Dynamic Risks for Cascading Himalayan Hazards

Lead Research Organisation: University of Edinburgh
Department Name: Sch of Geosciences


There are a wide range of natural hazards that impact communities living within, and at the edge of the Himalayan mountains; these are dominated by earthquakes, landslides and floods. In order to reduce the risk from landslides and floods, communities have developed early warning systems to downstream villages and towns, enabling pre-planned responses. Early warning systems require local authorities to be aware of potential dangers. For example, a steep hillslope that is known to be unstable with evidence of past landslides should be monitored, particularly during periods of heavy rainfall. However, in order for the local District Disaster Management Authorities (DDMAs) to know where to monitor, medium-term forecasts of the likely risk from different hazards need to be known. For example, certain areas are more prone to earthquakes, and others to landslides and flashfloods. If these risks from different hazards remain constant through time, then the forecasts and monitoring for each community remains steady. However, hazards do not act in isolation, but form cascades, each event triggering another. As a result, the risk from multiple hazards is not stable, but dynamic, and changes in response to upstream triggers. For example a landslide, that leads to a dam that breaks out to form a debris flow that then increases subsequent risk to floods due to choking of river channels with sand and gravel.

This project aims to provide the first fully quantitative forecasts of multihazard cascades using a range of new modelling techniques constrained by a history of field observations from the Garwhal Himalaya, Uttarakhand. This area has been devastated by recent landslides and flashfloods such as the Kedarnath disaster in 2013 and the Chomli landslide in 2021. Thick accumulations of sediment in these steep mountain valleys are known as 'sediment bombs' as they pose a danger to downstream communities; such sediment bombs may form where glaciers retreat or where landslides block valleys. In this project, the Indian and UK teams will combine to integrate new methodologies from digital topography, remote sensing, computer models and field monitoring to understand how sediment yield from glaciers and landslides initiate sediment bombs, and how these accumulations are then mobilised to form debris flows, flash floods and downstream flooding. Through understanding the distribution and rates involved in these processes, we will generate medium term forecasts that feed into early warning systems developed in the communities of the Alaknanda Valley.

The approach as outlined above suggests that the physical science models will be the sole input into consideration of dynamic risk; but it can't be as simple as that. The communities that live with this risk, and the DDMAs that manage the early warning systems have to be involved in the generation and iteration of the scientific methodology. Consequently, we are working with social scientists in the UK and India who have experience working with communities in the Himalaya through workshops and interviews that respect the diverse cultural, ethnic and gender-based perspectives. By the end of the project, we will have generated a decisional workflow for district authorities that integrates dynamic risk into their medium term forecasts in response to cascading hazards. Having demonstrated this process in the Garwhal Himalaya, we intend to work with the National Disaster Management Authorities in India and Nepal to promote national strategies for dynamic risk assessment.


10 25 50