Evaluation of reanalyzed precipitation variability and trends using the GPCP product over the CONUS

To better understand the long-term rainfall variability and trends over the continental United States (CONUS), an accurate temporally and spatially homogenous precipitation data should be used. Atmospheric reanalysis datasets appear to fit these criteria, however, their uncertainties must first be addressed. In this study, precipitation estimates from five global reanalyses (ERA-Interim, MERRA2, JRA-55, CFSR, and 20CR) and one regional reanalysis (NARR) are compared against the GPCP Satellite-Gauge combined product over the CONUS during the period from 1980 through 2013. Reanalyses capture the variability of precipitation distribution over the CONUS as derived from GPCP, however, large regional and seasonal differences exist. Global reanalyses generally overestimate the precipitation over the western part of the country throughout the year and underestimations are noticed over the northeastern US during fall and winter seasons, which may be associated with the difficulties within the models to accurately reproduce the precipitation over the complex terrains and during snowfall events. Furthermore, systematic errors found in five global reanalyses suggest that their physical processes in modeling precipitation need to be improved in the future. Even though negative biases exist in NARR, it has very similar precipitation pattern as observed in GPCP; this is anticipated because it assimilates rain gauge analysis unlike other global reanalyses. We also conduct trend analysis of precipitation over the CONUS using GPCP and reanalyzed precipitation products from 1980 to 2013. Based on the linear regression of GPCP data, there is a decreasing trend of 2.00 mm/year. For spatial distribution, only north central and northeastern parts of the county show positive trends, while rest of the regions show negative trends through course of a year. Compared to the GPCP observed long-term trend of precipitation, all reanalyses except for 20CR exhibit similar inter-annual variation, and a significant decreasing trend is found in ERA-Interim. Although comprehensive global reanalyses that assimilate both satellite and in-situ observations can provide more reasonable precipitation estimates, substantial efforts are still required to further improve the reanalyzed precipitation over the CONUS.