Simulating avalanche-triggered lake overspill and downstream impacts at Birendra Lake using RAMMS and HEC-RAS

Thapa, Sujan; Vaidya, Ragini; Chand, Mohan Bahadur; Kayastha, Rijan Bhakta

doi:10.5194/egusphere-2025-4454

Preprints

https://doi.org/10.5194/egusphere-2025-4454

Preprints

29 Sep 2025

| 29 Sep 2025

Simulating avalanche-triggered lake overspill and downstream impacts at Birendra Lake using RAMMS and HEC-RAS

Sujan Thapa, Ragini Vaidya, Mohan Bahadur Chand, and Rijan Bhakta Kayastha

Abstract. The study presents the first comprehensive quantitative assessment of avalanche-triggered GLOF hazards at Birendra Lake using integrated RAMMS-HEC-RAS modelling to evaluate cascading risks from avalanche release to downstream flood propagation. Three scenarios representing small (5.1 × 10⁴ m³), medium (5.3 × 10⁵ m³), and large (1.2 × 10⁶ m³) avalanche releases from steep slopes (30°–48.8°) surrounding the lake were simulated. The modelling framework demonstrates that all scenarios reach Birendra Lake with substantial mass retention (62–86 %), generating maximum velocities of 33.8–72.8 m/s and flow heights of 11.2–36.8 m. The displacement-driven overspill mechanism displaces 0.01–0.18 % of total lake volume (4.7 × 10⁶ m³), producing peak discharge rates of 615.7–3,151.8 m³/s. HEC-RAS flood modelling reveals rapid downstream propagation, with flood arrival times of 0.15–0.43 hours at Samagaon and 4.6–19.76 hours at Jagat, accompanied by maximum flood depths of 0.96–12.69 m and velocities of 1.94–15.62 m/s. The modelling results demonstrate strong qualitative alignment with the April 2024 event, validating the overspill mechanism. Medium to large avalanche scenarios pose severe threats to downstream communities, with the large scenario producing catastrophic conditions at Samagaun, where depths exceed 12 m with velocities above 15 m/s. The findings establish Birendra Lake as an imminent high-risk system where steep avalanche-prone terrain, lake proximity to unstable glacier zones, and significant downstream exposure create catastrophic cascading hazards. This research provides essential quantitative foundations for early warning systems and risk reduction strategies in avalanche-prone glacial lake environments across High Mountain Asia.

Received: 10 Sep 2025 – Discussion started: 29 Sep 2025

Publisher's note: Copernicus Publications remains neutral with regard to jurisdictional claims made in the text, published maps, institutional affiliations, or any other geographical representation in this paper. While Copernicus Publications makes every effort to include appropriate place names, the final responsibility lies with the authors. Views expressed in the text are those of the authors and do not necessarily reflect the views of the publisher.

Download & links

Sujan Thapa, Ragini Vaidya, Mohan Bahadur Chand, and Rijan Bhakta Kayastha

Status: closed (peer review stopped)

RC1:
'Comment on egusphere-2025-4454', Anonymous Referee #1, 19 Oct 2025
The manuscript requires significant restructuring to clearly convey the objective, methodology, results, limitations, and conclusions of the study. Currently, many technical terms, topics, and section headings are introduced abruptly without prior explanation or context, making it difficult for the reader to follow the narrative.

There is a noticeable lack of alignment between the study’s stated objectives, the flowchart presented in Figure 3, the methodology described in Section 2, and the Results and Discussion section. These components should be revised to ensure coherence and logical flow throughout the manuscript.

It is suggested to restructure the methodology into two separate sections as follows:

Section 2 – Study Area

Section 3 – Materials and Methods, with the following subsections:
1 Avalanche Susceptibility Mapping

2 Glacial Lake Volume Estimation

3 RAMMS Modeling

4 HEC-RAS Modeling

5 Exposure Analysis

Each subsection should clearly describe its role in the study, including the input parameters used and references where applicable. A similar structure should also be followed in the Results and Discussion section for consistency. Subsections can be further divided as needed to highlight important components.
Results and Discussion Section

The Results and Discussion section should focus solely on the outputs and findings of the present study. Currently, this section often reads like an extended version of the methodology. The descriptive elements already covered in the methodology should not be repeated. Instead, this section should include clear interpretation and analysis of the model outputs, supported by relevant figures and tables.

There are two unnumbered subsections presented under Results and Discussion which were not introduced or discussed in the methodology:

Temporal Characteristics and Warning Implications

Sensitivity Analysis and Hazard Assessment Implications

These need to be formally integrated into the manuscript structure. Additionally:
The source of the outflow timing estimate (8–10 minutes) should be clarified. Including a figure showing the progression of the flood from initiation to downstream points would greatly enhance clarity.

The parameters considered for sensitivity analysis should be clearly identified and their impact discussed.

Section 3.7, titled Exposure Analysis of Avalanche-Induced Flood Scenarios at Different Sites, currently discusses flood arrival time, maximum flow depth, and velocity. However, it does not include any actual exposure analysis, which by definition involves identifying and quantifying the elements at risk (e.g., population, infrastructure, land use). This section should be revised to include or refer to such analysis, or the section title should be changed to accurately reflect its content.

Figures

The following revisions are suggested for the figures:

The Study Area Map should be labeled as Figure 1.

Figures 4 and 5 present similar information except for the three different release zones; these could be combined into a single comparative figure for better clarity and reduced redundancy.

Figures 6, 7, and 8 ( RAMMS simulation outputs) can be consolidated into one composite figure with clearly labeled panels. Each panel should include color bars indicating the dynamic parameter being shown (e.g., height, velocity, momentum). Be sure to mention that these represent RAMMS outputs in the figure caption.

RAMMS: What is the basis and background for identification, demarcation and consideration of the release zones and the release depth? If the main implementation of the RAMMS model has been done to estimate the volume of material reaching at the lake, then it is not clear why the 2^nd/3^rd scenario was proposed/assumed. Different scenario might have also formulated using different release depth and initial volume at the scenario-I (Small) .

Input Data: Make a table to show all the input data used in this work. (DEM resolution, time, frictional parameters, entrainment (if any), release condition (block/hydrograph) for RAMMS & HECRAC.

HEC RAS: Why was a Manning’s n value of 0.06 considered appropriate for the modelled Himalayan stream reach?

-Why was roughness calibration considered limited or not performed in detail for this model?
-How can sensitivity analysis be incorporated within HEC-RAS modeling to improve flood hazard assessment?
Citation: https://doi.org/10.5194/egusphere-2025-4454-RC1
RC2:
'Comment on egusphere-2025-4454', Adam Emmer, 14 Nov 2025
This study written by Sujan Thapa and colleagues aims at modelling different avalanche scenarios and subsequent dam overtopping-induced floods at Birendra Lake and downstream valley in Nepal. The study would be of interest for the GLOF research community as well as disaster risk management authorities, however, there are several issues that should be addressed before this study is further processed:
Structure – the study suffers from rather unusual structure (intro section seems unusually long; separate discussion section is definitely needed; results section should be structured according to objectives / methods used and needs to only present results; study area section is not a part of the methods, …); as a result, it is difficult to read it and understand the work done

Methods – the use of some of the methods doesn’t seem suitable / justified (the Bühler et al. 2013 methodology was developed for snow avalanches but here the authors model ice-avalanches (different mechanisms / processes); since the time of famous Huggel et al., 2004 lake area-volume scaling relationship, many other Himalaya-focusing methods for estimating lake volume have been developed and published since, providing better performance).

Modelling parameters and assumptions – parameters that are used need justification other than the use in previous studies (e.g. Manning) - please not only use the values but also discuss the performance / sensitivity evaluations from previous studies; what is the unit of curvature 50? and standard deviation of terrain roughness 15m? - please check; Calculated % of released volume are not correct (800,000 m3 of a large scenario is not 0.18% of the lake volume); Why is x axis and the shape of all hydrographs the same, regardless the scenario? How they were created (considering dam overtopping mechainsm and lake dynamics, the shape would be a series of waves rather than one several minutes-lasting wave)?

Terminology – the terminology is not used properly (Exposure – the authors write about exposure in abstract and text but no exposed elements are mapped at the end, flow depth is not a characterics of exposure; vulnerability – the lake is not vulnerable to avalanches but prone to avalanche impacts, …) undermining the work done

The 2024 GLOF event mentioned in the intro provides an opportunity for evaluation the performance of used models; however, no detailed info of this event and its impacts nor further analysis are provided

Recommendations - the list of recommendations are predominantly general and tru for all potential GLOF sites; what site-specific recommendations can be derived from the results of the study?

To sum up, this study needs substantial revisions before it can be considered for publication.
Citation: https://doi.org/10.5194/egusphere-2025-4454-RC2
CC1: 'Comment on egusphere-2025-4454', Nitesh Khadka, 12 Dec 2025

I am happy to read the manuscript by Thapa et al. which have used RAMMAS modeling. Initially, I anticipated this research to be a reconstruction of the 2024 Birendra Lake GLOF event, however, it appears to be focused on a hazard assessment instead. This distinction should be made explicit to avoid potential misunderstandings. Given the numerous published studies after the Birendra GLOF event (Khadka et al., 2024; Banerjee et al., 2025; Chaulagain et al., 2025; Poudel et al., 2025), it is essential for this study to acknowledge these prior works to build up the study and articulate what new insights or findings it brings to show the credibility.
My main concern is that the authors claim 0.01-0.18% of total lake volume 4.7×10⁶ m³ (=8460 m³ max) is spilled from lake which seems speculative and small when comparing to downstream discharges and time given in the manuscript. Please make me clear if I missed anything. Further, impact/exposure assessment as shown in Figure 3 has not been conducted.
L49 What does ice-debris avalanche mean? The feeding glacier is a clean glacier
L50 vulnerability to susceptibility
L52 I think this event is a small event and has not affected or had impacts on downstream community/settlements
L55-L58 cite previous studies that have investigated this event
L61 two semi colon
L94 check the exact cause of the 2016 Gongbatongsaco GLOF
L119 mention the glacial lake
L134 once the full form is abbreviated it is wise to use abbreviation throughout the ms
L145 use another term for vulnerability
L167 glacier detach: is it recently or since the end of LIA? Prove if its recent from satellite imagery
Study area is poorly described. It would be great to know the climatology, glaciers, glacial lakes and their characteristics in the region. Why studying such event is important in the region?Authors could have used also reviewed more published studies in Nepal Himalaya for structuring the background/introduction and for study area (some maybe https://doi.org/10.1007/s10113-023-02142-y, https://doi.org/10.3390/rs9070654, https://doi.org/10.3390/rs10121913)
L226 which high resolution DEM?
L242 in repeated
L285 et al. is missing in the reference when there are more than 2 authors
Table 1 How was depth determined? Is it mean depth or max depth?
L346 Why not to use HMA DEM of better resolution than ALOS?
L348 mention the weblink or cite it
For figures 9, 10 and 11, merge them and describe giving sub-number
L715 cite the published paper not the preprint
Banerjee A, Meadows EM, Yadav N, et al. Glacier and glacial lake dynamics from 1990 to 2024 and their impact on flood hazard in the central Nepal Himalaya[J]. Journal of Mountain Science: 2025, 22:1926-1943. 10.1007/s11629-024-9298-0
Chaulagain M, Chand MB, Pradhananga D, et al. Recurring avalanche hazards at Birendra Lake, Manaslu region: Interdisciplinary insights from the April 21, 2024, avalanche event[J]. 2025, 7:59-77.
Khadka N, Zheng G, Chen X, et al. An ice-snow avalanche triggered small glacial lake outburst flood in Birendra Lake, Nepal Himalaya[J]. Natural Hazards: 2024, 121:6357-6365. https://doi.org/10.1007/s11069-024-07014-0
Poudel U, Gouli MR, Hu K, et al. Multi-breach GLOF hazard and exposure analysis of Birendra Lake in the Manaslu Region of Nepal[J]. Natural Hazards Research: 2025. 10.1016/j.nhres.2025.03.007

Citation: https://doi.org/10.5194/egusphere-2025-4454-CC1

Status: closed (peer review stopped)

RC1:
'Comment on egusphere-2025-4454', Anonymous Referee #1, 19 Oct 2025
The manuscript requires significant restructuring to clearly convey the objective, methodology, results, limitations, and conclusions of the study. Currently, many technical terms, topics, and section headings are introduced abruptly without prior explanation or context, making it difficult for the reader to follow the narrative.

There is a noticeable lack of alignment between the study’s stated objectives, the flowchart presented in Figure 3, the methodology described in Section 2, and the Results and Discussion section. These components should be revised to ensure coherence and logical flow throughout the manuscript.

It is suggested to restructure the methodology into two separate sections as follows:

Section 2 – Study Area

Section 3 – Materials and Methods, with the following subsections:
1 Avalanche Susceptibility Mapping

2 Glacial Lake Volume Estimation

3 RAMMS Modeling

4 HEC-RAS Modeling

5 Exposure Analysis

Each subsection should clearly describe its role in the study, including the input parameters used and references where applicable. A similar structure should also be followed in the Results and Discussion section for consistency. Subsections can be further divided as needed to highlight important components.
Results and Discussion Section

The Results and Discussion section should focus solely on the outputs and findings of the present study. Currently, this section often reads like an extended version of the methodology. The descriptive elements already covered in the methodology should not be repeated. Instead, this section should include clear interpretation and analysis of the model outputs, supported by relevant figures and tables.

There are two unnumbered subsections presented under Results and Discussion which were not introduced or discussed in the methodology:

Temporal Characteristics and Warning Implications

Sensitivity Analysis and Hazard Assessment Implications

These need to be formally integrated into the manuscript structure. Additionally:
The source of the outflow timing estimate (8–10 minutes) should be clarified. Including a figure showing the progression of the flood from initiation to downstream points would greatly enhance clarity.

The parameters considered for sensitivity analysis should be clearly identified and their impact discussed.

Section 3.7, titled Exposure Analysis of Avalanche-Induced Flood Scenarios at Different Sites, currently discusses flood arrival time, maximum flow depth, and velocity. However, it does not include any actual exposure analysis, which by definition involves identifying and quantifying the elements at risk (e.g., population, infrastructure, land use). This section should be revised to include or refer to such analysis, or the section title should be changed to accurately reflect its content.

Figures

The following revisions are suggested for the figures:

The Study Area Map should be labeled as Figure 1.

Figures 4 and 5 present similar information except for the three different release zones; these could be combined into a single comparative figure for better clarity and reduced redundancy.

Figures 6, 7, and 8 ( RAMMS simulation outputs) can be consolidated into one composite figure with clearly labeled panels. Each panel should include color bars indicating the dynamic parameter being shown (e.g., height, velocity, momentum). Be sure to mention that these represent RAMMS outputs in the figure caption.

RAMMS: What is the basis and background for identification, demarcation and consideration of the release zones and the release depth? If the main implementation of the RAMMS model has been done to estimate the volume of material reaching at the lake, then it is not clear why the 2^nd/3^rd scenario was proposed/assumed. Different scenario might have also formulated using different release depth and initial volume at the scenario-I (Small) .

Input Data: Make a table to show all the input data used in this work. (DEM resolution, time, frictional parameters, entrainment (if any), release condition (block/hydrograph) for RAMMS & HECRAC.

HEC RAS: Why was a Manning’s n value of 0.06 considered appropriate for the modelled Himalayan stream reach?

-Why was roughness calibration considered limited or not performed in detail for this model?
-How can sensitivity analysis be incorporated within HEC-RAS modeling to improve flood hazard assessment?
Citation: https://doi.org/10.5194/egusphere-2025-4454-RC1
RC2:
'Comment on egusphere-2025-4454', Adam Emmer, 14 Nov 2025
This study written by Sujan Thapa and colleagues aims at modelling different avalanche scenarios and subsequent dam overtopping-induced floods at Birendra Lake and downstream valley in Nepal. The study would be of interest for the GLOF research community as well as disaster risk management authorities, however, there are several issues that should be addressed before this study is further processed:
Structure – the study suffers from rather unusual structure (intro section seems unusually long; separate discussion section is definitely needed; results section should be structured according to objectives / methods used and needs to only present results; study area section is not a part of the methods, …); as a result, it is difficult to read it and understand the work done

Methods – the use of some of the methods doesn’t seem suitable / justified (the Bühler et al. 2013 methodology was developed for snow avalanches but here the authors model ice-avalanches (different mechanisms / processes); since the time of famous Huggel et al., 2004 lake area-volume scaling relationship, many other Himalaya-focusing methods for estimating lake volume have been developed and published since, providing better performance).

Modelling parameters and assumptions – parameters that are used need justification other than the use in previous studies (e.g. Manning) - please not only use the values but also discuss the performance / sensitivity evaluations from previous studies; what is the unit of curvature 50? and standard deviation of terrain roughness 15m? - please check; Calculated % of released volume are not correct (800,000 m3 of a large scenario is not 0.18% of the lake volume); Why is x axis and the shape of all hydrographs the same, regardless the scenario? How they were created (considering dam overtopping mechainsm and lake dynamics, the shape would be a series of waves rather than one several minutes-lasting wave)?

Terminology – the terminology is not used properly (Exposure – the authors write about exposure in abstract and text but no exposed elements are mapped at the end, flow depth is not a characterics of exposure; vulnerability – the lake is not vulnerable to avalanches but prone to avalanche impacts, …) undermining the work done

The 2024 GLOF event mentioned in the intro provides an opportunity for evaluation the performance of used models; however, no detailed info of this event and its impacts nor further analysis are provided

Recommendations - the list of recommendations are predominantly general and tru for all potential GLOF sites; what site-specific recommendations can be derived from the results of the study?

To sum up, this study needs substantial revisions before it can be considered for publication.
Citation: https://doi.org/10.5194/egusphere-2025-4454-RC2
CC1: 'Comment on egusphere-2025-4454', Nitesh Khadka, 12 Dec 2025

I am happy to read the manuscript by Thapa et al. which have used RAMMAS modeling. Initially, I anticipated this research to be a reconstruction of the 2024 Birendra Lake GLOF event, however, it appears to be focused on a hazard assessment instead. This distinction should be made explicit to avoid potential misunderstandings. Given the numerous published studies after the Birendra GLOF event (Khadka et al., 2024; Banerjee et al., 2025; Chaulagain et al., 2025; Poudel et al., 2025), it is essential for this study to acknowledge these prior works to build up the study and articulate what new insights or findings it brings to show the credibility.
My main concern is that the authors claim 0.01-0.18% of total lake volume 4.7×10⁶ m³ (=8460 m³ max) is spilled from lake which seems speculative and small when comparing to downstream discharges and time given in the manuscript. Please make me clear if I missed anything. Further, impact/exposure assessment as shown in Figure 3 has not been conducted.
L49 What does ice-debris avalanche mean? The feeding glacier is a clean glacier
L50 vulnerability to susceptibility
L52 I think this event is a small event and has not affected or had impacts on downstream community/settlements
L55-L58 cite previous studies that have investigated this event
L61 two semi colon
L94 check the exact cause of the 2016 Gongbatongsaco GLOF
L119 mention the glacial lake
L134 once the full form is abbreviated it is wise to use abbreviation throughout the ms
L145 use another term for vulnerability
L167 glacier detach: is it recently or since the end of LIA? Prove if its recent from satellite imagery
Study area is poorly described. It would be great to know the climatology, glaciers, glacial lakes and their characteristics in the region. Why studying such event is important in the region?Authors could have used also reviewed more published studies in Nepal Himalaya for structuring the background/introduction and for study area (some maybe https://doi.org/10.1007/s10113-023-02142-y, https://doi.org/10.3390/rs9070654, https://doi.org/10.3390/rs10121913)
L226 which high resolution DEM?
L242 in repeated
L285 et al. is missing in the reference when there are more than 2 authors
Table 1 How was depth determined? Is it mean depth or max depth?
L346 Why not to use HMA DEM of better resolution than ALOS?
L348 mention the weblink or cite it
For figures 9, 10 and 11, merge them and describe giving sub-number
L715 cite the published paper not the preprint
Banerjee A, Meadows EM, Yadav N, et al. Glacier and glacial lake dynamics from 1990 to 2024 and their impact on flood hazard in the central Nepal Himalaya[J]. Journal of Mountain Science: 2025, 22:1926-1943. 10.1007/s11629-024-9298-0
Chaulagain M, Chand MB, Pradhananga D, et al. Recurring avalanche hazards at Birendra Lake, Manaslu region: Interdisciplinary insights from the April 21, 2024, avalanche event[J]. 2025, 7:59-77.
Khadka N, Zheng G, Chen X, et al. An ice-snow avalanche triggered small glacial lake outburst flood in Birendra Lake, Nepal Himalaya[J]. Natural Hazards: 2024, 121:6357-6365. https://doi.org/10.1007/s11069-024-07014-0
Poudel U, Gouli MR, Hu K, et al. Multi-breach GLOF hazard and exposure analysis of Birendra Lake in the Manaslu Region of Nepal[J]. Natural Hazards Research: 2025. 10.1016/j.nhres.2025.03.007

Citation: https://doi.org/10.5194/egusphere-2025-4454-CC1

Sujan Thapa, Ragini Vaidya, Mohan Bahadur Chand, and Rijan Bhakta Kayastha

Viewed

Total article views: 673 (including HTML, PDF, and XML)

HTML	PDF	XML	Total	BibTeX	EndNote
488	152	33	673	16	20

HTML: 488
PDF: 152
XML: 33
Total: 673
BibTeX: 16
EndNote: 20

Views and downloads (calculated since 29 Sep 2025)

Month	HTML	PDF	XML	Total
Sep 2025	141	4	1	146
Oct 2025	132	26	12	170
Nov 2025	75	35	7	117
Dec 2025	82	50	13	145
Jan 2026	58	37	0	95

Cumulative views and downloads (calculated since 29 Sep 2025)

Month	HTML	PDF	XML	Total
Sep 2025	141	4	1	146
Oct 2025	132	26	12	170
Nov 2025	75	35	7	117
Dec 2025	82	50	13	145
Jan 2026	58	37	0	95

Viewed (geographical distribution)

Total article views: 663 (including HTML, PDF, and XML) Thereof 663 with geography defined and 0 with unknown origin.

Country	#	Views	%

Latest update: 26 Jan 2026

Short summary

This study examines how avalanches striking a glacial lake can cause sudden floods downstream in Nepal's Manaslu region. We used computer simulations to simulate various avalanche sizes and their effects on the lake. These simulations revealed that medium and large avalanches could lead to severe flooding in a short amount of time. This highlights the pressing need for early warning systems and better disaster preparedness to protect at-risk communities.


Total:	0
HTML:	0
PDF:	0
XML:	0