Noise Reduction Correction Research Articles

Detection of changes in coral reef communities using Landsat-5 TM and Landsat-7 ETM+ data

Satellite remote sensing is increasingly used to map and monitor coral reefs. From 1984 to the present, Landsat-5 thematic mapper (TM) and Landsat-7 enhanced thematic mapper plus (ETM+) images provide the longest time series available for change detection analysis over coral reefs. A time series of four Landsat-5 images and one Landsat-7 image spanning 1984‐2000 was analyzed to detect changes in "coral-dominated", "sand", "algae", and "substrate" benthic classes for Carysfort Reef in the Florida Keys. To properly analyze this time series, a set of corrections was undertaken, which included noise-reduction correction, atmospheric correction, and TM‐ETM+ data normalization. All images were classified with a Mahalanobis distance classifier using statistics from the 1984 image to identify the four benthic classes. The results were compared with historical ground-truthing data, a combination of high-resolution aerial photography and Ikonos satellite data, and results from a temporal texture change detection analysis. All data sets provided consistent results, with an extreme loss in coral cover between 1982 and 2000. The Landsat time series provided across-time progression and locations of the coral-dominated zones for all of Carysfort Reef. This study demonstrates the feasibility and utility of combining Landsat-5 TM and Landsat-7 ETM+ images for coral reef community scale change detection studies at a decadal scale. It opens the possibility of a cost-effective larger scale study, which could include an entire reef tract