ECMWF estimates the state of the global ocean via the operational system Ocean-S4. Ocean-S4 gives an estimate of the history of the ocean from September 1957 to present (with a few days delay) via the Ocean Reanalysis System 4 (ORAS4), as well as the latest ocean conditions, provided by the real-time extension Ocean Real Time Analysis System 4 (ORTA4).
Initialisation of coupled forecasts and ocean monitoring
The main purpose of the ocean analysis at ECMWF is to provide initial conditions for the extended range forecasts (seasonal and monthly), which are produced with a coupled ocean-atmosphere general circulation model. As well as providing initial conditions for coupled model forecasts, ocean reanalyses are valuable resources for climate monitoring, climate variability studies, and verification of climate models.
Ocean initial conditions are the main source of predictability at seasonal time scales, being especially important in the prediction of ENSO (El Niño Southern Oscillation) and its impact on the climate system. More generally, the correct initialisation of the upper ocean thermal structure is considered instrumental in the prediction of the tropical SST at seasonal timescales. At the monthly time scales, the prediction of phenomena such as the MJO (Madden Julian Oscillation) requires the correct representation of the ocean-atmosphere interactions. The current operational ocean reanalysis is ORA-S4, and it goes back to September 1957. The period 1981-2009 is used to initialise the calibrating hindcasts of the S4 seasonal forecasting system (Molteni et al., 2011). The earlier period of the reanalysis has been used to initialise CMIP5 decadal predictions.
A historical ocean reanalysis is required to provide initial conditions for the calibration of the coupled forecasts. The a-posteriori calibration of model output requires an estimate of the model climatology, which is obtained by performing a series of coupled hindcasts during some historical period. A historical record of hindcasts is also needed for skill assessment. The interannual variability represented by ocean reanalysis will have an impact on both the calibration and on the assessment of the skill.
An ensemble of ocean analyses (five in total) is performed to sample the uncertainty in the ocean initial conditions. The ensemble of ocean initial conditions provided by the five analyses contributes to the creation of the ensemble of forecasts for the probabilistic predictions at monthly and seasonal ranges. The ensemble can also be used to estimate the uncertainty in the climate signals from the ocean reanalysis. By construction, the ensemble samples uncertainty in the wind stress, in the observation coverage, and in the ocean state chosen to initialise the reanalysis.
The ingredients: ocean model, surface fluxes, data assimilation and ocean observations
Ocean initial conditions for the global ocean could in principle be estimated by forcing an ocean model with atmospheric fluxes of heat, momentum and fresh water. However both ocean models and atmospheric fluxes are far from perfect, and the estimation thus obtained (first guess) can have substantial uncertainty. In order to improve the estimation of the state of the ocean, this first guess is combined with ocean observations via a data assimilation procedure.
The current operational ocean estimation (Ocean-S4) is based in the NEMO ocean model and in the NEMOVAR data assimilation system. This is the first time that NEMOVAR is used operationally, and it is the first time that ECMWF uses the NEMO ocean model.
The NEMOocean model spatial discretisation is that of the tripolar ORCA1 configuration of NEMO, which has a horizontal resolution of about 1 degree, with equatorial refinement. It uses 42 levels in the vertical, 18 of which are in the upper 200m. ORCA1 configuration used in Ocean-S4 has been prepared at NOCS. Ocean-S4 uses version 3.0 of NEMO.
Daily surface fluxes of heat, momentum and fresh water are used to force the ocean model and to produce the first guess or the state of the ocean. The source of fluxes used in Ocean-S4 is shown in the diagram above. Prior to 1989, the surface fluxes are from the ERA-40 atmospheric reanalysis. From the period 1989-2009, the surface fluxes are from ERA-Interim reanalysis. From 2010 onwards, when Ocean-S4 started operational running, daily surface fluxes were derived from the operational ECMWF atmospheric analysis.
The NEMOVAR ocean data assimilation, in its 3D-var FGAT mode, is used to assimilate temperature and salinity profiles as well as along track altimeter derived sea level anomalies. The assimilation window is 10 days for the reanalysis stream ORAS4, while a variable length window is used for the real time stream ORTS4 (for more information about the Ocean-S4 streams see below). A bias correction scheme is used to correct the model/forcing errors. The bias correction is needed to ameliorate the spurious variability that can arise from changes in the observing system. The spatial distribution and time variations of the assimilation increments and bias correction terms is displayed in the Ocean-S4 ocean reanalysis product web pages
Ocean Observations are used to improve the ocean estimate given by the first guess. The figure above offers an schematic view of the different data sources used in Ocean-S4. The following types of ocean observation are used:
- Temperature and salinity profiles (T/S). The profiles are from the quality-controlled EN3 v2a data-set with XBT depth corrections until 2010, and from the Global Telecommunications System (GTS) thereafter. The T/S profiles consisting of the several data-types: XBTs (T only); CTDs (T/S); moorings (T/S); Argo profilers (T/S); and APBs (or elephant seals, T/S).
- Altimeter derived sea level anomalies (SLA), from AVISO, available from November 1992 onwards. The along track product is assimilated via NEMOVAR. Global mean values from the AVISO gridded maps of sea level anomaly (MSLA) are used to constrain the global fresh water budget.
Sea Surface Temperature (SST)and sea-ice. From September 1957 until November 1981) the SST/sea-ice are taken from the ERA-40 archive. From December 1981 until December 2009, SST/sea-ice are taken from the NCEP OI v2 weekly product, (Reynolds in the figure) and from January 2010 onwards the OSTIA SST/sea-ice is used. The SST and sea-ice information is used to constrain the upper level ocean temperature via a newtonian relaxation scheme.
The reanalysis and real time streams
The current ECMWF operational ocean analysis system is Ocean-S4. As for previous systems, Ocean-S4 consists of two analysis streams:
- A historical ocean reanalysis from 19570901 to present, updated every 10 days with a 6 day delay. It is used to initialise the coupled hindcasts needed for calibration of coupled model output. This stream is called ORAS4, for Ocean Re-Analysis System 4.
- A real time ocean analysis, produced daily, used to initialise the coupled forecasts. The real time stream is called ORTA4, for Ocean Real Time Analysis System 4.
The figure below shows schematically the operational schedule followed in the production of the two streams. Every 10 days, ORAS4 ocean analysis is advanced by 1 assimilation cycle (10 days). It runs with a delay of 6 days, in order to wait for the arrival of ocean observations (in particular, the 6 day delay is set by the retrieval of the sea-level products to constrain the global mean of the model sea level). This means that initial conditions from the reanalysis system could be up to 16 days old and thus not suitable to initialise monthly and seasonal forecasts. In order to create real-time ocean initial conditions on a daily basis, ORTA4 brings the latest ORAS4 state up to real time every day, using the available observations in a variable assimilation window. The length of the window is determined by the time difference between the latest ORAS4 reanalysis and the present day plus 1 day, so the output initial conditions will be valid on 0Z the following morning.
The data archive
The Ocean S4 products are not in the ECMWF catalogue, and are not archived in MARS.
Monthly mean output from selected fields will be publicly available here.
The output is in netCDF, and the ocean variables follow the naming convention of the NEMO model. For example:
- votemper: Ocean Potential Temperature (deg C)
- vosaline: Ocean Salinity (psu)
- vozocrtx: zonal velocity (m/s)
- vomecrty: meridional velocity (m/s)
Contact address: magdalena.balmaseda at ecmwf.int
References and further reading
Balmaseda, M. A., K. E. Trenberth, and E. Källén (2013), Distinctive climate signals in reanalysis of global ocean heat content, Geophys. Res. Lett., 40, 1754–1759, doi:10.1002/grl.50382.