Water Resources Planning under Non-Stationary Hydroclimate in a Snow Dominant Watershed

Milly et al. in a recent article (Science, Vol319, 1February, 2008, pp573–574) declared that "stationarity is dead and should no longer serve as a central, default assumption in water-resources risk assessment and planning." They went on stating, "Finding a suitable successor is crucial for human adaptation to changing climate." Noting that the fundamental cause of this climatic change is the warming temperature and further noting no apparent change in the runoff volume for the last century, the authors hypothesize that a temperature based hydrology can explain most of the non-stationary behavior of the runoff in a snow dominant watershed like the Upper Feather Basin in California. The observation of the historical temperature data suggests that this region is warming consistent with the global trend. The projections of precipitation by various GCM's are wide spread and uncertainties on the wetness (or dryness) abound whereas the future temperature projections, through also widely spread, are unanimous in directional sense—going up or getting warmer over time. Noting this robust nature of the future temperature projections and also noting that the cause of the future precipitation changes between rain and snow is due to the rising temperature, the authors take an approach that the temperature, rather than the precipitation, should be the commencing point in the development of the changing future hydrology. We claim that the main cause of the "death" of the stationarity in a snow dominant watershed may be the warming temperature. Therefore, by commencing with the temperature in the hydrologic process, either the form of precipitation or the melting of the accumulated snow can be simulated and the non-stationary aspects of the future hydrology can be captured for more adequate water resources planning and management. The USGS under a contract to the California Department of Water Resources completed development of the Precipitation-Runoff Modeling System (PRMS) application for simulating daily streamflow for the Upper Feather River Basin. PRMS simulates all the major snowmelt/precipitation related physical processes including snowpack accumulation/melting, sublimation, evapotranspiration, surface runoff, subsurface flow, and ground water flow. The calibrated model simulates Water Years 1971–2001. This analysis perturbs the historical period daily minimum and maximum temperatures by 1°C, 2°C, 3°C, and 4°C, respectively, to determine the impact on snowmelt/runoff processes and ultimately streamflow at Oroville; all other parameters were unchanged from the historical simulation (Base Case). Precipitation spatial and temporal distribution was unchanged. Model output parameters studied include evapotranspiration, groundwater flow, groundwater recharge, interception evaporation, precipitation (rain/snow), runoff, snow cover percent, snow evaporation/sublimation, snowmelt, snowpack water equivalent, surface streamflow, subsurface flow, and subsurface recharge. For these varying degrees of warming, the outflows of the Basin were examined and compared to the base historical simulation. The timing of the center of the mass, the April through July runoff as a percent of the annual runoff, and the April snowpack water equivalent are shown to change appreciably with rising temperature.