Simulated relationship between air temperature and precipitation over Europe: sensitivity to the choice of RCM and GCM

This paper presents an evaluation of the relationship between near-surface air temperature (T) and precipitation (P) over Europe simulated by regional climate models (RCMs) from the EURO-CORDEX project. Simulations of two RCMs (RCA4 and CCLM4-8-17) driven by four different global climate models (GCMs) in the period 1971-2000 were analysed. Two different coefficients, the Pearson correlation coefficient based on monthly anomalies and the T-P index (TPI) based on daily values, were used for this evaluation. First, the comparison of T-P correlation and TPI patterns calculated from E-OBS data was made. Afterwards, simulated T-P relationship patterns were briefly evaluated. We especially focused on evaluating the RCM and GCM influence on the simulated T-P relationship pattern. Our results indicate that RCM simulations captured the basic seasonal dependence of the T-P relationship over Europe represented by both T-P correlation and TPI patterns. The simulated T-P relationship pattern is more strongly influenced by an RCM than a driving GCM in contrast to temperature means in winter and precipitation and temperatures changes, which are more strongly influenced by driving GCM as shown in previous studies. The simulations of the CCLM4-8-17 RCM demonstrated noticeably better agreement with observed T-P relationship fields than the RCA4 RCM simulations. The T-P relationship simulated by RCA4 tends to shift to positive values in comparison to E-OBS values. These findings were found for both TPI and T-P correlation.