Assessment of wind‐stress errors using bias corrected ocean data assimilation

Abstract

AbstractCurrent seasonal forecasting methods make extensive use of wind‐stress products and ocean thermal profile observations to produce ocean analyses for retrospective forecasts. The combination of inaccurate wind‐stress forcing and relatively accurate ocean thermal profile observations can have significant detrimental impacts on the analysis in equatorial regions as it leads to an imbalance between the zonal pressure gradient and wind stress. Model integrations have been performed to assess various wind‐stress products using a Pacific region ocean general‐circulation model in three assimilation modes. The assimilation of thermal profile data has been shown to impact negatively on the zonal and vertical current structure. A bias correction scheme has been used to illustrate how these errors can be mitigated. The method used to correct for these systematic errors allows for the errors to evolve in time and space. The strength of the equatorial undercurrent is brought to a realistic level, but the scheme does reduce its variability. The European Remote Sensing satellite scatterometer wind products have been used to illustrate the methodology and comparisons have been drawn with several other wind‐stress products. © Royal Meteorological Society, 2004. Contributions by M. R. Huddleston, M. J. Bell and M. J. Martin are Crown copyright.