UCLA

Increasing risk of many natural hazards including floods, wildfires, and reduction of agricultural production are often due to extreme climate events such as heavy precipitation, heat waves and drought. In this study, the variable-resolution Community Earth system model (VR-CESM) was used to investigate potential changes in extreme climate events at refined-resolution (14 km) in the western U.S. and eastern China. The model performance for simulating climate extremes is evaluated by various extreme metrics and assessed against the PRISM reference dataset in the U.S. VR-CESM is also compared with other models at different resolutions, including the uniform coarser resolution model CESM with the same physical parameterization, and HadGEM2-WRF, one RCM in the CORDEX-NA ensembles with the same SST forcing as the GCMs. Among those precipitation extreme metrics, VR-CESM simulates the short-term daily metrics (e.g., 95th percentile precipitation) in the western U.S. well (with greater uncertainties in the Sierra Nevada and Cascade Mountains), but underestimates the long-term metrics (e.g., number of dry years) during the historical period. For temperature extremes, VR-CESM also overestimates the number of hot temperature extremes in California's Central Valley and southwestern desert regions; this is consistent with the warm seasonal average JJA biases. In model intercomparisons, VR-CESM has smaller biases than CESM and HadGEM2-WRF for both extreme precipitation and temperature events. Additionally, the Palmer Drought Severity Index (PDSI) for all models is simulated at four watersheds and compared to the PRISM reference dataset with statistical testings for their similarities. Under RCP8.5 through 2050, VR-CESM projects stronger and more frequent extreme precipitation in the mountains and hot events in the deserts of the western U.S., with more wet periods in California as indicated by PDSI. In another VR-CESM refined-resolution region in China, VR-CESM projection indicates an increase in strength and magnitude of extreme precipitation in southern China and hot temperature extremes in eastern China. Geographic heterogeneity in PDSI projections is also evident as the North China Plain is projected to have more dry periods while the Three Gorges Basin is projected to not change much. This study demonstrates that the analysis of multiple extreme metrics is essential to provide an adequate understanding of how extremes may change in the future in a particular place.