on the changing pattern of high river flows in recent decades, in a region highly exposed to TCs, will be of great interest to strategic planners and flood managers. We conclude with a discussion on the impact of global climate model precipitation projections for this region, contrasting past/present (1980–2014), and future (2016–2050), GM-HYPE model results.
- Preprint
(2295 KB) - Metadata XML
-
Supplement
(1701 KB) - BibTeX
- EndNote
Status: final response (author comments only)
- RC1: 'Comment on egusphere-2025-3506', Anonymous Referee #1, 08 Oct 2025
-
RC2: 'Comment on egusphere-2025-3506', Anonymous Referee #2, 15 Apr 2026
This manuscript investigates the influence of tropical cyclone (TC)–induced precipitation on high river discharges across the Greater Mekong Region using the GM-HYPE hydrological model forced with ERA5 data, with and without TC precipitation data. In the following, some points that would benefit from further consideration and clarification.
- A central element of the study is the identification and removal of TC-induced precipitation using a fixed 500 km radius around the cyclone centre. While this approach follows previous studies, it represents a strong simplification as the use of a fixed buffer may therefore lead to the inclusion of non-TC precipitation and, conversely, the exclusion of TC-related rainfall occurring outside the selected radius. This introduces a source of uncertainty that could affect the estimated contribution of TCs to streamflow. It would be helpful if the authors could discuss this limitation more explicitly and, if possible, explore the sensitivity of the results to the choice of the buffer size or consider alternative, more physically based approaches for attributing precipitation to cyclones.
- While the overall performance of the GM-HYPE model is described as satisfactory, the authors does not provide a targeted validation of the TC-related signal itself. It would strengthen the study to demonstrate that the model is capable of reproducing observed discharge responses during TC events. For example, a comparison with documented flood events associated with specific cyclones, or an event-based evaluation of model performance during TC periods, would provide stronger support for the interpretation of the results
- The analysis of the drivers of excess high flows identifies antecedent soil moisture as the dominant explanatory factor. This finding is consistent with hydrological understanding; however, it is derived from a relatively simple regression framework that does not include interaction terms and may not fully capture the non-linear nature of these hydrological processes. Clarifying this aspect and exploring temporal dynamics (e.g. antecedent conditions with lags) would strengthen the analysis.
- The characterization of high flows is mainly based on the 95th percentile of discharge, which provides a useful first-order indicator of extremes but does not fully describe flood behaviour. Approaches based on on return periods, for istance, could allow a more reliable assessment of flood hazard and its changes under TC influence.
- The analysis is based on the use of long-term dataset spanning 1970–2019. The authors acknowledge that both ERA5 and IBTrACS data are less reliable before the satellite era, but this potential limitation is not explored further in the analysis. Given the known limitations of pre-satellite data, the authors could maybe consider restricting the analysis to a more reliable period, applying correction techniques, or at least explicitly testing the sensitivity of the results to the inclusion of earlier data.
- From my understanding, the analysis of future conditions is not fully consistent with the historical framework as TC precipitation cannot be isolated in the GCM simulations. Future projections, indeed, include all sources of precipitation, making the comparison with the TC-specific historical results not consistent. Clarifying this point and carefully framing the interpretation would improve the coherence of the study. I also feel that this analysis focused on the use of future projections should be deepened if the authors would include it within the manuscript (e.g., which projections have been used, have they been corrected in some way, etc.) as otherwise this part seems to be extremely marginally.
In summary, the study addresses an interesting topic and presents a promising modeling strategy, but some key methodological assumptions and limitations should be more thoroughly examined.
Citation: https://doi.org/10.5194/egusphere-2025-3506-RC2
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 2,079 | 159 | 35 | 2,273 | 121 | 54 | 50 |
- HTML: 2,079
- PDF: 159
- XML: 35
- Total: 2,273
- Supplement: 121
- BibTeX: 54
- EndNote: 50
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
The paper quantifies the influence of TC activity on high river flows within a 1.2 million km2 area of southeast Asia encompassing the Mekong and Red River catchments, plus 13 smaller catchments located along the coastal fringe of Vietnam. In 2020, this Greater Mekong Region (GMR) supported a population 85 of over 160 million people.
1.The use of a 500km radius to crop TC related precipitation in the article does not fully demonstrate the applicability of this radius in the GMR region.
2.Table 1 only considers three variables (precipitation, soil moisture, slope), ignoring possible important factors such as land use, reservoir regulation, and previous rainfall. Suggest explaining the possible impacts of these potential factors in the discussion.
3.The author points out that the reliability of data in the 1970s and 1980s was low, but does not evaluate the specific impact on trend analysis.
4.Figure 5 shows future changes, but does not provide confidence intervals or inter model differences (such as multi model sets).
5.The abstract should succinctly summarize the research objectives, methods, key findings, and conclusions. I suggest reducing background information in the abstract and focusing more on the study's highlights and outcomes.
6.Some newest research work related with this paper can be added in the introduction. Diffusion evolution rules of grouting slurry in mining-induced cracks in overlying strata. Water injection softening modeling of hard roof and application in Buertai coal mine.
7.The conclusion mentions “useful for planners and managers”, but does not provide specific recommendations (such as which areas should prioritize strengthening flood control facilities). The authors should provide a summary of their main findings, the limitations of the study, and recommendations for future research in the conclusion.
8. There are several grammar mistakes in the paper, please revise and double-check throughout the whole manuscript. The language could benefit from further editing and polishing.