The effect of temperature on forecasting of daily discharge by Fuzzy Based Models in Lighvan watershed

In this research, fuzzy based models were used for river discharge forecasting in lighvan watershed.To forecast the daily river flow, three parameters of the amount of rainfall, temperature, and daily discharge of Lighvanchai basin were used. Assessment of prediction results by using various error estimations methods such as Nash-Sutcliff Coefficient (CNS) showed that ANFIS model (CNS=0.998) was able to predict discharge of these rivers with higher accuracy and low dispersal than FIS model (CNS=0.993). It was also observed that by applying FIS model, daily air temperature affected the river discharge in two months of December (Azar) and April (Farvardin), and the accuracy of the model were increased. It can show the effects of air temperature on the variety of precipitation form in these two months.