Construction of wind stress fields using optimum interpolation with self-determined autocorrelation function

In order to construct gridded wind stress fields from observations by spaceborne scatterometers, an optimum interpolation method is utilized with spatial autocorrelation function of the wind field which is determined from the scatterometer-derived swath data every month. The autocorrelation function represents the nature of wind field. The e-folding scale of the wind field shows large seasonal and geographical variations. Using an optimum interpolation method with the derived autocorrelation function, daily surface wind and stress vector fields gridded in 1 deg. x 1 deg. (lat. x long.) are constructed from the NSCAT observations over the North Pacific. The derived wind fields are validated by comparison with ocean buoy data and are compared with other gridded wind products. It is concluded the optimum interpolation technique with the self-determined autocorrelation function works well to remove influence of the swath sampling of the scatterometer winds on the constructed wind fields and also preserves detailed structure of the fields.