One of the most robust impacts of a disrupted climate is the enhanced melting of snow with warmer temperatures. In mountainous areas with snow dominated hydrology the increased proportion of winter precipitation falling as rain instead of snow and the earlier melt of the snow pack have important consequences on the reliability of managed water systems. The earlier snow melt caused by warmer temperatures is accelerated at the interface of snow-covered and snow-free areas by the snow-albedo feedback, where newly exposed bare ground with a low albedo increases sensible heat transfer to the air and the adjacent snow pack. The simulation of a snow-albedo feedback and its evolution in a changed climate requires coordinated simulation of both the near surface atmosphere and the surface hydrology, usually achieved with a land surface representation in a computationally expensive coupled atmospheric model. In this study, we investigate the detectability, magnitude, and significance of the snow albedo feedback on surface runoff at different spatial scales, and the ability to capture the feedback using dynamical downscaling and different statistical downscaling methods informed by coupled model simulations and gridded observations. At the spatial scale of larger rivers draining California’s Sierra Nevada and supplying the reservoirs feeding the State’s managed water system, the ability to represent snow albedo feedback is not strongly evident in the simulated hydrology or projected changes in streamflow. When the focus is placed on smaller areas at elevations near the snow line its representation becomes more important for hydrologic projections.