Runoff sensitivity of Indian sub-continental river basins

Abstract

Understanding the sensitivity of water availability in the current and future climate in the Indian sub-continent is vital for food and water security. Using the Variable Infiltration Capacity (VIC) model and Budyko's framework with two observational datasets, we estimated water budget and mean annual runoff sensitivity to precipitation and potential evapotranspiration (PET) over 18 major river basins and 222 sub-basins in the Indian sub-continent. The river basins located in the north experienced a decline in mean annual precipitation while the basins in the south witnessed an increase in mean annual precipitation. Declined precipitation and increased PET resulted in a decrease in mean annual runoff in Brahmaputra, Ganga, and Indus basins during 1980–2014. On the other hand, mean annual runoff has increased in Sabarmati, South Coast, Subernarekha, Tapi, Mahanadi, East coast, Cauvery, and Brahmani river basins. Mean annual AET estimated using the Budyko's framework was underestimated while mean annual total runoff was overestimated for the majority of the basins in comparison to the estimates from the VIC model. Moreover, the Budyko's framework with both observational datasets underestimated runoff sensitivity to the changes in precipitation and PET in comparison to the VIC model. Runoff is more sensitive to change in precipitation than PET for the majority of the river basins highlighting the importance of changes in precipitation for water availability in the Indian sub-continent. The VIC model simulated runoff and evapotranspiration are in better agreement with the observations in comparison to the estimates from the Budyko's framework. However, a large uncertainty was found in water budget and runoff sensitivity estimated using the VIC and Budyko's models, which highlights the importance of considering multiple models for estimation of the water budget and runoff sensitivity in the sub-continental river basins.