Application of statistical downscaling models to assess climate change impacts on the East Rapti River Basin using the Rx5day index

A landlocked, mountains-dominated country, Nepal is one such country that is highly susceptible to climate change. The East Rapti River Basin (ERRB), a sensitive corridor, is under extreme threats due to flooding as well as rainfall variability. The study aims to evaluate the performance of the Statistical Downscaling Model (SDSM) in simulating historical rainfall and to project future changes in extreme rainfall intensity (Rx5day) and inter-seasonal variability within the ERRB. The SDSM was first calibrated and validated (1980–2005) using NCEP/NCAR reanalysis predictors to establish local-to-regional relationships. Subsequently, the validated model was forced with CanESM2 GCM outputs to project future rainfall series under RCP 4.5 and RCP 8.5 scenarios for the 21st century. In validation, the SDSM yielded a Nash-Sutcliffe Efficiency (NSE) of 0.605, classified as "good" model performance. Projections indicate a substantial rise in the dry-season extreme rainfall: Q1 Rx5day values for the January–March period will increase from a historical 12.5 mm to 64.5 mm by the 2080s under RCP 4.5. While Q3 monsoon intensity remains dominant (260.9 mm historical vs. ~236 mm projected), a significant decrease in extreme intensity is projected for Q2. There will be a significant shift to an oscillating hydrological cycle with high "off-season" multi-day peaks because of climate change. Overall, the results indicate that traditional drainage system designs and mitigation practices might not be effective and that effective agricultural practices and flood-resilient crops should be implemented in the affected regions based on climate change projections.