Modelling climate in the world’s highest mountains
The Hindu Kush Himalaya is one of the most climate-sensitive regions on Earth, yet its steep terrain makes it one of the hardest to model. Global models are too coarse to resolve individual valleys and ridges, and observations at high altitude are scarce. My work uses a range of models, from global to kilometre scale, and tests them against the observations we do have.
Global and regional climate models
During my PhD at IIT Roorkee, I set up and ran the regional climate model RegCM4.7 and NASA’s GISS ModelE2.1 general circulation model on the PARAM Ganga supercomputer to study black carbon and its radiative impacts over the Hindu Kush Himalaya. I also evaluated how well CMIP5 and CMIP6 models reproduce observed precipitation and temperature trends over Himalayan sub-regions (1981–2014) compared with ERA5 reanalysis.
High-resolution weather modelling
At the kilometre scale, I have used the WRF model to reconstruct the meteorological conditions around the 7 February 2021 Chamoli rock-ice avalanche in Uttarakhand. The simulations showed an unusual drop in temperature and a rainfall event in the days before the disaster, signals that could be useful for anticipating similar hazards.
Observing the extremes
Models need data from the places they struggle with most. I contributed to observations from the world’s highest weather stations near the summit of Mount Everest, which reveal extreme wind chill, rapid frostbite risk and strong sublimation, and help constrain how models represent high-altitude climate.
Tools: RegCM4.7, GISS ModelE2.1, WRF, CMIP5/CMIP6 output, ERA5, SBDART radiative transfer, Python and MATLAB on HPC systems.