Statistical analysis of NRCS curve number (NRCS-CN) in arid basins based on historical data

CN method is a commonly used technique for estimation of direct runoff for flood mitigation studies. In Saudi Arabian (SA) arid basins, there is no analysis of NRCS-CN obtained from measured rainfall-runoff events. The research objectives are to find out the actual range of CN values in arid basins, the statistical distribution of CN, the confidence intervals of CNs, and the relation between the CN and initial abstraction factor, λ. Five basins with 19 sub-basins located in the southwest of Saudi Arabia were considered, and 161 rainfall-runoff events were analyzed during the period 1984–1987. The least squares method was used to obtain the optimum range of values for λ and CN based on three estimation methods. The rainfall and runoff exhibited log-normal distribution. The analysis showed that CN varied between 45 and 85 at λ = 0.2. The low CN values account for the transmission losses which is a typical phenomenon in arid regions. The initial abstraction ratio λ = 0.01 is found to be more representative to arid basins rather than λ = 0.2. The Beta distribution is the best to fit CN at both λ = 0.2 and 0.01. The confidence intervals are estimated and tabulated at different significant levels for flood risk assessment studies in arid basins.