On the effects of climate downscaling for projecting hydrologic response of catchments in High Mountain Asia

The coarse spatial resolution of Global Climate Models (GCMs) makes the assessment of future hydrological dynamics particularly challenging over complex terrain regions like High Mountain Asia (HMA). Climate downscaling is therefore essential to better understand the impacts of climate change on water resources in HMA. Owing to this need, we investigate the effect of downscaling climate models on the hydrological projections of glacierized catchments in HMA with a specific focus on streamflow projections, relative contribution of streamflow components, and peak flows over five study basins in HMA. For this, four CMIP6 GCMs under SSP5-8.5 scenario were downscaled using two statistical techniques—parametric Cumulative Distribution Function (CDF) matching and Generalized Analog Regression Downscaling (GARD)—and used for streamflow simulations using the Hydrological Model for Distributed Systems. The original GCMs exhibited a wet bias across all basins in the dominant precipitation season and a cold bias (−4.3 °C to −13.6 °C) over the glacier-dominated basin. CDF matching and GARD performed well in reducing the bias from original GCMs over high precipitation seasons. Future precipitation increases over winter and spring seasons for basins in Pakistan and over summer and fall seasons for basins in Nepal and Bhutan. An increase in rainfall-runoff component is anticipated in the future across all basins, while the contribution to streamflow from snowmelt decreases over central and eastern basins. The overall water availability across the basins is projected to increase, along with extreme flows. The results revealed the impact of choice of downscaling techniques and GCMs on the catchment climatology, affecting the hydrological projections.