Temperature and precipitation extremes in climate model outputs over central Europe

For human society, knowledge of extreme events is as important as knowledge of the mean state of climate.Changes to the magnitude, character and spatial distribution of extreme temperature and precipitation events mayhave serious environmental, social and economic implications. The present study evaluates the simulation of tem-perature and precipitation extremes (high quantiles of distributions of daily maximum air temperatures and dailyprecipitation amounts) over central Europe in an ensemble of regional climate model (RCM) simulations withthe 25 km resolution, taken from the ENSEMBLES database. The statistics are compared for the recent climatesimulations (1961-1990) against data gridded onto the models’ grid from a high-density station network of dailyobservations,inordertoidentifymaindrawbacksoftheRCMsinreproducingobservedcharacteristicsofextremes(including their spatial patterns). Climate change scenarios (under the SRES A1B forcing) are then evaluated forthe mid-21st century (2020-49) and the late-21st century (2070-99) time slices, and the projected changes in ex-tremes are compared with changes in mean temperature and precipitation characteristics. The agreement amongmodels is better for temperature than precipitation extremes, and for the latter, in winter than summer. In sum-mer, increases in high quantiles of precipitation are projected in spite of pronounced drying (declines in meanprecipitation totals) in most RCMs.