Comparative diagnostic analysis of runoff generation processes in Oklahoma DMIP2 basins: The Blue River and the Illinois River

Abstract

Summary This paper presents the results of a comparative diagnostic study of runoff generation processes in two test basins in Oklahoma: The Blue River at Blue and the Illinois River near Tahlequah. This study involves analysis of signatures of spatio-temporal runoff variability, extracted from both observed rainfall–runoff data and from predictions of a distributed, physically based rainfall–runoff model. Analysis of observed data in both basins indicates that event runoff coefficients are systematically higher in the wet season than in the dry season. Model predictions indicate that the transition from high to low runoff coefficients in the Blue River basin is linked to variations of water table depth and surface soil moisture, contributing to a seasonal switching of surface runoff generation mechanisms, from saturation excess to infiltration excess. In the Illinois River basin, however, due to more permeable soils, infiltration excess runoff occurs rarely. The differences in intra-annual patterns of runoff coefficients and runoff generation mechanisms can be partly explained by the seasonality of climate forcing and water table position. Despite the significant differences of runoff generation mechanisms between the two basins, spatial analysis of the model results reveals that in both watersheds, but especially so in the more humid Illinois River basin, saturation excess runoff and subsurface stormflow coexist in competition throughout the year. This competition is quantitatively shown to be controlled by the relative magnitudes of the saturated hydraulic conductivity of the soils and the topographic slope. In addition, the spatial variabilities of runoff generation processes also impact the spatial scaling behavior of runoff ratios, indicating the existence of a threshold watershed size beyond which the variability is averaged out.