• Home
• What is GODAE
• Activities
• Working Groups
• Data & Products
• Science
• Final Symposium
  • Announcements
  • Poster abstracts and pdfs
  • GODAE Ocean briefings
  • Invited papers
  • Symposium committees
• Special Issues
• Publications
• GODAE OceanView
• What's New
• Calendar
• Documents

GODAE is sponsored by

• 
• 
• 
• 
• 
• More Sponsors >

Your location: Home / Final GODAE Symposium / Invited papers / 3.5 PO abstract

3.5 Observing System Evaluation

Lead author: Peter Oke (CSIRO)

Author/co-authors: Peter R. Oke¹, Magdalena Balmaseda², Mounir Benkiran³, Jim Cummings⁴, Eric Dombrowsky³, Yosuke Fujii⁵, Stephanie Guinehut⁶, Gilles Larnicol⁶, Pierre-Yves Le Traon⁷, Matthew J. Martin⁸

Abstract

Global ocean forecast and reanalysis systems, developed under GODAE are a powerful means to assess the impact of different components of the global ocean observing system (GOOS). GODAE systems can be exploited to help identify observational gaps and to ultimately improve the efficiency and effectiveness of the GOOS forecasting eddy-resolving ocean models. Many tools are currently being used by the GODAE community to evaluate the GOOS. Observing System Experiments (OSEs, e.g. Figure 1), where different components of the GOOS are systematically withheld, can help quantify the extent to which the skill of a model depends of each observation type. Observing System Simulation Experiments (OSSEs) where the impact of hypothetical observations types, often through some form of twin experiment, help to evaluate the potential benefits of future observation types. Singular vector analysis can quantify the sensitivity of the ocean circulation to small perturbations, potentially identifying regions where additional observations are likely to benefit ocean forecasts. Analysis sensitivity, via the influence or information matrix, can be used to routinely assess the influence of individual observations and observations types on an analysis that can identify the most and least beneficial components of the GOOS. Examples using these tools developed and employed under GODAE will be presented in this paper.

Figure 1: Observed (column 1) and modelled SST (columns 2-6) from a reanalysis of the Tasman Sea using different observation types as labelled along the top row (NONE denotes no assimilation; ALTIM denotes altimetry; ALL denotes Argo+SST+ALTIM); adapted from Oke and Schiller (2007).

(Last Updated: 24-10-2008)