You are here:
The Comparative Accuracy of Two Hydrologic Models in Simulating Warm-Season Runoff for Two Small, Hillslope Catchments
Citation:
Zhang, Y. AND W. Shuster. The Comparative Accuracy of Two Hydrologic Models in Simulating Warm-Season Runoff for Two Small, Hillslope Catchments. JOURNAL OF AMERICAN WATER RESOURCES ASSOCIATION. American Water Resources Association, Middleburg, VA, 50(2):434-447, (2014).
Impact/Purpose:
Runoff prediction is a cornerstone of water resources planning, and therefore modeling performance is a key issue. This paper investigates the comparative advantages of conceptual versus process- based models in predicting warm season runoff for upland, low-yield micro-catchments on a continuous basis. Employed in this study are EPA Storm Water Management Model (SWMM), a semi-conceptual model, and Gridded Surface-Subsurface Hydrologic Analysis (GSSHA), a process-based model that concurrently simulates soil moisture and runoff.
Description:
Runoff prediction is a cornerstone of water resources planning, and therefore modeling performance is a key issue. This paper investigates the comparative advantages of conceptual versus process- based models in predicting warm season runoff for upland, low-yield micro-catchments on a continuous basis. Employed in this study are EPA Storm Water Management Model (SWMM), a semi-conceptual model, and Gridded Surface-Subsurface Hydrologic Analysis (GSSHA), a process-based model that concurrently simulates soil moisture and runoff. Two numerical experiments are conducted using each model wherein warm season water budget and runoff response are simulated for two micro- catchments near Coshocton, OH for every year in 1990-2003 less 1994. The first experiment entails no calibration and relies exclusively on parameter values taken from literature. The second experiment mirrors the first but uses boundary values via calibrating respective models against the runoff record of 1990. The results of these experiments indicate that, SWMM, with lesser detail in representation, provides runoff estimates closer to observed than GSSHA. Yet, the predictions from both models depart considerably from observations. Neither model, even with calibration, is able to reproduce the contrasts in runoff response characteristics during the spring and summer, and between the two watersheds. These findings further suggest that further efforts are required for effectively resolving the season-dependent effects of vegetation, and the spatial-temporal distribution of soil moisture to allow accurate and physically meaningful prediction of runoff. Meanwhile, lumped, semi-conceptual models such as SWMM are perhaps more suitable for predicting runoff for these small watersheds when calibrated on a seasonal basis. The implications of this work for runoff volume management are discussed.
