|Title||A multi-decadal daily sea surface temperature and sea ice concentration data set for the ERA-40 reanalysis.|
|Series/Collection||ERA-40 Project Report Series|
|Event Series/Collection||ERA Report|
|Place of publication||Shinfield Park, Reading|
The lower boundary condition of sea surface temperature (SST) and sea-ice concentration (SIC) is a critical forcing of the lower frequencies in multi-decadal global atmospheric reanalyses such as next ECMWF reanalysis ERA-40. Partly in response to the ERA-40 project, new SST/SIC data sets have been developed that are considerably improved over those available to the first-generation reanalyses. This paper documents the input SST/SIC data sets and the processing that created the daily SST/SIC for the ERA-40 period 1957-2002. The source data are: 1) the monthly mean HadISST data set from the UKMO Hadley Center for 1956-1981; and 2) the weekly NCEP 2DVAR data for 1982-present. Both data sets are reanalyses of satellite and conventional SST/SIC observations. The principal reason for the higher quality of these source data sets is the use of a common consensus SIC and a common SIC-SST relationship in the sea ice margins. The use of a common SIC resulted in a very smooth transition between HadISST and NCEP 2DVAR, despite differences in data assimilation techniques and monthly versus weekly analyses. No special action was required to insure consistency at the transition unlike as was necessary for the AMIP II experiment. The only special processing was application of the AMIP II mid-month calculation for the interpolation of monthly mean data to daily values. This scheme insures that the monthly mean of the daily-interpolated data is nearly identical to the input monthly mean. Detailed comparisons of the SST and SIC during the HadISST-NCEP transition, and other long time series, are given. We also compare the NCEP 2DVAR (circa 2000) to a newer version of the OISST (V2, circa 2001) and demonstrate that the small differences should have no impact on the ERA-40 atmosphere reanalyses. Finally, we document precisely which data sets were used in ERA-40 production through a comparison of input and ERA-40 post-processed SST/SIC.