CoAuthors

Hui Feng (Univ. of New Hampshire, United States); Lotfi Aouf (Meteo-France, France); Ngan Tran (CLS, France); Fabrice Ardhuin (IFREMER, France)

More accurate sea state bias (SSB) models that include global ocean surface wave model data have been shown to reduce error in the overall sea level variance signal and potentially alter the mean dynamic topography across ocean basins. However, there is a range of ocean wave models to choose from and their output is: i) highly dependent on the surface wind forcing product used to run them and ii) somewhat dependent on variations of the physics, optimization, data assimilation, and application that defines each specific wave model configuration. In order to consistently apply ocean wave model data to altimeter range correction, a strategy is needed to do this in a fashion that preserves the integrity and accuracy of each satellite’s measurements, and do this over the long term, i.e. from ~1992 to present. Ideally, this strategy can also provide pseudo-operational SSB corrections suitable for interim and science quality (IGDR and GDR) data records that are made available for use within weeks or less of real time. This study proposes and evaluates an approach that involves use of an existing hindcast WAVEWATCH III wave model database at IFREMER (using NCEP-CFSR wind data) that runs from 1990-2012 and then transitioning to the use of the operational Meteo-France WAM model (using ECMWF winds) starting in Dec. 2014. Potential problems with switching between models comes both in long term drift and spatial/temporal biases in the wind and wave field estimates that may in turn impact the SSB. In the present SSB model the focus is on the mean wave period parameter. Our results will address wave model differences observed in recent assessments, the proposed approaches to remedy differences and transitions in the context of the SSB correction, and assessment of the level of SSB error that results from transitioning between products. One encouraging preliminary result is the finding of improved SSB model performance when using wave period information from the data assimilating Meteo-France WAM output.