Watershed-scale hydrology: Influences of spatial variability of snowpack in a semiarid, mountainous environment

Water is a precious resource in the semiarid western U.S. rangeland. In order to effectively develop and allocate water, one needs to accurately know the volume and timing of streamflow from these watersheds. This study used a combination of extensive data analysis and modeling to understand the nature of annual hydrologic mass balance at a range of spatial scales within Reynolds Creek Experimental Watershed (RCEW). The study is presented as a collection of three papers, which focus on building a working hypothesis at a small scale, and then testing its transferability to a larger scale.; The dominant hydrologic process in RCEW is the highly spatially variable surface water input (SWI), a result of wind-induced snow drifting combined with variability of air temperature with elevation and spatial variability of net radiation due to terrain effects. The drift factor approach was used to parameterize wind-induced snow drifting. A quantitative basis was established for subdividing watersheds into modeling elements based on the distribution of the drift factors. The working hypothesis for the first-order watershed is that most of the runoff is generated by SWI into the deep snowdrift zone located on the north-facing leeward slope, whereas the SWI on the rest of the watershed is used mainly to satisfy evapotranspiration and subsurface storage demands.; It is impractical to conduct manual snow surveys at appreciably sized watersheds. Instead, a physically based, blowing snow model was used to simulate snow drift. Its performance was tested against manually surveyed snow water equivalence maps and the calibrated drift factor map at the first-order watershed. The simulated pattern of snow accumulation did not agree well with observations in a pointwise comparison. It was found, however, that drift factors obtained from the blowing snow model can be used to parameterize the distribution of the drift pattern within an error bound of 25% if the variability of precipitation between measurement sites is of the order of 70%.; The understanding of annual hydrologic mass balance was transferred to a larger, fourth-order watershed within RCEW. Parameters of the above-surface, SWI model calibrated at the first-order watershed could be transferred to the fourth-order watershed. Parameters of the below-surface, hydrologic mass balance model calibrated at the first-order watershed also transferred to the fourth-order watershed with minor modifications. Annual mass balance was found sensitive to two parameters and the initial conditions.