Models for the near-surface oceanic vorticity and divergence

Model-based scatterometer wind retrieval algorithms are based on parametric models for the near-surface wind field. Crucial to these models are the representation of the wind vorticity and divergence fields. As part of an effort to improve the modeling accuracy of these fields, the spectra of the vorticity and divergence fields has been computed using ERS-1 scatterometer winds. Over scales of 100 km to 1000 km the vorticity and divergence fields exhibit a power-law dependence on wavenumber k of approximately k^-2.6 and k^-1.5, respectively. This suggests that low-order numerical models can be used to model these fields with the level of accuracy required for model-based wind retrieval. The authors apply the Karhunen-Loeve (KL) transform to develop data-derived statistical models for the vorticity and divergence fields and compare the resulting wind field model to a previous model based on a polynomial expansion. The KL-based model provides some improvement in the model accuracy.