In CERA, atmosphere data are assimilated with the 4-dimensional variational method part of the ECMWF Integrated Forecast System (IFS), while ocean data are assimilated with a 3-dimensional variational method developed for the ocean model NEMO, NEMOVAR. In CERA, the coupling of the ocean and atmosphere components is achieved at the outer-loop level: this guarantees that assimilation of an ocean observation has immediate impact on the atmospheric state estimate, and, conversely, assimilation of an atmospheric observation affects the ocean state.
Within ERA-CLIM2, this coupled approach has been applied to produce the first European coupled global reanalysis of the 20th-century, CERA-20C, which will provide the community with a century-long record of low-frequency climate variability and change using a consistent set of observations. In CERA-20C, the evolution of the global ocean and atmosphere for the period 1901–2010 is represented by a ten-member ensemble of 3-hourly estimates for ocean, surface and upper-air parameters. The ensemble technique has been used to simulate the effect of the inevitable uncertainties in the observational record and the forecast model, and to provide an indication of the data confidence.
In this talk, the CERA system will be described, and some results from CERA-20C, which has just been completed in June 2016, will be presented.