ASSESSMENT OF WATER RESOURCES IN DUDH KOSHI RIVER BASIN AND IMPACTS OF FUTURE CLIMATE SCENARIOS

Abstract

Dudh Koshi river originates from the Himalayas of Mount Everest and other peaks of Eastern Nepal and joins with Sunkoshi to form a giant Saptakoshi, which is the major tributary of the Ganga river in India. The multiple tributaries originated from the high Himalaya, cross the mountains and hills in deep gorges and flowing south through the Middle Hills to form Dudh Koshi. The total basin area of Dudh Koshi river is 4018 km2 at the confluence point with Saptakoshiriver, whereas the basin considered in this study is 3674 Km2. Although a majority of the precipitation for this region is concentrated during the monsoon months, the snow covered and glaciated upper reaches of the basin contribute melt water to the streamflow throughout the year. Furthermore, other spring sources in the hilly and mountain terrains provide considerable perennial streamflow in the basin. The river is blessed with very good power generation potential in hilly regions and irrigation potentiality to command the substantial agricultural area by inter-basin water transfer to the Terai region. The hydrology of the Himalayan river is supposed to be more sensitive to climate change due to the effects of the seasonal, latitudinal and altitudinal shift in the freezing line. Rising air temperature, snow melting, irregular rainfall, desertification are key indicators for climate change within the basin. The basic influences of climate variability on the intensity and frequency of droughts and floods are likely to increase water and flood management changes. In this study, hydrological modelling of Dudh Koshi River Basin (DKRB) using physically based, spatially distributed, a continuous model known as ‘Soil and Water Assessment Tool’ (SWAT) for the assessment of water availability and impacts of climate change in the basin has been envisaged. The key objectives of this study were to set up a suitable model to describe the hydrology of the DKRB using the SWAT model, to investigate the rainfall-runoff relationship by simulating the regular flow using the hydrological model under various climate change scenarios, and to determine the water availability in the Dudh koshi river basin under various climate change scenarios. Hydrological and meteorological data were obtained from DHM, GWD and SWAT auto simulation. DEM-SRTM was downloaded from USGS, whereas Soil map and LULC were downloaded from the website of ICIMOD, Nepal. The SWAT model was applied on a daily and monthly time scale and calibrated and validated to observed discharge at Outlet O-18 for 1982 2000 and 2001-2006, respectively, using the SUFI-2 technique in the SWAT CUP program. The performance of the model was evaluated using various model performance criteria reported in the literature, such as coefficient of determination (R Square), Nash-Sutcliffe Efficiency (NSE), Percent bias (PBIAS). The results obtained during the model calibration (R2 = 0.67, NSE =0.65 and PBIAS=-8.5) and validation (R2 =0.65, NSE =0.63 and PBIAS=16.2) showed a reliable iii estimate of daily streamflow, which shows the SWAT model performance was ‘good’. The p factor for the calibration and validation period were found 0.41 and 0.27, respectively, whereas the r-factor for those periods were 0.12 and 0.08, respectively. But, the results obtained from the model calibration and validation showed a reliable estimate of monthly streamflow (R2 = 0.84, NSE =0.83 and PBIAS=0.8) for calibration period and (R2 =0.84, NSE =0.75 and PBIAS=23.7), which shows the SWAT model performance is ‘very good’. The p-factor for both calibration and validation period was 0.38 and 0.25, whereas the r-factor for those periods were 0.08 and 0.06, respectively. Though the result was good on the daily time scale, SWAT underestimated the streamflow during extreme flow periods, which might be because the curve number technique used could not predict accurate runoff for a day that experienced several storms.