Melting glaciers in high mountain regions are creating vast new lakes that can hold millions of cubic metres of water, yet scientists cannot reliably predict when these lakes will drain catastrophically or how much water they store for future use. Current estimates of glacial lake volume rely on fewer than 100 field surveys worldwide, producing uncertainties of 20–50 percent. This makes it impossible to assess downstream flood risks or water availability as climate change accelerates lake formation and urban expansion pushes communities into vulnerable mountain valleys. The problem is acute in High-Mountain Asia, where countries like Nepal face potential flood disasters from lakes that can exceed 200 metres in depth. This Fellowship will deploy a multibeam sonar system to map lakebed shapes and underwater glacier structures in Nepal, generating the first high-resolution bathymetry datasets for dozens of lakes. Machine learning models trained on these data will estimate water storage at unsurveyed lakes without costly field expeditions, while flood models using high-resolution topography will identify which buildings and infrastructure are exposed to catastrophic drainage events or extreme rainfall. The results will feed into an open-access online platform—the Glacial Lake Observatory—that tracks lake dynamics globally in near-real time using satellite data. This should allow governments and development agencies in Nepal and beyond to target disaster prevention resources where they will deliver the most benefit, rather than spreading them thinly across all lakes.
View original technical description
A global trend of glacier loss is leading to the development of high-mountain glacial lakes that can exceed kilometres in length and over 200 m in depth, therefore storing vast quantities of water. However, their poorly constrained estimates of current and future water stores restricts assessments of water resource availability and potential downstream flood risks. Glacial lakes can drain seasonally and catastrophically, leading to downstream flooding with high socio-economic impacts, particularly across High-Mountain Asia in countries such as Nepal. These flood events cause widespread concern, spanning mountain communities to development agencies and government departments. However, historical records suggest that most glacial lakes are inherently stable. To ensure that disaster risk-reduction resources are targeted to deliver maximum benefit, and that water resource trends are understood, it is therefore essential to develop a robust evidence base in the context of climate change, accelerating lake development, and urban expansion into mountain regions. High-mountain glacial lake water storage is measured for a small proportion of lakes globally due to logistically challenging survey requirements and the inability to derive depth observations using satellite data. Instead, glacial lake bathymetry datasets are produced through field surveys, and are subsequently used to inform empirical scaling relationships that relate lake area to volume. Estimating water storage using these relationships that are based on few datapoints globally (n~100) contains large uncertainties (>20-50%). Glacial lakes also promote a positive feedback, whereby the thermal energy stored in lake water and buoyancy forces acting on the glacier can accelerate glacier recession. However, similarly sparse observations of lake-glacier interactions mean these mechanisms are not parameterised in models predicting glacier evolution and downstream runoff trends. This Fellowship presents an integrated, interdisciplinary approach to assess both glacial lake development and downstream floods in topographically complex catchments. I will develop an innovative survey methodology to derive the bathymetry of glacial lakes using both a single beam sonar, combined with a cutting edge multibeam sonar system. The latter will produce complete maps of lakebed morphology and reveal the subaqueous glacier structure. Extensive bathymetry surveys in Nepal will underpin numerical modelling and machine learning approaches that conceptualise glacial lake geometry and development trajectories, and quantify current and future water resource trends. The models derived from these data will also provide a scalable solution to robustly estimate dynamic water storage at unsurveyed lakes, therefore reducing the requirement for costly and difficult field surveys. I will also address the critical requirement for high-resolution topographic data to enable robust flood modelling downstream of glacial lakes. These models will identify socio-economic exposure of buildings and infrastructure to flood events caused by precipitation extremes or glacial lake drainage events. The Fellowship's outputs will be operationalised in an online open access Glacial Lake Observatory (GLO) platform that will underpin a new era of collaborative glacial lake research by removing barriers to data access and knowledge exchange. The GLO will catalogue glacial lakes globally and will monitor near-real-time lake dynamics using optical, radar, and altimetry satellite data. Our research culture will advocate for ethical and inclusive overseas fieldwork practices that strengthen partnerships, research collaborations, and knowledge exchange, therefore maximising the long-term benefits of the Fellowship's outputs. Collaboration with leading academics, development agencies, and government departments in Nepal will enable co-production of knowledge that addresses global water resource challenges.
Plain English summaries and category classifications on this site are generated by AI and may not perfectly reflect the original research.
Is something wrong? Let us know