Hazard cascades initiated by landslides sourced above glaciers and glacial lakes


In the Kingdom of Bhutan growing glacial lakes pose a known threat of Glacial Lake Outburst Floods (GLOFs) that pose a risk to remote communities and large urban areas (Carrivick and Tweed, 206). Most of these lakes are surrounded by extremely steep, high slopes that meet the conditions for large landslides failure, and, rock avalanches are known from other parts of the Kingdom (Dunning et al. 2006).

Hazard cascades that originate from the failure of slopes above glaciers and glacial lakes pose a significant, and, debatably, an increasing risk to life and infrastructure. The Nov. 2020 Elliot Creek, Feb. 2021 Chamoli (Shugar et al. 2021), and, March 2021 Sendongpu hazard cascades all originated as ice:rock avalanches on high, remote mountains where a ‘normal’ runout would have caused few issues, but, displacement waves from glacial lakes; transformation into mobile debris flow/floods; temporary landslide dams, and, large-scale sediment transport greatly extended their risk footprints, with potential for long-term impacts. In all instances it is the presence of water that allows the initial solid dominated landslide to become a far more mobile flow/flood, this can be from the rapid melting of glacial ice during landslide motion, or, from a glacial lake.

In recent years there has been rapid development in our ability to detect slopes that pose threats, and, in modelling landslide and flood inundation either as separate components of a hazard cascade, or, all within one modelling environment (Frey et al. 2018; Mergili et al. 2020) . In many instances the modelling of a hazard cascade is responsive, a back analysis to yield better understanding of an event. In these cases, we often have validation data we expect our models to fit. However, if we want to help Governments and agencies tasked with disaster risk reduction before an event, it is far more challenging (Marlovits et al. 2021).

Here we aim to use ensemble modelling approaches using cutting edge codes (RAMMS; a.avaflow) to characterise down valley inundation and arrival times from landslides and their resultant hazard cascades in the Kingdom of Bhutan. We will then use this information to quantitatively specify the most beneficial locations of early warning systems (EWS) and sensors to provide life information on the nature of the threat.


There are two fundamental questions to be answered:
Q1: Can we use ensemble modelling to quantify locations at persistent risk of inundation during landslide-initiated hazard cascades?
Q2: Are there locations along potential hazard cascades that not only allow detection of an ‘event’, but, also the most likely model scenario it is following?

To do this the successful student will:

O1: Undertake a critical review of current modelling approaches for hazard cascades
O2: Undertake fieldwork to measure parameters essential to modelling hazard cascades that cannot be quantified remotely
O2: Use appropriate runout codes such as (but based on O1-O2 findings) RAMMS, a.avaflow, HEC-RAS to model hazard cascades from landslide source, over glaciers, and through glacial lakes, and then along valley floors
O3: Develop methods to assess the siting of sensors and early warning systems to both detect hazards cascades, and, link them to the ensemble model runs

Fieldwork and model setup/verification, and the potential siting of monitoring will be undertaken in close collaboration with the Royal Government of Bhutan (The National Center for Hydrology and Meteorology (NCHM)) and the United Nations World Food Programme (WFP).

Project Timeline

Year 1

T1a: Undertake a critical review of existing regional scale (DEM/remote sensing) approaches to locate slopes that pose a risk, apply to Bhutan
T1b: Review approaches to modelling inundation and time to arrival of hazard cascades from ice:rock landslides
T1c: Determine what variables require on-site measurement and monitoring to feed into models

Year 2

T2a: Carry out fieldwork in the Kingdom of Bhutan alongside project collaborators
T2b: Undertake ensemble numerical modelling of the propagation of hazard cascades in Bhutan

Year 3

T3a: Complete numerical modelling of hazard cascades
T3b: Develop methods to quantify the siting of sensors and monitoring to link field data to model scenarios.

Year 3.5

T3.5a: Finalise model outputs and present to the Royal Government of Bhutan
T3.5b: Finish thesis write up

& Skills

The successful student will develop skills in the application of numerical modelling, which will be supported through appropriate training.
Field skills will be developed through integration into a team of scientists researching GLOFs; flooding; glacial lake change; real-time monitoring.

References & further reading

Carrivick, J.L. and Tweed, F.S. (2016). A global assessment of the societal impacts of glacier outburst floods. Global and Planetary Change, 144. pp. 1-16. ISSN 0921-8181
Dunning, S.A., Rosser, N.J., Petley, D.N. et al. (2006). Formation and failure of the Tsatichhu landslide dam, Bhutan. Landslides 3, 107–113 https://doi.org/10.1007/s10346-005-0032-x
Frey et al. (2018). Multi-Source Glacial Lake Outburst Flood Hazard Assessment and Mapping for Huaraz, Cordillera Blanca, Peru. Frontiers in Earth Science 6.
Marlovits, N., Mergili, M., and Glade, T.: Challenges in the predictive simulation of cascading landslide processes , EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-14333, https://doi.org/10.5194/egusphere-egu21-14333, 2021.
Mergili et al. (2020). Reconstruction of the 1941 GLOF process chain at Lake Palcacocha (Cordillera Blanca, Peru). Hydrol. Earth Syst. Sci. 24, 93-114
Shugar, D. H et al. (2021). A massive rock and ice avalanche caused the 2021 disaster at Chamoli, Indian Himalaya, Science, 373, 6552: 300-306

Further Information

Please contact Dr Stuart Dunning for further information stuart.dunning@ncl.ac.uk

Apply Now