The Godae Intercomparison Project

Recently the GODAE Intercomparison Project has allowed to intercompare and perform accuracy and consistency assessment. The objectives of the project were to (a) demonstrate GODAE operational systems in operations; (b) share expertise and design validation tools and metrics endorsed by all GODAE operational centers; (c) evaluate the overall scientific quality of the different GODAE operational systems (results are summarized in Hernandez et al. 2009).

This project involved the majority of operational centres worldwide delivering daily ocean products, such as: BLUElink (Australia), HYCOM (USA), MOVE/ MRI.COM (Japan), Mercator (France), FOAM (United Kingdom), C-NOOFS (Canada), and TOPAZ (Norway) systems (Dombrowsky et al. 2009; Hurlburt et al. 2009). It provides a diversity of ocean models -4 types; global, or regional; based on different vertical discretizations; eddy-permitting to eddy-resolving; coupled or not with sea-ice models; using different types of air-sea flux modelling. It provides also some diversity of assimilation techniques, using not the same kind of observations; proceeding to weekly or daily analysis or updates; based on sequential or varia-tional approaches; based on single or ensemble analysis and predictions; applying or not "close to data" schemes like First Guess At Appropriate Time (FGAT) and Incremental Analysis Update (IAU) techniques (Bloom et al. 1996).

It was initially decided to analyse similarly the operational outputs of the different OFS involved. February, March and April 2008 was the selected period. In practice, all output could not be provided in real-time, and the scientific evaluation has been performed with some month delay.

A series of observations and reference dataset have been used to assess the accuracy and consistency of the ocean products, using Class 1 and Class 2:

• Weekly maps of Sea Surface Height (SSH) or Sea Level Anomalies (SLA) from AVISO satellite altimetry25.

• Weekly maps of surface currents derived from satellite altimetry (Larnicol et al. 2006).

• The Levitus WOA 2005 climatology (Antonov et al. 2006; Locarnini et al. 2006).

• The Mixed Layer Depth climatology (D'Ortenzio et al. 2005; de Boyer Monte-gut et al. 2004, 2007).

• Daily sea-ice concentration from satellite from OSI-SAF26.

• OSTIA GHRSST SST products (Donlon et al. 2009).

• In-situ temperature and salinity, provided by CORIOLIS.

In practice, all groups could contribute to the intercomparison. Specific studies were carried out: in the north, the south and the tropical Atlantic, the western north Pacific, the tropical Pacific, and the Indonesian Seas. All group had access to all output and reference dataset. SST consistency and accuracy was verified against OSTIA maps. Water mass consistency was evaluated using the WOA 2005 climatology. Mixed Layer Depth consistency was verified with the climatology. Sea level and mean circulation were assessed against satellite altimetric maps. Mean and eddy kinetic energy were compared at the surface with SURCOUF maps (Larnicol et al. 2006).

Three months are rather short to infer the circulation patterns analysed in DYNAMO or reanalysis project. However, the consistency assessment allowed verifying if the "mean" circulation was like expected. For instance, in Fig. 23.8 the current analysis in the North Atlantic showed us the consistency of the subtropical and subpolar gyre circulation. One can note that the Azores Current appears for some systems, that the Gulf Stream extension does not spread similarly, or that the Labrador and the East-Greenland current are more or less intense. The use SURCOUF data allowed to status on the quality of the different outputs: eddy kinetic energy can be computed, and accuracy numbers given. However the high resolution systems

25 See footnote 23.

26 See footnote 22.

M'« i-,;, t ' M omi t'V 'Hi J jqohco imj 1 I ou* Sil: ..*!.', mi

M'« i-,;, t ' M omi t'V 'Hi J jqohco imj 1 I ou* Sil: ..*!.', mi

Fig. 23.8 From (Hernandez et al. 2008). Averaged eddy kinetic energy (m2/s2) from February to April 2008 for: TOPAZ (top-left), HYCOM (middle-left), C-NOOFS (middle-bottom-left), FOAM (top-right), Mercator High-res (middle-right), Mercator global (middel-bottom-right) and the observed SURCOUF product (bottom-left). Bottom-right: Time series of eddy kinetic energy box averaged in limited area around the Gulf Stream (80-60°W and 30-42°N)

IMTVOI fCP

Fig. 23.8 From (Hernandez et al. 2008). Averaged eddy kinetic energy (m2/s2) from February to April 2008 for: TOPAZ (top-left), HYCOM (middle-left), C-NOOFS (middle-bottom-left), FOAM (top-right), Mercator High-res (middle-right), Mercator global (middel-bottom-right) and the observed SURCOUF product (bottom-left). Bottom-right: Time series of eddy kinetic energy box averaged in limited area around the Gulf Stream (80-60°W and 30-42°N)

seem to provide more energetic features than SURCOUF. Here we can suspect that SURCOUF currents are smoother than HYCOM. Which indicate than reference dataset have also to be taken with caution. Similar limitation appeared using OSTIA SST: OSTIA map can be dubious when satellite data are lacking. Thanks to OSTIA error estimates provided together with SST values, the intercomparison could focus on "valuable" areas. The full overview of this first intercomparison experiment is given in (Hernandez et al. 2008, 2009).

This first international intercomparison of OFS was limited to a short period, and a short set of ocean parameters. Impact of forcing field was not studied, neither the time-varying aspect of ocean features (eddy propagation, waves.) or sea-ice. The analysis was also limited to hindcast: forecast and performance metrics could only be assessed in a limited way. This initiative should carry on in the framework of the GODAE Ocean View project. More reference dataset will be made available in real-time soon, and the methodology, the metrics, are now adopted by most groups. This experiment has shown that intercomparison and evaluation of OFS could be performed in any part of the ocean. The three-months limited period could address the consistency, and accuracy of OFS for this season. The performance of the system, with regard of their particularities (resolution, model approximations, assimilation method.) started to be evidenced. Next step would be to carry on the intercomparison in term of multi-model ensemble assessment.

Was this article helpful?

0 0
Guide to Alternative Fuels

Guide to Alternative Fuels

Your Alternative Fuel Solution for Saving Money, Reducing Oil Dependency, and Helping the Planet. Ethanol is an alternative to gasoline. The use of ethanol has been demonstrated to reduce greenhouse emissions slightly as compared to gasoline. Through this ebook, you are going to learn what you will need to know why choosing an alternative fuel may benefit you and your future.

Get My Free Ebook


Post a comment