Climate-driven variability and potential land use change effect on sources of water from Mountain Headwater Systems.

This study estimated the climate and land use change effects on hydrological processes of major Nepalese watersheds including Kali Gandaki, Tamor and Seti using the Soil and Water Assessment Tool. Future climate change effect was modeled using the outputs of temperature and precipitation changes derived from Special Report on Emission Scenarios (B1, A1B & A2) of 16 global circulation models for 2080s. The discharge of the Kali Gandaki River was approximately 39% higher than current values for the maximum temperature and precipitation changes of the A2 scenario and 22% less for minimum changes of the same scenario. With 7% of original forested land removed, sediment yield for the Tamor basin was estimated to be 65% higher, but increased to 124% for the SRES-B1 scenario. For the Seti basin, 4% deforestation yielded 33% more sediment for the SRES-A1B scenario. The effect appears to be geographically important with higher influence in the eastern Tamor basin potentially due to longer and stronger monsoonal period of that area. To better understand the potential groundwater dynamics in mountain stream system and also to better assess the model simulation outputs, I used δ¹⁸O and δ²H of water, solute concentrations, and ²²²Rn to determine which of these may best characterizes water derived from sub-surface sources present in the stream during summer recession flow in McDonald Creek watershed. This study was important since higher groundwater contribution to the stream discharge was assessed in the Himalaya Mountain systems. In main channel of the McDonald Creek, I measured peak ²²²Rn activity of 2,646 Bq/m³ at 7.9 km distance downstream located in middle of the watershed associated with streamflow constriction that corresponded to changes in local orientation of underlying rocks. Based on ²²²Rn mass balance calculations, I estimated that groundwater contributed between 0.3 and 29% of total flow. However, 60% of the total groundwater input was likely to have been produced from hyporheic zone of the stream channel. Finally, I estimated a 5.9% of groundwater contribution integrated for stream reach measured at McDonald Creek during recession flow period.