Predicting Hydrologic Effects of Land-Use Change: Problems with the Curve Number Approach

摘要

The Curve Number (CN) approach is ubiquitous in hydrologic model studies of land-use change effects. In this study we illustrate a specific case where the use of CN is misleading, counter-intuitive, and may be inappropriate for design without calibration data. Results compare observed rainfall-runoff data with simulation results from a fully-distributed physics-based Gridded Surface/Subsurface Hydrologic Analysis (GSSHA) model and a lumped CN-based model (HEC-HMS). Results demonstrate how the CN approach fails to accurately simulate an extreme event using standard-practice parameters. The CN approach does perform better with calibration, at the cost of implausible parameter values. The authors argue that given new interfaces and understanding, the Engineering community should use physics-based modeling approaches to overcome the limitations of the CN approach in simulating land-use change effects.