Runoff Estimation Using the NRCS Slope-Adjusted Curve Number in Mountainous Watersheds

In mountainous watersheds, rainfall generates runoff quickly because of the decreased depression storage, high downslope flow velocity, and smaller chance for rainwater infiltration. In order to obtain precise event-based runoff estimations in mountainous watersheds, a slope-adjusted curve number (${\mathrm{CN}}_{\mathrm{II}\alpha }$) with a smaller initial abstraction ratio ($\lambda$) is indispensable in the standard natural resources conservation service (NRCS) curve number (CN) model. Using measured rainfall-runoff data from 39 mountainous watersheds in South Korea, this study investigated two existing ${\mathrm{CN}}_{\mathrm{II}\alpha }$ approaches and suggested a new approach that was accompanied by a lower $\lambda$ value. The new ${\mathrm{CN}}_{\mathrm{II}\alpha }$ equation was calibrated with 1,402 measured rainfall-runoff events from 31 watersheds and validated with 377 rainfall-runoff events from the remaining eight watersheds. Most of the runoff events, used for both calibration and validation, were underestimated using a CN without a slope-adjusting factor. By considering the combined effect of the proposed ${\mathrm{CN}}_{\mathrm{II}\alpha }$ and setting $\lambda$ equal to 0.01, the performance measures based on the root mean squares error (in mm), Nash-Sutcliffe efficiency, and coefficient of determination were significantly improved from averages of 27.10, 0.64, and 0.75 to 18.69, 0.82, and 0.87, respectively, as compared to the standard NRCS model. The proposed modification exhibited superior results compared to the two existing ${\mathrm{CN}}_{\mathrm{II}\alpha }$ approaches. Findings from this study support the adjustment of both the CN and $\lambda$ in the NRCS model to increase its runoff prediction capabilities.