Dynamical downscaling for improved climate forecasts and projections
We work to develop methodological approaches to the generation of dynamically downscaled global climate model data, using the Weather Research and Forecasting Model (WRF). Toward improved seasonal forecasts for the warm season, we have downscaled a retrospective reforecast of the Climate Forecast System (CFS) global seasonal forecast model, in collaboration with the Climate Prediction Center and Environmental Modeling Center with NOAA National Center for Environmental Prediction. We generate dynamically downscaled climate change projections of “well performing” models of the Coupled Model Intercomparison Projects, which are the basis of climate change projections in IPCC reports. These data are being currently used in the Southwest for future projection of water resources and severe weather during the North American monsoon. How natural climate variability is potentially synergistically interacting with climate change to intensify climate extremes if of particular interest. We work with numerous regional stakeholders to develop customized climate impacts assessment deliverables, including the United States Air Force, National Weather Service, Bureau of Reclamation, Central Arizona Project, and Salt River Project.
Warm season climate variability
Similar to the cool season, the first order control on natural climate variability in western North America during the warm are atmospheric teleconnection patterns. Teleconnections are basically the quasi-stationary Rossby wavetrains that drive continental-scale patterns of precipitation and temperature variability, and control the distribution of precipitation during the North American monsoon. We statistically characterize the nature of these atmospheric teleconnections, in a unified context that consider the paleoclimate (tree ring) record, modern instrumental record, and climate change projections for the coming century. On the intraseasonal timescale, the burst and break character of warm season precipitation is largely governed by synoptic transients, and of particular importance to the Southwest are transient inverted troughs.
Improved simulation of warm season convection
We are using WRF at convective-resolving spatial scale (1-2 km grid spacing) to simulate thunderstorms during the North American monsoon. This spatial scale is necessary to reasonably represent the organized, propagating convection (mesoscale convective systems and squall lines) that accounts for the most extreme precipitation events and a greater proportion of precipitation in urban centers and western deserts in Arizona and Sonora. In collaboration with scientists at the Universidad Nacional Autonoma de Mexico, we are currently investigating the utility of assimilating atmospheric water vapor derived from global positioning system technology, to improve short-term severe weather forecasts during the monsoon. We apply the same convective-resolving WRF modeling paradigm to project how severe weather will change in the Southwest in association with climate change.
WRF model development
We strive to be “intelligent” users of WRF and active contribute to model parameterization development. Specific examples of recent work along these lines includes the implementation of a modified convective trigger function in the Kain-Fritsch scheme, testing of the new NOAH Multiparameter (NOAH-MP) land surface model and the impact of dynamic vegetation, and testing of an urban canopy model. We find that such improvements in model parameterization may have significant bearing on the regional model simulation results, from the weather forecast to climate projection timescales.