Evaluation of Different Climate Change Models for Future Precipitation Projections in Florida

Abstract

Precipitation being a vital input for many hydrological modeling studies has a direct bearing on the water resources modeling and management at different spatial and temporal scales. According to Intergovernmental Panel on Climate Change (IPCC), frequency of extreme precipitation events is expected to increase in future with no consistent trend in mean precipitation across the globe. To evaluate trends in precipitation, Global Circulation Models (GCMs) combined with statistical or dynamic downscaling techniques are generally used. However, it is agreed that skill of any climate change model is lower for precipitation compared to that for temperature. The model performance also depends on spatial and temporal resolution of the simulations. In the current study, precipitation projections based on fifteen GCMs from WCRP's(World Climate Research Program) Coupled Model Inter-comparison Project, phase -3 (CMIP3) project with different SRES (Special Report on Emission Scenarios) runs are analyzed for the state of Florida. Historical precipitation data is used for evaluation of the models via several performance measures and for selection of the best model. Long term historical precipitation data from United States Historical Climatology Network (USHCN) and GCM simulations from 20th and 21st century are used in this study. Efficacy and utility of Bias-Corrected Spatial Disaggregation (BCSD) procedure used in CMIP3 project for downscaling precipitation data for the state of Florida is assessed. Performances of models in two distinct seasons (wet and dry) that dominate tropical climate of Florida are also evaluated.