William Skirving

Remote sensing of sea surface temperatures during 2002 Barrier Reef coral bleaching

Early in 2002, satellites of the U.S. National Oceanic and Atmospheric Administration (NOAA) detected anomalously high sea surface temperatures (SST) developing in the western Coral Sea, midway along Australias Great Barrier Reef (GBR). This was the beginning of what was to become the most significant GBR coral bleaching event on record [Wilkinson, 2002]. During this time, NOAAs National Environmental Satellite, Data, and Information Service (NESDIS) provided satellite data as part of ongoing collaborative work on coral reef health with the Australian Institute of Marine Science (AIMS) and the Great Barrier Reef Marine Park Authority (GBRMPA). These data proved invaluable to AIMS and GBRMPA as they monitored and assessed the development and evolution of SSTs throughout the austral summer, enabling them to keep stakeholders, government, and the general public informed and up to date.

Satellite observation of Keppel Islands (Great Barrier Reef) 2002 coral bleaching using IKONOS data

An examination of IKONOS satellite imagery of the Keppel Islands (Great Barrier Reef) acquired before and during a coral bleaching event indicates that severe bleaching of reefs can be detected as an increase in brightness in the band 1 (blue) and band 2 (green) IKONOS spectral bands (4-m resolution). The bleaching was not detected in band 3 (red), band 4 (near-infrared), or in the 1-m panchromatic band data. A total of 0.74xa0km2 of bleached coral was identified, with detection occurring in waters as deep as 15xa0m. The procedure requires that one of the scenes be radiometrically normalized to match the reference scene prior to image differencing. A relative radiometric normalization was used in this case because variable cloud cover present in the image acquired during the bleaching event prevented reliable modeling of atmospheric effects. The success at coral bleaching detection at Keppel Islands represents both a “best-case” and a “cloud-challenged” scenario. It was a best-case scenario in that coral cover was extensive (70–90% live coral cover, mostly acroporids) and the bleaching level was extreme (92–95% of coral cover white bleached). It was a cloud-challenged scenario in terms of having extensive and highly variable cloud cover present in the image acquired during the bleaching event. Color difference images reveal extensive areas of bleached coral at sites away from our study area, indicating that this platform and methodology may be a valuable tool for mapping high coral cover areas during bleaching events. Additional studies and technique refinements would be required to test the detection limits of bleaching with IKONOS imagery or to develop a spectrally based bleaching detection index.