Downscaling hydroclimatic changes over the western US based on CAM subgrid scheme and WRF regional climate simulations

Abstract

This study examines two dynamical downscaling methods, a subgrid parameterization and a regional climate model, to compare their impacts on simulating orographic precipitation and surface hydrology in mountain regions. A global climate model was first applied at 1° × 1.25° grid resolution with a subgrid orographic precipitation scheme. The global simulations were then used to drive a regional climate model at 15-km grid resolution over the Western United States. By comparing the global and regional simulations for two 10-year periods, 1993-2003 and 2039-2049, this study assesses the two downscaling methods in the context of simulating both the present climate and climate change signals, and the implications of the relatively short simulation length to investigate differences in current and future climate simulated by the models are discussed. The model results show that improving the representation of surface topography through higher spatial resolution or a subgrid method has a large impact on the simulations. Both the subgrid scheme and the regional model significantly improved the simulation of snowpack in the mountains. The spatial distributions of precipitation and snowpack are generally consistent between the subgrid and regional simulations, since they were driven by the same large-scale circulation from the global simulations. However, because rain-shadow effects are not represented in the subgrid scheme, the regional simulations produced much more realistic spatial variability in precipitation and snowpack than the subgrid simulations in narrow mountain ranges. In the climate change experiments, both downscaling procedures preserved the large-scale patterns of temperature and precipitation changes in the global simulations. However, the regional simulations show larger changes in precipitation and snowpack along the coastal mountains than the subgrid simulations. This is attributed to the fact that the regional model explicitly simulates the interactions of atmospheric circulation and the underlying topography, so changes in wind directions with respect to the orientations of the mountains may lead to changes in orographic precipitation that cannot be explained by changes in atmospheric temperature and moisture alone. Hence differences between the precipitation changes simulated by the regional model and the subgrid method are larger in narrow mountains such as the Cascades and the Sierra Nevada because the subgrid method does not account for the influence of mountain orientations at the subgrid scale. As precipitation is an important driver of surface hydrological processes, differences between the precipitation changes simulated by the two methods lead to important differences in the surface hydrological processes under climate change. © 2009 Royal Meteorological Society.