The National Resources Conservation Service (NRCS) Curve Number (CN) method holds immense significance in various critical analyses, including floodplain management, land use planning, and water quality management. The proposed change of the initial abstraction ratio (lambda) to 0.05 from 0.2 in the USDA National Engineering Handbook Part 630 Hydrology indicates uncertainty in the approach to curve number. The objective of this study is to evaluate three approaches to Curve Number calibration and the effects of changing lambda on the accuracy of Curve Number determination. This study utilizes rainfall-runoff datasets from eight locations around the United States representing a wide geographic and climate distribution. Three distinct methodologies, namely least square error (LSE), asymptotic, and the NRCS National Engineering Handbook (NEH) method, are employed for this purpose. The effects of the lambda value and two distinct data ordering approaches (natural and ordered data ordering methods) are also evaluated. Preliminary results have shown that when compared to the NEH (median) CN values, the different approaches underestimate CN for both lambda values and ordering methods. The estimation efficiency (NSE) of the runoff depth does not differ significantly when comparing the combinations of lambda and CN values. CN calibration for forested watersheds showed an unsatisfactory performance only when using natural ordered data. Final results should contribute to the discussion about the need and pertinence of changing the standard value of lambda.
