Description
Summary:The aim of this paper is to review how the Global Ocean Data Assimilation Experiment (GODAE) centers will proceed to ensure the value and quality of their ocean products and to evaluate the performance of their systems. The strategy is to define a set of standard internal verification tests and metrics. The scientific credibility will rely on careful checks of the consistency of the system outputs with state-of-the-art knowledge of the ocean state and its variability. The quality assurance of the products will rely on systematic verification of key parameters and computation of statistical indexes by reference to both climatologies and real-time data, and, in a delayed mode, to quality controlled observations. The performance of the systems will rely on diagnostics based on key indicators such as estimates of forecasting skill, ability to constrain a sparsely observed field or non-assimilated field, and evaluation of real-time versus re-analysis products. The internal metrics discussed in this paper rely on the following diagnostics: consistency analysis, including time-mean climatologies, space and time variability, and physical balance; quality assurance, including time series stations, satellite remote sensing, and VOS and floats; and performance analysis, including error estimates and model-data differences. The sharing of experience between the GODAE centers is critical. Intercomparison exercises between the different Centers is one way to respond to this need. A few examples of metrics for intercomparison are given from operational systems that are specific to the Atlantic and the Pacific basins, where knowledge of ocean characteristics is the most advanced and where comparison exercises are underway as part of the GODAE common. (8 refs.) Presented at an International Symposium "En route to GODAE," held in Biarritz, France, on 13-15 Jun 2002. Prepared in collaboration with LEGOS/CNRS, Toulouse, France; Met Office, Berkshire, United Kingdom; RSMAS/MPO, University of Miami, Miami, FL; Propulsion Laboratory, Pasadena, CA; and the Meteorological Research Institute, Tsukuba, Japan.