Discussion of “Quantification of Curve Numbers for Arid Watersheds”

The runoff curve number (CN) method for estimating runoff from storm rainfall is widely recognized in hydrologic engineering and environmental impact assessments.Its popularity can be attributed to its simplicity and convenience.However, the primary limitation of this approach is its empirical foundation.The underlying physical mechanisms remain a subject of ongoing investigation, and the original paper represents an important contribution to this area.The authors analyzed historical rainfall and runoff data from 1975 to 2021 for two arid to semiarid watersheds, namely, the Santa Rita and Lucky Hills experimental watersheds, and computed the CN based on rainfall events.The authors assumed that the potential maximum retention is equivalent to the cumulative infiltrated water after the beginning of runoff, which was estimated using the Green-Ampt infiltration model.The purpose of this discussion is to highlight certain aspects of the computation that warrant attention and to review the assumptions made by the authors.Specifically, this discussion focuses on the following issues: (1) the authors neglected important preprocessing steps when applying the Hawkins (1993) method to calculate the curve number corresponding to the maximum precipitation, CN ; (2) it is inappropriate to compute CN using only larger rainfall events, as the Hawkins (1993) method requires the use of rainfall events across all frequencies; (3) the assumption made by the authors-that the potential maximum retention equals the cumulative infiltrated water during each rainfall event-has already been proposed in previous studies, with a more reasonable computational approach.The original paper, following the method of Hawkins (1993), computed the curve number corresponding to the maximum precipitation, CN .However, they used natural paired rainfall and runoff data, rather than rank-ordered (i.e., sorted) pairs of rainfall and runoff events.This distinction is emphasized in the first step of the Hawkins (1993) method, which involves ordering the datasorting the rainfall and runoff depths independently in descending order and then rematching them.Since each of the ordered pairs corresponds to the same return period, this approach ensures frequency matching between rainfall and runoff.The CN values