Abstract

The hydrological sensitivity of snowmelt-dominated, high mountain headwaters to climate change was investigated using a physically based model to diagnose snow processes and headwater basin runoff response to perturbations of the current climate in three well-instrumented mountain research basins spanning the northern North American Cordillera. High-resolution hourly meteorological observations were perturbed using air temperature increases and precipitation changes and then used to force comprehensive, mountain hydrological models created using the modular, process-based Cold Regions Hydrological Modelling Platform (CRHM) for each basin. Simulations using multiple elevations show that both peak snowpack and annual runoff respond to warming and precipitation changes and these responses vary with latitude. In all three basins, the timing and magnitude of peak snowpack were sensitive to changes in temperature and precipitation, but timing was most sensitive to temperature. Annual runoff was far less sensitive to temperature than the snow regime. The impacts of the range of warming expected from North American climate model simulations on annual runoff, but not peak snowpack, can be offset by the size of precipitation increases projected for the future period 2041-2070. To offset the impact of 2°C warming on annual runoff, precipitation would need to increase by less than 5% in all three basins. To offset the impact of 2°C warming on peak snowpack, however, precipitation would need to increase by 12% in Wolf Creek-Yukon Territory, 18% in Marmot Creek-Canadian Rockies and an amount greater than the maximum projected at Reynolds Mountain-Idaho. The role of increased precipitation as a compensator for the impact of warming on mountain snow hydrology is more effective at the high elevations and high latitudes. Increased precipitation leads to resilient and strongly coupled snow and hydrological regimes in cold regions and sensitive and weakly coupled regimes in the low elevations and temperate climate zones.