Sensitivity Analysis of Meteorological Parameters in Runoff Modelling Using SWAT (Case Study: Kasillian Watershed)

Determination of river runoff is essential in design and construction of most hydraulic structures including dams. In rivers with no measurement stations, the hydraulic models can be used for data estimation.  SWAT is one of the most widely-used numerical models. In this model, input influential meteorological data as precipitation, temperature, wind speed, solar radiation and relative humidity as well as watershed data including the curve number and roughness coefficient are required to calculate the watershed runoff. The lack of weather stations in some watersheds increase the risk that the registered data in a station do not represent the whole watershed. Consequently, runoff estimation error should be determined. This research evaluates the sensitivity of the river runoff estimation to variations of the meteorological parameters such as precipitation, solar radiation, wind, humidity and temperature using SWAT numerical model. The results indicated that with a 30% decrease in the average monthly precipitation, solar radiation, relative humidity, wind and temperature, a 64.27% decrease, 114.67% increase, 45.93% decrease, 126.12% increase, and 39.21% increase was observed in the modeled runoff, respectively.