Modeling Hydroclimatic Processes in Large Watersheds.

Spatial and temporal differences in the processes that transform precipitation into runoff are a common source of error in watershed models. For large watersheds, spatial disaggregation of the drainage area can improve efforts to model the precipitation-runoff relationship when the model employs physically explainable variables to represent horizontal and vertical variations in moisture fluxes and storages. The paper presents the conceptual structure of a hydroclimatic model that simulates the runoff process as four functionally discrete subsystems linked in series. Moisture inputs are routed through each subsystem employing environmentally dictated rules and readily available data. Dividing the basin into homogeneous subunits addresses the problem of horizontal heterogeneity. Residence times for two groundwater storage subsystems are calculated by an optimization algorithm included in the model. Modeled hydroclimatic characteristics for the Deschutes River Basin in Oregon illustrate the ability of the model to reveal important spatial and temporal differences in the runoff process. Sectors representing 20 percent of the basin area contribute 57 percent of modeled groundwater recharge and 61 percent of total modeled runoff for the watershed, but some other sectors make a relatively small quantitative contribution to basin runoff. Moisture exchange processes in the basin display explicit seasonal characteristics that explain the surprisingly large proportion of annual precipitation allocated to groundwater recharge and the dampening of the seasonal moisture input by the groundwater storages. This knowledge of the basin runoff process provides a rational basis for formulating water management strategies for the watershed. [ABSTRACT FROM AUTHOR]