Since the SEASAT mission in 1978 (Fu and Holt, 1982), spaceborne synthetic aperture radars (SARs) have acquired millions of high-resolution images of ocean scenes, which have been used for applications such as wave and wind retrievals, oil pollution monitoring, ship detection, sea ice monitoring, and the interpretation of signatures of surface current gradients over oceanic fronts, internal waves, and shallow-water bathymetry. Unfortunately, despite the fact that a SAR is a Doppler radar, conventional SAR images do not provide direct information on target veloci-ties, since the Doppler information in the raw data is normally utilised to obtain the highest possible spatial resolution in flight (azimuth) direction. In a process called aperture synthesis, targets are mapped to azimuthal locations in the image where their contribution to the spectrum of the received signal during the SAR overpass appears at a Doppler frequency of 0. This implies the assumption that targets have a radial (line-of-sight) velocity of 0. Targets with a nonzero radial velocity will appear shifted in azimuth direction, and it is sometimes possible to retrieve their velocity from the visible displacement (e.g. between train and track or between ship and wake), but this is not possible for distributed targets such as the ocean surface.
展开▼
