| Abstract | Global coverage of the oceans is now available from synthetic aperture radar (SAR) sensors on board such satellites as ERS-1, ERS-2 and RADARSAT. With different applications in mind, it is useful to
extract as much ocean surface information as possible from each SAR scene. Scene statistics, calibrated backscatter, image spectra, and inter-look cross-spectra can all play a role in determining the wind and wave conditions present in the scene.
These, in turn, affect the detectability oftargets such asships and slicks on the ocean surface. Quantitative analysis of data from ERS-1, ERS-2, and RADARSAT, along with in situ measurements of wind and wave conditions at the time the data were
acquired, are combined to validate several information extraction techniques which apply to each of these C-band sensors. The techniques include the use of low wavenumber image spectra to resolve wind direction and the combination of these directions
with calibrated ocean backscatter to resolve wind speed. Another technique investigated is the use of inter-look cross spectra to reduce speckle noise contributions and to resolve the wave propagation direction in SAR image spectra prior to their
inversion to wave spectra. The use of scene statistics and the K-distribution order parameter in combination with the wind and wave information to predict the detection performance of the SAR for ships and other targets with unusual statistical
properties is also considered. While the potential exists to obtain a detailed snapshot of many aspects of the ocean surface and the targets on it with the SAR data alone, ancillary data may be required to fully resolve the wind direction and to
supply wave information outside of the azimuth passband. Synergistic use of wind and wave eld predictions together with the SAR imagery, should allow a more complete picture of the ocean to be developed than would be possible with either information
source alone. A case study is made of an atmospheric gravity wave which was observed in both a RADARSAT and a CCRS CV580 airborne SAR image. In this case, both the SAR data and other information sources are combined to provide a better picture of the
dynamics of the wave. |