Effects of climate change on runoff in a representative Himalayan basin assessed through optimal integration of multi-source precipitation data

Abstract

Study region: The Upper Koshi River basin (UKRB), Eastern Nepal, Central Himalayas. Study focus: Climate change is exerting significant effects on the Himalayan region. Understanding the response mechanism of the runoff of the Upper Koshi River, the largest river in Nepal, to climate variability is essential for water resource management. However, all existing precipitation products have some bias in the estimation of precipitation of this region and subsequent runoff modeling. Therefore, this study first utilized a stacking machine learning approach to obtain more reliable precipitation data. Then, the response of runoff to climate change was investigated using the J2000 hydrological model and de-trend analysis. New hydrological insights for this region: The results show that the integrated precipitation data are comparable with existing products in terms of gauging data validation and hydrological model performance. The de-trend analysis highlights the dominance of precipitation changes in driving runoff changes in the UKRB, but with notable spatial variability. Below 4000 m, runoff decline is primarily attributable to reduced precipitation. Within 4000–5000 m, runoff is hindered by warming but promoted by the increase of precipitation. Furthermore, this study establishes a demonstration framework for assessing the impacts of climate change on hydrological regimes in data-scarce alpine basins.

Keywords

• Himalayan glaciated basin
• Precipitation
• Stacking machine learning
• Hydrological modeling
• De-trend analysis