Basin hydrologic response relations to distributed physiographic descriptors and climate

The long-term (climatic) hydrologic response of a basin may be quantified u sing the evaporation efficiency (E/E-p or actual evaporation ratio with pot ential evaporation) and runoff ratio (RIP or ratio of runoff loss to precip itation). A key question is the degree to which the basin's physiographic f eatures and regional climate can explain or predict these: hydrologic respo nse measures. In this paper we present the results from 10 basins in diverse climates and terrains. The long-term hydrologic response is estimated using an equilibr ium surface water-groundwater interaction model. We investigate variability between basins with an examination of the relationships between various ph ysical characteristics and the hydrologic properties of basins. Neither cli mate nor physiography alone can explain observed interbasin variability. Si x variables are selected to represent the basins' climate, geomorphology, a nd lithology, each of which has a conceptual relationship to basin-scale eq uilibrium hydrology. The parameters include median slope, relief ratio, dra inage density, wetness ratio, infiltration capacity, and a saturated zone e fficiency index. Two hydrologic variables (runoff ratio and evaporation eff iciency) are selected from the output of a distributed hydrologic equilibri um model. We perform a stepwise regression to identify which combinations o f variables are valuable in predicting the basin-average hydrologic fluxes. A combination of two variables estimate the runoff ratio with an R-2 or ex plained-variance fraction of 0.76: use of ail six variables increases the p rediction to an R-2 of 0.90. The stepwise regression technique fails to ach ieve a statistically significant model for evaporation efficiency, but a re gression model using all six variables nonetheless achieves an R-2 of 0.79. This paper demonstrates that physiographic and climate descriptors can expl ain a large fraction of basin-to-basin differences in modeled hydrologic re sponse. The case has been built on modeled surface water-groundwater intera ction and should next be extended to hydrologic response descriptors derive d from observations alone. (C) 2001 Elsevier Science B.V. All rights reserv ed.