Using LOCA to downscale precipitation over Europe

Localized Constructed Analogs (LOCA) is a statistical downscaling technique that uses a multiple scale approach to determine appropriate local analogs from historical data. It was developed with a particular focus on the preservation of extreme events that were dampened or lost altogether when employing earlier analog methods. The LOCA method has been used to produce relatively high-resolution projections of precipitation over the coterminous United States for use in hydrologic applications but has never been applied over Europe. In this presentation we will describe the method in detail and show how it is being utilized to downscale CMIP6 precipitation to 1 arcmin x 1 arcmin horizontal resolution over the continent using the European Meteorological Observations (EMO-1) gridded dataset as the analog pool. Lastly, we will compare the LOCA output to that from other downscaled products.