SPARC Reanalysis Intercomparison Project (S-RIP) Final Report

Abstract

The S-RIP activity focuses predominantly on reanalyses, although some chapters include diagnostics from operational analyses when appropriate. Many of the chapters focus primarily on newer reanalysis systems that assimilate upper-air measurements and produce data at relatively high resolution (i.e., ERA-Interim, JRA55, MERRA, MERRA-2, and CFSR). The ERA5 reanalysis, which was released during the latter stages of the activity, is not fully evaluated but is included in some intercomparisons. Selected long-term reanalyses that assimilate only surface meteorological observations (e.g., NOAA-CIRES 20CR, ERA-20C, and CERA-20C) are also evaluated where appropriate. Some chapters include comparisons with older reanalyses (NCEP-NCAR R1, NCEP-DOE R2, ERA-40, and JRA-25/JCDAS), because these products have been extensively used in the past and are still being used for some studies, and because such comparisons can provide insight into the potential shortcomings of past research results. Other chapters only include a subset of these reanalysis data sets, since some reanalyses have already been shown to perform poorly for certain diagnostics or do not extend high enough (e.g., pressures less than 10hPa) in the atmosphere. At the beginning of each chapter an explanation is given as to why specific reanalysis data sets were included or excluded. The minimum intercomparison period is 1980-2010. This period starts with the availability of MERRA-2 shortly after the advent of high-frequency remotely sensed data in late 1978 and ends with the transition between CFSR and CFSv2. Some chapters also consider the pre-satellite era before 1979 and/or include results for more recent years. Some chapters use shorter intercomparison periods for some diagnostics due to limitations in the observational record available for comparison and/or computational resources.