An aggregate spectral model for ERS-1 wind retrieval

The wind field at the ocean surface has been estimated via radar scatterometry for nearly two decades. Approximately one decade ago, several efforts were undertaken that investigated in detail the effect of the wind on the radar signature. These efforts were based on the use of an ocean spectral model in conjunction with a model of electromagnetic scattering. These techniques led to an improved understanding of the air/sea interface and its interrogation by radar remote sensing, but their success was hampered in part due to limitations of the ocean spectral models used in these efforts. Ocean spectral models that are employed in scatterometer-based wind retrievals must accurately model waves on scales spanning several decades. The waves responsible for the preponderance of the radar signature are on the order of a few centimeters, but these waves are tilted, modulated, stretched and compressed by the large-scale underlying waves. Due to the numerous difficulties associated with obtaining spectral measurements at high frequencies, particularly on the open ocean, the authors' understanding of the relevant theory and hence the modeling of ocean waves at high frequencies is still far from adequate. In addition, computational limitations still present considerable challenges to modeling of ocean particularly at the smallest scales that demand resolution. To address some of these issues, the authors are developing an ocean spectral model that combines in situ measurements of the directional wave spectrum with physically-based predictive models. This spectral model is described.