Maritime hyperspectral imaging and coastal bathymetry from space

Spaceborne maritime hyperspectral imagery promises to be a powerful technique for the retrieval of coastal bathymetry and other environmental parameters worldwide. In hyperspectral imaging, a contiguous spectrum of the light leaving each ground sample is recorded. Bathymetry retrievals are performed by modeling the water-leaving spectral radiance of sunlight scattered by the water column and bottom, and matching the modeled data to the measured spectra. This can be done using a look-up-table (LUT) in which modeled spectra are stored for many combinations of water and bottom parameters, which are then searched to find the modeled spectrum that best matches the measured data for each ground sample. This method is applied to 4 m resolution airborne hyperspectral imagery of the Indian River Lagoon in Florida, and the retrieved bathymetry is compared to 30 m resolution ground validation data. LUT bathymetry retrievals for regions identified as a having uniform bottom type correlated well with measured bathymetry to a depth of 5 m. These preliminary results suggest that the LUT method has the potential for accurate bathymetry determination from space.