Water Availability of Soan River Basin under Climate Change using Swat Model

Abstract

Stream flow estimation under changing climate and land use practices is an immense challenge. This study estimated the runoff of Soan river basin using a semi distributed hydrological model SWAT. Sentinel-2 landuse imagery was selected for unsupervised classification of study area via ENVI model. ALOS PALSAR 12.5m resolution DEM was used for slope calculation and generation of stream networks. The hydrological models were calibrated by using observed stream flows of 3 years (2013–2015), while validated for 3 years (2016–2018). Overall, the simulated stream flow results showed that the performance of SWAT was acceptable, as depicted by Nash–Sutcliffe coefficient (NS) and coefficient of determination (R2) of 0.52 and 0.801for calibration period and R2of 0.66 and NS of 0.87 for validation periods. Further, the potential stream flows during decades of 2025s, 2055s and 2075s were projected for RCPs scenarios 4.5 and 8.5 for Access0.1, BCC-CMS, CMCC-CMS and INMCM4 models. Firstly biased corrected climatic data showed significant increase in precipitation and maximum temperature values station wise for RCP 4.5 and RCP 8.5especially for Islamabad and Attock station where this increase is of 5⁰C to 10⁰C. Due to increase in value of minimum temperature for Murree and Khanpur, a less cold winter is expected. However, there was a decrease in value of minimum temperature for Attock, Islamabad and Juharabad in contrast to other stations therefore a colder winter expected for these stations. Secondly, increasing trends of stream flows were found in consistent with the climatic dataset for three decades (2025s, 2055s and 2075s). A CMCC-CMS model, as compared to other models used, depicts a high amount of flow during post summer season for all three decades. Access0.1 model trajectories were close to base time flow. However two flood peaks were presented by BCC-CMS and CMCC-CMS model for July and august month. BCC-CMS model showed moderate increase (i.e. ≤ 500 cumec) in peak flows for post summer. There observed a slight shift of flow peaks post summer towards the start of summer. Post winter flood peaks with increasing trend shown by all models for both RCPs scenarios 4.5 & 8.5. Overall this study reveals in increase in stream flow for Soan river basin.