Site content

2.3 GODAE Systems in operation

Lead author: Eric Dombrowsky (Mercator-Ocean)


Author/co-authors: E. Dombrowsky1, L. Bertino2, G. Brassington3, E.Chassignet4, F. Davidson5, H. Hurlburt6, M. Kamachi7, T. Lee8, M. Martin9, S. Mei10, M. Tonani11

 

1Mercator-Ocean, Toulouse, France
2Nansen Environmental and Remote Sensing Center, Bergen, Norway
3Centre for Australian Weather and Climate Research, BOM, Melbourne, Australia
4Florida State University, COAPS, Tallahassee, USA
5Fisheries and Oceans, St Johns, Canada
6Naval Research Laboratory, Stennis Space Center, Slidell, USA
7Japan Meteorological Agency, Meteorological Research Institute, Tsukuba, Japan
8Jet Propulsion Laboratory, Pasadena, USA
9Met Office, Exeter, UK
10National Marine Environment Forecast Center, Beijing, China
11Istituto Nazionale di Geofisica e Vulcanologia, Bologna, Italy


Abstract

During the last 15 years, operational oceanography systems have emerged in several countries around the world, thanks to progresses made in ocean modelling and data assimilation, in the large increase of computing power, and on the development of global observing systems such as altimetry and ARGO. This emergence has been largely fostered by the GODAE experiment, during which each nation engaged in this activity have organised partnership and constructive competition. This trans-national coordination was very beneficial for the development of operational oceanography, leading to economies of scales and more targeted actions. In addition, initiatives are taken to further develop the coordination of these national efforts. For example, Europe is building through the GMES initiative the Marine Core Service, aiming at providing operational oceanography services to the citizen and to serve -among others- the environmental policies.

Today, several systems provide routine real-time ocean analysis and forecast and/or reanalysis products. They are all based on (i) state-of-the-art primitive equation baroclinic Ocean General Circulation Model (OGCM) configurations, either global or regional (basin-scale), with resolutions that range from coarse to eddy resolving and (ii) data assimilation techniques whose complexity ranges from simple analysis correction to advanced 4D variational schemes. They assimilate altimeter sea level anomalies, remotely sensed SST such as GHRSST products and in situ profiles of T and S, including ARGO. Some systems have implemented downscaling capacities in specific regions of interest including shelf/coastal seas. Some also have implemented coupling with the atmosphere and/or the prognostic sea ice in polar regions.


The GODAE systems in operation are:

 

  1. Australia: BLUElink> (http://www.bom.gov.au/bluelink/) Ocean Model, Analysis and Prediction System (OceanMAPS) is Australia's first operational global ocean prediction system forecasting out to seven days. The system consists of the Modular Ocean Model version 4, the BLUElink Ocean Data Assimilation System and the Bureau of Meteorology global numerical weather forecasts. OceanMAPS has been operational since August 2007 and supports a range of services including an online graphic service, a data product service and naval defense applications.
  2. Canada: C-NOOFS (http://www.c-noofs.gc.ca/home_e.php/) is a downscaling initiative within GODAE using a free surface version of NEMO. It runs daily nested one way within available weekly MERCATOR global runs. C-NOOFS provides a 6day forecast for the North West Atlantic based on the latest Environment Canada wind forecast.
  3. China: NMEFC system, (http://dell1500sc.nmefc.gov.cn/argo-sz/argo4n.asp), NMEFC ocean reanalysis system provides monthly tropical Pacific Ocean reanalysis products. The system applied 3D-Var schemes and assimilated altimeter sea level anomalies, in situ profiles of T and S.
  4. France: MERCATOR system (http://www.mercator.eu.org/), Mercator Ocean is providing weekly global ocean forecast (2-week) with enhanced capacity in the North Atlantic and Mediterranean (higher resolution, daily forecast) and reanalysis products, using eddy resolving NEMO configurations and sequential multivariate assimilation (SEEK Kalman filter)
  5. Italy: MFS (http://gnoo.bo.ingv.it/mfs/) The Mediterranean Forecasting system provides a 10 days Mediterranean Sea forecast every day, using an eddy resolving model - NEMO - and 3DVAR for the assimilation of all available in situ and satellite data. Once a week the system is off-linecoupled with an ecosystem model, BFM, providing 10 days of biogeochemical forecast.
  6. Japan: MOVE/MRI.COM (http://goos.kishou.go.jp/ and http://godae.kishou.go.jp/), MOVE/MRI.COM is a real-time operational system at the Japan Meteorological Agency. It provides high resolution one-month forecast in the western North Pacific and reanalysis products, using eddy resolving model MRI.COM and multivariate 3DVAR assimilation method
  7. Norway: TOPAZ (http://topaz.nersc.no/) system provides weekly ocean and sea-ice forecasts out to ten days in the North Atlantic and Arctic. TOPAZ uses HYCOM at a resolution between 11 km and 16 km coupled to a dynamic-thermodynamic sea-ice model and the EnKF for assimilation of ocean and sea-ice observations.
  8. United Kingdom: FOAM (http://www.ncof.co.uk/) provides daily analyses and 5-day forecasts of the global ocean with enhanced resolution nested models of the North Atlantic, Mediterranean and Indian Ocean. The model component has recently been changed to NEMO which is forced by Met Office surface fluxes and assimilates data using an Optimal Interpolation type scheme.
  9. United States of America:
    1. ECCO system (http://www.ecco-group.org/), The ECCO-GODAE ocean reanalysis systems includes a delayed-mode estimation using the adjoint method and a near-realtime system based on a Kalman filter/smoother method
    2. NOAA/NCEP(http://polar.ncep.noaa.gov/ofs/) operates a real-time HYCOM Atlantic prediction system with 4-18 km resolution.
    3. U.S. Navy prediction systems:
      1. Operational linked 2-model global 1/32º NLOM - 1/8º NCOM system (http://www.ocean.nrlssc.navy.mil/global_nlom) that makes daily nowcasts and 4-day forecasts plus 30-day NLOM forecasts once a week. NCOM has 15 km mid-latitude resolution and 40 levels, while NLOM has 3.5 km mid-latitude resolution and 7-layers, including the mixed layer. NLOM steric sea surface height anomalies and mixed layer depth are used in creating synthetic T&S profiles assimilated by NCOM.
      2. Pre-operational 1/12º global HYCOM/NCODA system (7-km midlatitude resolution with 32 layers) that makes daily nowcasts and 7-day forecasts (http://www.hycom.org). The NOCDA data assimilation uses multivariate OI.

We will review the major characteristics of these systems. We will then illustrate their capacity showing a selection of recent scientific results.

 

(Last Updated: 13-10-2008)