California Wintertime Precipitation Bias in 1 Regional and Global Climate Models

5 In this paper, wintertime precipitation from a variety of observational datasets, 6 regional climate models (RCMs), and general circulation models (GCMs) is aver7 aged over the state of California (CA) and compared. Several averaging method8 ologies are considered and all are found to give similar values when model grid 9 spacing is less than 3◦. This suggests that CA is a reasonable size for regional 10 intercomparisons using modern GCMs. 11 Results show that reanalysis-forced RCMs tend to significantly overpredict CA 12 precipitation. This appears to be due mainly to overprediction of extreme events; 13 RCM precipitation frequency is generally underpredicted. Overprediction is also 14 reflected in wintertime precipitation variability, which tends to be too high for 15 RCMs on both daily and interannual scales. 16 Wintertime precipitation in most (but not all) GCMs is underestimated. This is 17 in contrast to previous studies based on global blended gauge/satellite observations 18 which are shown here to underestimate precipitation relative to higher-resolution 19 gauge-only datasets. Several GCMs provide reasonable daily precipitation distri20 butions, a trait which doesn’t seem tied to model resolution. GCM daily and 21 interannual variability is generally underpredicted. 22