An assessment of SeaWinds on QuikSCAT wind retrieval

The scatterometer ocean wind retrieval process produces several possible solutions or ambiguities at each point, requiring a separate ambiguity selection step to infer a unique wind vector field. An ambiguity selection error occurs when the selected wind vector is not the closest ambiguity to the true wind. The current ambiguity selection routine for SeaWinds is ad hoc, but performs well under most circumstances. Factors such as instrument noise and rain can also cause the estimated wind flow to deviate from the true wind. A quality assurance (QA) analysis is performed to assess the ambiguity selection effectiveness and noise level of the retrieved wind using a low‐order wind field model. The wind field model is data‐driven and shown to be rather insensitive to the training data set. The QA analysis demonstrates that the SeaWinds ambiguity selection process is at least 95% effective. Ambiguity selection errors are correlated with storms and rain corruption. A subjective analysis on a set of cyclonic storm passes confirms that the wind retrieval is somewhat less effective in storm regions.