The Upper Indus River Basin (UIB). Understanding long-term hydrological responses in High Mountain Asia remains challenging under climate change due to complex topography, heterogeneous cryospheric processes, and limited in-situ observations. This study applied the Geomorphology-Based Hydrological Model (GBHM), a fully distributed process-based hydrological model tailored for cryosphere-dominated catchments to simulate the long-term hydrological processes in the UIB during 1951–2019, particularly the spatio-temporal water flux partitioning shaped by topography and elevation-dependent climatic influences. The UIB received a mean annual precipitation of 718 mm, with annual runoff (513 mm) and evapotranspiration (191 mm) as the dominant fluxes. Snowmelt and glacier melt together contributed approximately 65 % of total runoff, highlighting the basin’s strong reliance on cryospheric sources. Trend analysis during 1951–2019 revealed statistically significant increases in annual precipitation (+1.39 mm yr⁻¹), temperature (+0.013 °C yr⁻¹), evapotranspiration (+0.22 mm yr⁻¹), and runoff (+1.22 mm yr⁻¹). Seasonal decomposition indicated an earlier onset of snowmelt in recent decades, with spring runoff dominated by snowmelt while glacier melt contributing ∼15 % during summer. Elevation-based analysis showed that runoff generation peaked near 5000 m and declined in higher glacierized zones, where increased temperature sensitivity was offset by reduced hydrological responsiveness. The modeling framework and results provide a foundation for improved hydrological projections and sustainable water management under future climatic and cryospheric change. • Process-based hydrological modeling was employed in the Upper Indus Basin. • Long-term hydrological dynamics over nearly seven decades were examined. • A clear intensification of the hydrological cycle was observed in the study area. • Snowmelt and glacier melt together contributed ∼65 % of total annual river runoff. • Elevation-dependent hydrological responses were identified across four zones.
Umer et al. (2026) studied this question.