Comparison between Flaash and Atcor Atmospheric Correction Modules on the Basis of Worldview-2 Imagery and in Situ Spectroradiometric Measurements

Highly fragmented landscape, small plots and ownerships, as well as the undulated topography of Crete cause a challenge for land use/cover mapping via the satellite imageries. Both the enhanced spectral and spatial resolution of WorldView-2 satellite sensor is expected to increase the discrimination potential of the classification process. Field spectroradiometric measurements were acquired using an ASD FieldSpec® 3 spectroradiometer handheld and/or mounted on a prototype carrier mobile platform MUFSPEM@MED (Mobile Unit for Field SPEctroradiometric Measurements at the MEDiterranean) that can reach heights up to 9.6 meters, concurrently with a WorldView-2 satellite overpass. Spectral responses from selected common plants and other surface features were registered. This study is part of a strategic program to elaborate the potential of hyperspectral field data for the improvement of land use/cover mapping, and image analysis and interpretation in the South-East Mediterranean. The present paper specifically focuses on the performance of three atmospheric correction methods, namely, simple regression analysis, FLAASH and ATCOR. The processed image radiometry has been compared with the field spectral data acquired at the same time. The resulting statistics display the high capacity of removing residual solar irradiance and atmospheric path radiance effects. Performance variations are presented and discussed. Ideas and recommendations tackling challenges for the further development of a spectral library are presented towards the enhancement of the radiometry calibration and validation capacity for the South-East Mediterranean. * Corresponding Author. Tel.: +3