Hyperspectral Remote Sensing of Coastal Environment

Remotely sensed earth observation (EO) has revolutionized our understanding of our dynamic environment. Hyperspectral remote sensing is, in many ways, the ultimate optical remote sensing technology. Hyperspectral remote sensing is suited especially well for environmental studies due to it’s capability to discriminate between species and quantify the abundance of different materials and chemicals. In this thesis remote sensing methods for hyperspectral data, applicable in environmental monitoring of coastal environment are developed and tested.

The planning of hyperspectral flight campaign HYPE08 in South-West Finland raised the need to validate and develop data processing methods for HYPE08 as well as for future hyperspectral flight campaigns. The research presented in this thesis can be largely considered as a response to this need. The study concentrates on four main topics: wetlands mapping, benthic mapping, water quality and atmospheric correction. The study was done using airborne hyperspectral data and field spectroscopy measurements.

The results of this study emphasize the importance of local calibration and validation of methods used. Water quality retrieval algorithms developed in local environmental conditions outperformed those validated elsewhere. The results also show that hyperspectral remote sensing of benthic cover types is limited to rather shallow areas, indicating the need to use state of the art methodology in order to increase operational depth range. The proposed atmospheric correction algorithm produced very good results. The accuracy of model-based algorithm increases when Empirical-Line (EL) correction using spectral field measurements was applied to spectra generated by the model.