• low precipitation pattern;
  • climate change;
  • GCM;
  • multi-model ensemble;
  • model consistency


This study assesses future changes in low precipitation patterns over land around the globe under the Special Report on Emissions Scenarios A1B scenario. We use global precipitation data sets derived by the super-high-resolution Atmospheric General Circulation Model of the Japan Meteorological Agency and Meteorological Research Institute (MRI-AGCM3.1S) and Atmospheric Ocean coupled General Circulation Models compiled in the phase 3 of the Coupled Model Intercomparison Project (CMIP3 AOGCMs). The low precipitation patterns refer to the temporal and spatial distribution of annual minimum of accumulated precipitation over a given time interval using occurrence probability and observation windows in unfixed seasons. As for the low precipitation patterns, this study assesses low precipitation quantiles in six probability levels approximated by the Weibull distribution using annual minima of monthly accumulated precipitation over 1-, 3-, 6- and 12-month time intervals. Also, low precipitation occurrence season were assessed from the centroid of the frequency distribution of the ending months with annual minima of the low precipitation over the four time intervals. The model capability of reproducing current low precipitation patterns was examined in reference to the global precipitation observation data set of VASClimO, from which MRI-AGCM3.1S showed the best performance among all models. The MRI-AGCM3.1S projections, as well as the multi-model ensemble mean calculated over 16 GCMs, indicate that 0.1 probability low precipitation quantiles in 3- and 6-month intervals would decrease by 10 to 50% in Mexico, southern Brazil, southern Argentina, Mediterranean area and southern Africa with high model consistency. The projected shifts in low precipitation occurrence seasons were of concern; however, they show a considerable degree of uncertainty with low model consistency. Copyright © 2013 John Wiley & Sons, Ltd.