Bathymetry of shallow coastal regions derived from space-borne hyperspectral sensor

Hyperion is a hyperspectral sensor on board NASA's EO-1 satellite. Its spatial resolution is about 30 meters with a swath of /spl sim/7 Km. Though Hyperion was not designed for ocean studies, its unique spectral configuration (430 nm-2400 nm with a /spl sim/10 nm step) makes it especially attractive to study the effectiveness of such kind of sensor for observing complex coastal waters. In this study, Hyperion data over two sites of the Florida coasts were acquired, with one focused on the clear Key West waters, and the other focused on the relatively turbid Tampa Bay waters. From both data sets, water properties and bottom bathymetry were simultaneously derived from atmosphere-corrected Hyperion data using a spectral matching technique. More importantly, in the top-to-bottom processing of Hyperion data, there was no use of any a prior or ground truth information. For the Key West site, derived bathymetry and water properties were validated with NAVOCEANO CHARTS (active bathymetric LIDAR system) and field measurements, respectively. It is found that the retrieved depths (in a range of /spl sim/1-20 m) match LIDAR depths very well (/spl sim/15% average error), indicating significant potential of using hyperspectral satellite sensor for efficient and repetitive observation of shallow coastal regions.