Answers are grounded in this paper and cite their sources.
Hydrologic modeling study reveals higher predictive accuracy for SCS-CN than Green-Ampt in a semi-arid basin, indicating limitations of physically based methods under blind validation.
Pronounced spatio-temporal variability of rainfall in semi-arid basins remains a central challenge for rainfall–infiltration–runoff modeling, particularly in ungauged or newly instrumented catchments where continuous soil-moisture data are unavailable. This study evaluates the comparative predictive performance of the SCS-CN and Green-Ampt (GA) infiltration methods, implemented within HEC-HMS at the sub-basin scale, using eight rainfall–runoff analysis time windows recorded during the 2020–2025 rainy seasons within a monitoring network operational since October 2019 in an instrumented semi-arid basin in Mexico. A blind-validation framework was adopted: parameters for both methods were derived a priori from tabulated sources indexed by land use, hydrologic soil group, soil textural class, and locally supported by textural analysis at three depths per sub-basin, in situ testing of saturated hydraulic conductivity, and gravimetric determination of field capacity; the initial moisture content required by GA was set equal to the measured field capacity (θi = θfc) to equate initial conditions between the two methods. Spatially distributed rainfall was captured by four monitoring stations under a one-to-one gauge–sub-basin assignment scheme, with monthly rainfall depth varying from 22.8 to 204.9 mm across the four sub-basins. Both methods reproduced observed discharge with varying levels of agreement: SCS-CN yielded very good performance (Pearson R = 0.95; Nash–Sutcliffe efficiency NSE = 0.76), whereas Green-Ampt yielded moderate correlation but unsatisfactory NSE (R = 0.70; NSE = 0.45) against the Levelogger records. Contrary to the initial expectation that the physically based GA would outperform SCS-CN, SCS-CN yielded substantially higher performance across windows, with the two simulated discharge series differing by a mean absolute deviation of 36.9 m3/s. A systematic sensitivity analysis (±6%, ±10%, ±20% perturbations) revealed an asymmetric response: SCS-CN was highly sensitive to Curve Number perturbations (mean-deviation amplitude 65.9 m3/s), whereas Green-Ampt was nearly insensitive to its compound soil-hydraulic parameterization (amplitude 3.2 m3/s), indicating a structural limitation of the physically based method under blind validation.
No takes yet. Share an insight, caveat, or question.
Lara et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: