Assessment and Calibration of the Curve Number Method (SCS-CN) for Estimating Runoff in the Aras Sub-basins, North West of Iran

The SCS-curve number is one of the common methods to estimate drirect runoff in the watershed scale. There is often an uncertainty in predicting runoff in many areas due to to the model’s assumption on the initial abstraction ratio (l= 0.20). This study was conducted to evaluate the accuracy of SCS-CN method in estimating runoff in some semi-arid subbasins of the Aras river consits of the Livar, Kalaleh and Shekaralichay in northwest of Iran. Runoff values were obtained from hydrometery stations for a 30-year period from 1987-2017) and compared with the estimated values which were determined using the maximum retention capacity (S). Based on the results, mean annual runoff in the Livar, Kalaleh and Shekaralichay subbasins is is 6.4, 1.45 and 9.5 mm, respectively. The model accuracy of the Livar, Kalaleh and Shekaralichay was -0.44, -0.25 and -0.27 percent, respectively. It showed over-estimations for small runoff values and down-estimation for large values. The model was calibrated usingthe revision of the initial abstraction ratio value (l= 0.08) and the model accuracy improved to 95, 96 and 93 percent, respectively. As a conclution, the calibration of l is essential to acuurate prediction of runoff in the the Aras river and similar basins.