Change detection in shallow coral reef environments using Landsat 7 ETM+ data

This paper aims to clarify the potential of the new Landsat 7/Enhanced Them atic Mapper Plus (ETM+) sensors for change detection in coral reef environm ents. We processed images of two reef sites in Florida and Hawaii acquired over short time intervals (2 weeks and 3 months). During these periods, ree fs were not affected by major disturbances (phase shift, strategy shift, bl eaching, and hurricanes). This stability allowed us to assess the bias in c hange detection analysis. Two methods for change detection analysis were ap plied. The first one estimates the atmospheric conditions (Rayleigh and aer osol radiances, ozone and diffuse transmittances) using an ETM+/SeaWiFS mul tisensor approach. The second method is an empirical correction based on ps eudoinvariant features that compensates for different atmospheric condition s as well as for any sensor (noise) or environmental (water column, sea sur face state) conditions. The atmospheric correction alone did not provide an accurate match in images across time due to significant whitecaps and poss ible sun glint and its products required an empirical adjustment. Therefore , for the images in this study there was not substantial benefit in perform ing an atmospheric correction compared to an empirical correction alone. Bo th methods resulted in a minimum uncertainty of 4, 3, and 3 digital counts, respectively, in ETM+ Bands 1-3. Finally, we completed the study of real i mages by the analysis of ETM+ reflectance spectra for a large variety of co ral reef objects. We concluded that the assessment of the rates of change i n three ubiquitous classes 'sand,' 'background' (including rubble, pavement , and heavily grazed dead coral structure), and 'foreground' (including liv ing corals and macroalgae) emerges as the most reproducible and feasible ap plication for the ETM+ sensor. (C) 2001 Elsevier Science Inc. All rights re served.