Assessment of the soil conservation service–curve number method performance in a tropical Oxisol watershed

Authors: ALVES, G.J. - Federal University Of Lavras; MELLO, C. - Federal University Of Lavras; BESKOW, S. - Federal University Of Pelotas; JUNQUEIRA, J.A. - Federal University Of Minas Gerais; Nearing, Mark

Submitted to: Journal of Soil and Water Conservation
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 2/21/2019
Publication Date: 9/10/2019
Citation: Alves, G., Mello, C., Beskow, S., Junqueira, J., Nearing, M.A. 2019. Assessment of the soil conservation service–curve number method performance in a tropical Oxisol watershed. Journal of Soil and Water Conservation. 74(5):500-512. https://doi.org/10.2489/jswc.74.5.500.
DOI: https://doi.org/10.2489/jswc.74.5.500

Interpretive Summary: The USDA Curve Number method is used in the United States and around the world for estimating the volume of surface water runoff generated by rainfall. It is important for many applications, including flood forecasting, design of infrastructure such as roads and culverts, and for water conservation planning. In this study data from a watershed in southern Brazil was used to assess various techniques for applying the method. The study found that the use of tabulated values from the USDA-NRCS handbook did not perform particularly well in predicting runoff from individual events when applied to this type of watershed. Recommendations were presented for obtaining better equation parameter values in order to better use this method in these types of environments. It should have implications for use of the equation outside of the United States and within the country where tropical soils exist and data are scarce, such as Puerto Rico or Hawaii.

Technical Abstract: The SCS-CN method (Curve Number – Soil Conservation Service) is a rainfall-surface runoff model based on an empirical approach between rainfall (P) and ground conditions (soils, management and moisture) to estimate direct surface runoff (Q). The model can be applied using published tables of CN for each combination of soil hydrology, soil use and management, and the last 5-day rainfall occurrence. However, its performance is increased when observed rainfall-surface runoff events are used for model calibration. The CN approach generally uses a fixed ratio, ', between initial rainfall abstraction and the soil-water storage potential equal to a factor of 0.2. In this study, this coefficient was evaluated using ten methods for CN characterization in a tropical Oxisol watershed, based on both CN values published by NEH-4 and 166 selected rainfall-surface runoff events, and using ' values of 0.2 from the standard approach and 0.02, which was obtained from the 166 observed events. The results demonstrated that a fixed CN for representing the entire watershed gave poor performance even when antecedent soil moisture was considered. In contrast, the approach based on calibration to the observed hydrological events gave good results, based on the statistics. The use of ' = 0.02 rather than 0.2 resulted in improvements in direct surface runoff estimates for this landscape. Using a spatially-distributed CN model, considering the heterogeneity of the watershed, ' = 0.02 increased the model estimates of runoff, but resulted in a greater level of fit with the data compared to the use of ' = 0.2.