Application of Modified SME-CN Method for Predicting Event Runoff and Peak Discharge from a Drained Forest Watershed on the North Carolina Atlantic Coastal Plain

HighlightsProposed new method to assess subsurface and surface runoff from drained forested watershed.Factor Fp in the graphical peak discharge method is less in forested wetland watershed than recommended by USDA.Modified graphical peak discharge method correctly approximated observed peak discharges.Abstract. The NRCS curve number (CN) method is a widely used event-based model for estimating runoff using readily available watershed parameters and rainfall data from upland agricultural catchments. However, there is limited literature on application of the CN method in drained forest systems. This study proposes an application of the modified Sahu-Mishra-Eldho (SME) CN method developed and tested in earlier studies. In this study, the SME method was further modified by redefining the maximum potential retention to assess subsurface drainage and surface runoff, which are parts of total outflow, separately for a pine forest watershed with a high water table soil drained by ditches spaced 100 m apart in coastal North Carolina. Assuming that the measured outflow from the drained watershed was dominated by subsurface drainage, computed event outflow using the modified SME-CN (MSME-CN) model showed good agreement with the observed outflow data (without extreme rainfall events) for the study watershed, yielding a Nash-Sutcliffe coefficient of 0.97, R2 = 0.97, and RMSE = 3.46 mm. Linking the direct runoff from the MSME model into the SCS graphical peak discharge method (GPDM) also improved event peak flow estimates compared to those from the GPDM using SCS-CN based outflow, with calculated RMSE of 11.93 and 31.35 L s-1 and modeling efficiency (EF) of 0.79 and -0.45, respectively. In addition, based on analysis, the wetland factor (Fp) of 0.72 recommended in the GPDM was found to be very large and unsuitable for the study watershed with its high retention capacity. The authors suggest multi-site-year validation of the MSME-CN model, which is sensitive to input parameters such as PET5, P, CN, and a, to gain more confidence in it and the associated GPDM. Keywords: Peak discharge, Pine forest, Poorly drained soil, Potential retention, Subsurface runoff, Surface runoff.