Improving scatterometer ocean wind vector data assimilation in Environment Canada's global variational analysis and forecast system

Dedicated efforts have recently been undertaken toward enhancing the variational assimilation of scatterometer ocean wind vectors in the Canadian Global Deterministic Prediction System. Surface wind retrievals provided by KNMI from the SeaWinds sensor onboard the QuikSCAT satellite have been routinely assimilated since May 2008. Before the failure of SeaWinds in November 2009, observations from the more recent Advanced SCATterometer (ASCAT) onboard the Metop-A satellite were introduced in an effort to enhance the spatio-temporal coverage of ocean surface wind observations going into the assimilation system. Further improvements to various aspects of the assimilation system related to scatterometer winds were subsequently investigated. Modifications to observation processing and quality control procedures have been tested, including the improved screening of sea ice-contaminated observations and improved thinning of ASCAT data. The impact of adjustments to the observation error statistics has also been assessed, and a new equivalent-neutral wind observation operator has been developed. An outline of results from assimilation sensitivity tests related to these issues and associated impact on forecasts will be presented.