Climate change poses significant threats to mountain ecosystems in North America (Barnett et al., 2005) and will subsequently impact water supply for human and ecosystem use. To assess these threats, we must have an understanding of the local variability in hydrometeorological conditions over the mountains. This thesis describes the continued development and application of a fine scale spatial hydrometeorological model, GENESYS (GENerate Earth SYstems Science input). The GENESYS model successfully simulated daily snowpack values for a 10 year trial period and annual runoff volumes for a thirty year period. Based on the results of these simulations the model was applied to estimate potential changes in snowpack over the St. Mary River watershed, Montana. GCM derived future climate scenarios were applied, representing a range of emissions controls and applied to perturb the 1961-90 climate record using the “delta” downscaling technique. The effects of these changes in climate were assessed for thirty year time slices centered on 2020s, 2050s, and 2080s. The GENESYS simulations of future climate showed that mountain snowpack was highly vulnerable to changes in temperature and to a lesser degree precipitation. A seasonal shift to an earlier onset of spring melt and an increase in the ratio of rain to snow occurred under all climate change scenarios. Results of mean and maximum snowpack were more variable and appeared to be highly dependent on scenario selection. The results demonstrated that although annual volume of available water from snowpack may increase, the seasonal distribution of available water may be significantly altered. / viii, 93 leaves ; 29 cm

A modelling approach focused on snow hydrology was developed and applied to project future changes in spring streamflow volumes in the St. Mary River headwaters basin, Montana. A spatially distributed, physically-based, hydrometeorological and snow mass balance model was refined and used to produce snow water equivalent (SWE) and rainfall surfaces for the study watershed. Snowmelt runoff (SR) and effective rainfall runoff (RR) volumes were compiled for the 1961-2004 historical period. A statistical regression model was developed linking spring streamflow volume (QS) at Babb, Montana to the SR and RR modelled data. The modelling results indicated that SR explained 70% of the variability in QS while RR explained another 9%. The model was applied to climate change scenarios representing the expected range of future change to produce annual QS for the period 2010-2099. Compared to the base period (1961-1990), average QS change ranged from -3% to -12% for the 2020s period. Percent changes increased to between -25% and -32% for the 2050s, and -38% and -55% for the 2080s. Decreases in QS also accompanied substantial advances in the onset of spring snowmelt. Whereas the spring pulse onset on average occurred on April 8 for the base period, it occurred 36 to 50 days earlier during the 2080s. The findings suggest that increasing precipitation will not compensate for the effects of increasing temperature in watershed SWE and associated spring runoff generation. There are implications for stakeholder interests related to ecosystems, the irrigation industry, and recreation. / xii, 136 leaves : ill. ; 28 cm. --