New Modeling Flags Previously Unrecognized Glacial Lake Flood Risks Across Bhutan

Bhutan counted 567 glacial lakes as of 2021, according to an inventory by the country's National Centre for Hydrology and Meteorology. A new study from Newcastle University has now asked what would happen downstream if some of them burst, and the answer reaches further than earlier assessments suggested.

The work, published in Natural Hazards and Earth System Sciences, is described as the first comprehensive modeling of glacial lake outburst flood risk across Bhutan. These floods, known as GLOFs, begin when warming temperatures cause glaciers to retreat and meltwater pools at the glacier's front. Such lakes can fail suddenly, sending a fast-moving flood down a valley. In some cases the water can travel as far as 400 kilometers (250 miles) from where it started, damaging property, infrastructure and farmland, and sometimes killing people. An earlier study found that Bhutan's glacial lakes have grown in number by 17% since the 1960s.

The team modeled hypothetical outburst scenarios for 278 lakes. Each was larger than 0.05 square kilometers and sat within one kilometer of a glacier terminus. Hydrodynamic simulations tracked how floodwater would move through downstream valleys and how it would meet settlements and farmland. The researchers then linked each lake to the people, buildings and transport networks in the flood's path. Their approach combines exposure and vulnerability in a single risk calculation, two factors that earlier assessments often treated separately or only in part.

The resulting classification placed one lake in the "very high hazard" category, five as "high hazard" and 22 as "moderate hazard." Several of the lakes newly rated high hazard had not been singled out as dangerous in previous studies.

The human and economic totals are substantial. Areas that could fall within a flood zone after an outburst contain more than 11,000 people and more than 2,500 buildings. They also hold more than 250 kilometers (155 miles) of road, roughly 20 square kilometers of farmland and over 400 bridges.

The findings also go beyond individual lakes to administrative regions downstream. The researchers identified five local government areas with very high GLOF risk and eight with high risk. Some of these had not been recognized in national assessments.

Lead researcher Sonam Rinzin, a doctoral student at Newcastle, said that combining flood modeling with mapped infrastructure and population data gives a more detailed picture of potential downstream impacts. He argued that judging risk only by a lake's physical characteristics is less robust than integrating hydrodynamic modeling with socioeconomic data, which he said can help refine and prioritize preparedness planning.

The study sits within a wider pattern. Research from 2023, also led by Newcastle, found that High Mountain Asia, a region that includes Bhutan and stretches from Kyrgyzstan to China, faces the world's highest GLOF danger, with 9.3 million people potentially at risk. The Bhutan work was made possible by satellite data on lake numbers and locations, combined with terrain data, which let the team simulate multiple flood scenarios at national scale.

Professor Rachel Carr, a glaciologist at Newcastle, said the approach could be applied in other mountain regions facing similar hazards. She said future efforts should build on the findings by strengthening monitoring systems and extending mitigation measures to protect lives and livelihoods. The researchers present the results as a basis for prioritizing where monitoring, mitigation and preparedness work should go first in Bhutan.