Feasible Ranges of Runoff Curve Numbers for Korean Watersheds Based on the Interior Point Optimization Algorithm

Rainfall runoff is a complex phenomenon in nature. It differs from place to place due to different topographical features and rainfall patterns. The Natural Resources Conservation Service - Curve Number (NRCS-CN) is a well-adopted model to account for direct runoff volume from storm events. There are several studies on determining the initial abstraction and the CN from observed rainfall-runoff data; however, few studies demonstrate their statistical characteristics. The major aim of this study is to determine the feasible range and the confidence intervals of the CN. We examined 660 rainfall-runoff events collected from six medium sized watersheds in South Korea. The interior point optimization algorithm was adopted to ascertain the optimum value of CN and the initial abstraction coefficient (lambda). The obtained results show that the CN value ranged from 45 to 90 and the average lambda = 0.12 was best suited for Korean watersheds. The estimated confidence intervals were highly significant and strongly recommended for Korean watersheds.