| Winter-time hydrology plays many important roles in the over-all hydrology of many watersheds. Soil frost and snow are two elements unique to winter-time hydrology, both of which have been well modeled for a point but have seen little attention in the arena of watershed modeling. One of the major difficulties with modeling winter-time hydrology across a watershed is the problem of extreme spatial variability.; The primary objective of this project was to explore the possibility of utilizing Geographic Information System technology to model winter-time hydrology. The Winter-time Hydrology Model (WHYM) was developed to facilitate this objective. The processes simulated by WHYM were: soil freeze and thaw, snow accumulation, melt, and drift, and infiltration, runoff, and soil water balance. Another objective of this project was to design WHYM to be a practical model with as few inputs as possible that would run on a PC.; WHYM was corroborated using three data sets, two from data collected near Pullman, WA and one from Danville, VT. Snow processes generally showed good agreement between simulated and observed values. Snow drift results were the poorest of the modeled snow processes, but considering there has been little work toward modeling this process at the landscape level the results were encouraging. Soil frost was well simulated for deep, persistent frost conditions and poorly simulated for shallow, intermittent frost situations. The soil frost problems were primarily due to the simplifications made in modeling soil frost penetration which are not valid under shallow frost conditions. There were also errors associated with erroneous snow drifting estimates which over-estimated the insulative effects of the snow. The runoff and infiltration components of WHYM were not fully developed to account for a broad range of conditions, but within their current design limitations appeared adequate for handling winter-time conditions. The soil water balance performed adequately. The biggest problem WHYM faced was in unacceptably long run-times. |