Multi-sensor ocean colour atmospheric correction for time-series data: Application to LANDSAT ETM+ and OLI data.

This paper includes developments to a multi-sensor Atmospheric Correction (AC) that can be applied to marine and terrestrially-orientated medium resolution optical sensors, with a focus on the processing of imagery for coastal applications. The AC was originally developed for the airborne Compact Airborne Spectrographic Imager (CASI), and has since been expanded to include Compact High Resolution Imaging Spectrometer on the PRoject for On-Board Autonomy satellite mission (CHRIS-PROBA) with this paper detailing the extension to Landsat; examples shown are for the Landsat 7 Enhanced Thematic Mapper Plus (ETM+) and Landsat 8 Operational Land Imager (OLI). The underlying approach is a Coastal Zone Color Scanner (CZCS) style aerosol estimation, using an Angstrom exponent, where the water's contribution to the Top Of Atmosphere (TOA) signal within the aerosol estimation wavebands is subtracted using a Bright Pixel (BP) estimation. This BPAC was originally developed for MEdium Resolution Imaging Spectrometer (MERIS), and includes a near infrared (NIR) optical water model. For Landsat, there is also an additional approach that extends the AC over the land using shortwave infrared (SWIR) vegetation 'dark targets' i.e. akin to the Dark Pixel (DP) approach over water. The preliminary results appear promising, but further research is needed. Although the suspended particulate matter (SPM) maps have realistic patterns, spectral plots comparing the Landsat results to MERIS and Moderate Resolution Imaging Spectroradiometer on the Aqua spacecraft (MODISA) atmospherically corrected data appear to show that the reflectance values are too high and the spectral shape indicates the BPAC is failing in some areas for the Landsat ETM+ data.