SWAT MODELING FOR THE PISHIN LORA BASIN, BALOCHISTAN, PAKISTAN

The Balochistan Province in Pakistan is largely an arid desert (~100 mm/yr), is one of the poorest and most dangerous provinces, where political instabilities and water shortages are giving rise to local and regional disputes. Afgani war immigrants and decreased precipitation over the past decade have contributed to a dramatic rise in the local population of the Quetta valley, part of the Pishin Lora Watershed (16,671 km²) encompassing areas in Pakistan and Afganistan. Given the difficulties in accessing the region, and the paucity of field data, we adopted methodologies that heavily rely on readily available remote sensing technologies as viable alternatives and useful tools for the assessment and management of the water resources of these remote regions. The Soil and Water Assessment Tool (SWAT) was selected for continuous (1998-2007) rainfall-runoff modeling of the Pishin Lora watershed. Because only one Quetta rain gauge station data is available in and around the Pishin Lora Watershed, we used satellite-based, TRMM 3-hourly precipitation data that is available for every 0.25° x 0.25° (1998 till present). Additional model inputs include: (1) Soil types from geologic maps; (2) Land use types (Pakistani Agricultural Department); (3) Hydraulic parameters (published data); (4) Elevation data (SRTM); (5) Meteorological data (e.g., solar radiation, temperature, humidity) from the Quetta climatic station. Model calibration for the period 1998 to 2005 was attained by comparing simulated flow to that reported from the the Burj Aziz Khan station (Lat: 30⁰20’; Long: 66⁰35’). Preliminary findings reveal a good correlation (RMSE: 0.71; Nash-Sutcliffe efficiency: 0.72) between observed and simulated stream flow values at the Burji Aziz Khan station. First order estimates for average annual precipitation (1998-2007), runoff, and recharge are 2,190x10⁶ m³, 58x10⁶ m³, 141x10⁶ m³, respectively. Attempts are under way to further refine our initial calibration by adjusting sensitive model parameters (e.g., SWAT snow and flow parameters) and by comparing simulated flow to that observed at two additional stations (Beleli Nala: Lat: 30⁰16’, Long: 66⁰54’; Sariab Lora: Lat: 30⁰11’, Long: 66⁰58’).