Satellite‐enhanced dynamical downscaling for the analysis of extreme events

The use of regional models in the downscaling of general circulation models provides a strategy to generate more detailed climate information. In that case, boundary‐forcing techniques can be useful to maintain the large‐scale features from the coarse‐resolution global models in agreement with the inner modes of the higher‐resolution regional models. Although those procedures might improve dynamics, downscaling via regional modeling still aims for better representation of physical processes. With the purpose of improving dynamics and physical processes in regional downscaling of global reanalysis, the Regional Spectral Model—originally developed at the National Centers for Environmental Prediction—employs a newly reformulated scale‐selective bias correction, together with the 3‐hourly assimilation of the satellite‐based precipitation estimates constructed from the Climate Prediction Center morphing technique. The two‐scheme technique for the dynamical downscaling of global reanalysis can be applied in analyses of environmental disasters and risk assessment, with hourly outputs, and resolution of about 25 km. Here the satellite‐enhanced dynamical downscaling added value is demonstrated in simulations of the first reported hurricane in the western South Atlantic Ocean basin through comparisons with global reanalyses and satellite products available in ocean areas. Key Points Blending two boundary‐forcing techniques to improve dynamical downscaling Assimilation of satellite‐based precipitation estimates for extreme‐event analysis Dynamical downscaling applied in regional hydroclimate reconstructions