An up-to-date assessment of temperature extremes over the MENA region from observational and CIMP5 data

We assess observed and modeled temperature extremes over the MENA region during the last four decades. The purpose of our analysis is two-fold: I) provide an up-to-date, observationally based estimation of recent past evolution and ii) evaluate the performance of global climate model simulations. A list of indices of temperature extremes, based on threshold, percentile, heatwave and coldwave characteristics is used, as defined by the Expert Team on Climate Change Detection and Indices (ETCCDI). We derive the indices. We use daily near-surface air (2-metre) temperature (Tmax and Tmin) to derive the extremes indices for the period 1980-2018 from: 1) re-analyses (ERA-Interim, MERRA2) and gridded observational data (Berkeley) and ii) 18 CMIP5 model runs combining historical (1950-2005) and scenario runs (2006-2018 under RCP 2.6, RCP4.5 and RCP8.5). Using these reanalyses, observational and CMIP5 multimodel ensemble data-sets we derived their statistics (climatological average, trends) and produced maps for the MENA region. In addition, the CMPI5 indices were compared with the indices derived from the observational and reanalyses and their biases were revealed through spatial (maps) and temporal (time-series) comparison. It is found, as expected, that the choice of the RCP does not make any difference in the calculations up to 2018 but nevertheless the use of the three “scenarios” provides a better model sample for the evaluation against observations. Finally, the best performing global model realizations for the temperature extremes are revealed by the comparison of individual models with the re-analyses and observational data.