|Title||The NEMOVAR ocean data assimilation system as implemented in the ECMWF ocean analysis for System 4|
|Year of Publication||2012|
|Authors||Mogensen, K, A. Balmaseda, WMAlonso|
|Secondary Title||Technical Memorandum|
NEMOVAR is a variational data assimilation software for the NEMO ocean model. Its development is a collaborative project among different partners, ECMWF included. NEMOVAR is the basis of the new ECMWF operational Ocean analysis System 4 (Ocean-S4). It is the first time that NEMOVAR is used operationally. Ocean-S4 uses NEMOVAR in its multivariate 3D-Var FGAT configuration. It assimilates temperature and salinity profiles and altimeter-derived sea level anomalies. In its current configuration Ocean-S4 includes an ensemble of NEMOVAR-derived reanalyses, consisting of 5 members. No attempt has been made to use the ensemble information in the specification of the background-error covariances. In Ocean-S4, the specification of the background-error covariances is somewhat heuristic, with computational efficiency and simplicity being important considerations. This report describes NEMOVAR and how it is used in both the reanalysis and real-time suites of Ocean-S4, called ORAS4 and ORTS4, respectively. The performance of NEMOVAR is evaluated using a set of objective metrics, including comparison with independent data and the impact on seasonal forecast skill. NEMOVAR improves the mean state and interannual variability of the ocean estimation in most regions, including the Equatorial Atlantic, a traditionally challenging area. NEMOVAR also improves the skill of seasonal forecasts. While the overall impact of NEMOVAR on the ocean state estimation is positive, some problem areas have been identified such as the representation of the Atlantic Meridional Overturning Circulation. It is expected that ongoing developments to include more versatile formulations of the background-error covariances will result in an improved data assimilation system.