Abstract. Using the fifth Coupled Model Intercomparison Project (CMIP5) model simulations and two observational datasets, the surface temperature trends and their discrepancies have been examined. The temporal-spatial characteristics for the surface temperature trends are discussed. Different from a constant estimated linear trend for the entire simulation period of 1850–2012, a dynamical trend using running linear least squares fitting with the moving 10 yr time windows are calculated. The results show that the CMIP5 model simulations are generally in good agreement with the observational measurements for the global scale warming, but the temperature trends depend on the temporal change and the regional differences. Generally, contrary to the small discrepancies on the global scale, the large discrepancies are observed in the south- and north-polar regions and other sub-regions.