Hyperspectral identification of coral reef features

Remote sensing is regarded as an efficient and accurate tool for mapping an d monitoring changes in coral reef extent and: well being consistently over large geographic areas. However, several operational restrictions limit th e accuracy with which coral reefs can be monitored remotely. A primary rest riction is that the spectral responses of numerous features in the coral re ef environment are optically similar, which has the potential of contributi ng to misclassification errors. In 1996, in situ spectral reflectance data were collected in Fiji using an underwater cosine receptor and a 10 m under water fibre optic cable, which permitted sampling at depth while scuba divi ng. In 1997, in situ spectral reflectance measurements of exposed coral ree f features with little or no water cover were collected in Indonesia using the same radiometer, but a nonwaterproof remote cosine receptor. These spec tral datasets were compared and analysed to test the following hypotheses. First, geographic location does not affect the spectral reflectance charact eristics; second, the morphology of reef features does affect the spectral reflectance characteristics; third, bleached coral and healthy coral have d istinct spectral reflectance characteristics; and finally, a spectral refle ctance index will aid image classification. Results indicate that the spect ra measured in Fiji and those measured in Indonesia are statistically simil ar, so all spectra were merged into one large spectral dataset and principa l components analysis was used to determine the most representative spectra . Derivative spectroscopy was then used to conclude that spectral discrimin ation is indeed possible between 654 and 674 nm, between 582 and 686 nm and between 506 and 566 nm. The proportion of correctly identified spectra usi ng the three-step procedure of first derivatives is 75% with the main sourc e of error resulting from spectral variability of algae reflectance. The re sults of this feasibility study indicate that hyperspectral remote sensing of a coral reef environment will lead to accurate identification and subseq uent monitoring of changes in coral health and overall well being.