Pacific northwest climate sensitivity simulated by a regional climate model driven by a GCM. Part II: 2xCO(2) simulations

Global climate change due to increasing concentrations of greenhouse gases has stimulated numerous studies and discussions about its possible impacts on water resources. Climate scenarios generated by climate models at spatia l resolutions ranging from about 50 km to 400 km may not provide enough spa tial specificity for use in impact assessment. In Parts I and II of this pa per. the spatial specificity issue is addressed by examining what informati on on mesoscale and small-scale spatial features can be gained by using a r egional climate model with a subgrid parameterization of orographic precipi tation and land surface cover, driven by a general circulation model. Numer ical experiments have been performed to simulate the present-day climatolog y and the climate conditions corresponding to a doubling of atmospheric CO2 concentration. This paper describes and contrasts the large-scale and meso scale features of the greenhouse warming climate signals simulated by the g eneral circulation modal and regional climate model over the Pacific Northw est. Results indicate that changes in the large-scale circulation exhibit strong seasonal variability. There is an average warming of about 2 degrees C, an d precipitation generally increases over the Pacific Northwest and decrease s over California. The precipitation signal over the Pacific Northwest is o nly statistically significant during spring, when both the change in the la rge-scale circulation and increase in water vapor enhance the moisture conv ergence toward the north Pacific coast. The combined effects of surface tem perature and precipitation changes are such that snow cover is reduced by u p to 50% on average, causing large changes in the seasonal runoff. This pap er also describes the high spatial resolution (1.5 km) climate signals simu lated by the regional climate model. Reductions in snow cover of 50%-90% ar e found in areas near the snow line of the control simulation. Analyses of the variations of the climate signals with surface elevation ranging from s ea level to 4000 m over two mountain ranges in the Pacific Northwest show t hat because of changes in the alitude of the freezing level, strong elevati on dependency is found in the surface temperature, rainfall. snowfall, snow cover, and runoff signals.