Shallow Water Bathymetry And Optics Parameter Inversion Study Based On Hyperion Data

Depth survey of coastal water is an important work in aspect of shipping, water conservation, and coastal project and so on. With the development of satellite remote sensing, marine hyperspectral characteristic research can obtain more detailed target feature, so as to improve knowledge of marine water column properties. Shallow water bathymetry inversion based on hyperspectral remote sensing data is studied in this thesis.It is necessary to take atmospheric correction before image processing. The FLA ASH module is used to perform the atmospheric correction of the Hyperion hyperspectral imagery, which changes the top of atmosphere radiance spectrum to sea surface reflectance spectrum. Through principal component analysis of the region of interest (ROI), image bottom classification is taken by spectral clustering based on scatter diagram with maximum likehood classification. Calculating Kd of every pixels by a model for diffuse attenuation coefficient of downwelling irradiance and calibrating it in the sake of optical shallow water, then using respective values of bottom reflectance for each masked region, it is easy to derive depth information through Bierwirth directly inversion algorithm based on radiative transfer theory. In addition, QA A algorithm is used to calculate inherent optics parameters of selected areas.Selected two hyperspectral remote sensing image data to carry on depth inversion is studies in this thesis, and directly inversion method can derive the depth quickly. It also shows that the classification which combined scatter diagram with maximum likelihood can ensure accuracy of classification.