The Application of Geographic Information System to the Coral Reef of Southern Okinawa

Geographic Information Systems (GIS) was used as a tool to study the spatial characteristics of the coral reefs of Southern Okinawa. The study involved the intensive use two main spatial analysis modules of the GIS namely the image processing module and the decision support system module. Image processing was undertaken on a Landsat satellite image to produce a marine benthic habitat map. This process utilized a depth compensation algorithm, and the unconventionally used Bayesian “soft and hard” supervised classification to produce not only a single thematic map, but a set of multiple maps showing the probability values for each of the categories selected. Each of these produced images was assessed for their accuracy with the results showing only moderate accuracy, which is typical in contemporary literature. These maps are intended to serve as baseline information that can be used for future spatial analysis processes. The next step involved the use of the images produced from the image processing together with other derived images, to undertake a spatial decision support analysis. The analysis utilized five spatial images as evidences (i.e., carbonate probability cover, sand probability cover, land areas, deep water areas and distance from Chibishi reefs image) to produce a suitability map showing the areas where coral larvae from Chibishi will most likely settle in the intertidal areas of Southern Okinawa. Another spatial evidence was added (i.e., chl. a image) to produce another map to showing the chances of surviving of these larvae. Results show that the bay areas in the eastern side of Okinawa as having very low probabilities of settlement and survivorship of larvae from Chibishi. It is argued that this might be one of the reasons why these areas with very good coral cover 30 years ago is now mostly just dead reefs. Key Words; Decision Support System, GIS and Environmental Modelling, Bayesian Probability, Okinawa, Marine Benthic Habitat Classification, Idrisi