Spatial Variability of Random Error and Biases in the FSU3 Winds

The new FSU3 wind climatology will be introduced and briefly described. The spatial patterns of random errors and biases are examined for the FSU3 wind product. The biases are assessed in comparison to satellite observations, and are small for must areas. Differences are largely due to the smoothing required in the in situ product. An objective technique has been developed to determine uncertainty (random errors) as a function of sampling and observational error, as well and the functional used in the variational method. This technique can be applied to any objectively gridded product for which sampling and observational errors can be estimated for all the input data. This technique will be briefly described. In the near future, this technique will be extended to all fields in the FSU3 in situ based flux product. The influences of sampling and observational errors are compared. This detailed error analysis will be of great use to users of the FSU3 gridded product.