DYNAMICALLY DOWNSCALED CLIMATE PROJECTIONS FOR ECOHYDROLOGICAL APPLICATIONS OVER THE SOUTHWEST

Predicting the impacts of climate change on terrestrial ecosystem requires an understanding of how climate variables will change at the regional scale. Unfortunately, state of the art climate change estimates are based on global climate projections using GCMs (global climate models), and while these models can give us a rough estimate of the climate at very large scales, they aren't useful at the local or regional scales – the scales that have the most important for ecosystems impacts. Hydrologic studies using raw coupled climate output yield poor results due to their coarse resolution and unrealistic land surface hydrologic representation. The goal of this work is to improve warm season climate projections in the North American Monsoon Region to be used in hydrological and ecological applications. To do this we are dynamically downscaling both historical and future global climate model data from the Hadley Centre for Climate Prediction and Research / Met Office UKMO-HadCM3 model. This model was chosen because it surpassed all other available models in representing the climate of the Southwest U.S. when evaluated using a set of relevant metrics. The dynamical downscaling uses the Weather Research and Forecasting (WRF) regional climate model and generates 6-hourly data at a 33km resolution covering the period from 1960-2081. Our hypothesis is that the regional models will improve upon the coarser resolution driving GCMs, particularly during the warm season, due to the fact that the physical mechanisms of rainfall during the summer are more related to mesoscale processes, such as the diurnal cycle of convection, low-level moisture transport, propagation and organization of convection, and surface moisture recycling. In general, these are poorly represented in global atmospheric models, and better captured in the regional models. Preliminary simulations show that WRF-downscaled simulations can provide a more realistic representation of convective rainfall processes. Thus a RCM can potentially add significant value in climate projections of the warm season provided the downscaling methodology incorporates spectral nudging to preserve the variability in the large-scale circulation while still permitting the development of smaller-scale variability in the RCM. With this condition, downscaled simulations can produce realistic continental-scale patterns of warm season precipitation. This includes a reasonable representation of the North American monsoon in the southwest U.S. and northwest Mexico, which is notoriously difficult to represent in a global atmospheric model. We anticipate that this research will help lead the way toward substantially improved projections of North American summer climate with a RCM.