عنوان مقاله [English]
The study aimed to evaluate the regional models in estimation of probable annual water yield in hydrologic units of Karkheh, Dez and Karun watersheds, located in south-west of Iran. Afterward, observational data of discharge and precipitation were used from 108 hydrometeric stations and 208 rain gauges. Area, equivalent rectangle length, slope, Gravilios coefficient, precipitation, main stream length, main stream slope, mean watershed elevation, drainage density and perimeter were accounted in regional analysis. The cluster analysis and discriminate function was used in homogeneity analysis of hydrologic units. The results of multiple regression showed a high correlation between observed and computed data (0.98 and 0.99). It was also found out that annual discharge had a direct correlation with the first four above variables, and an inverse correlation with the other factors. The results showed that the annual yield can be precisely predicted by the area. Finally, the results showed that the annual depth of runoff yield over watershed had a high correlation with area and precipitation in a bi-variable regression model with regression coefficients of 0.52 to 0.8.