Global assessment of modern coral reef extent and diversity for regional science and management applications: a view from space Serge ANDRÉFOUËT 1,2,* , Frank E. MULLER-KARGER 2 , Julie A. ROBINSON 3 , Christine J. KRANENBURG 2 , Damaris TORRES-PULLIZA 2 , Steven A. SPRAGGINS 2,3 , and Brock MURCH 2 1 Institut de Recherche pour le Développement, BP A5, 98848 Nouméa, New Caledonia 2 University of South Florida, College of Marine Science, Institute for Marine Remote Sensing, 140, 7th Av. South, St Petersburg, FL 33701, USA 3 Earth Sciences and Image Analysis Laboratory, Johnson Space Center, 2101 NASA Parkway, Mail Code SA15, Houston, TX 77058, USA *Corresponding author: S. Andréfouët Fax: +687 26 43 26, e-mail: andrefou@noumea.ird.nc Abstract In 2001, the Millennium Coral Reef Mapping Project initiated a global compilation of 30-meter spatial resolution Landsat 7 ETM+ satellite images that provides unprecedented coverage of coral reefs worldwide. From this unique image data set, it is possible to create globally consistent map products to serve local, regional and global applications. The project has examined ~1,500 images to design a thematically rich (966 classes) geomorphological classification scheme used to interpret and map every single reef of the planet. Compared to previously existing databases, the maps provide greatly enhanced localization, inventory and spatial analysis of reef structures. Distributed as Geographic Information Systems layers beginning in late 2003, the map products, have been used for a variety of applications. We present an overview of the milestones of the project and the methods we followed to create the image archive and use it to develop geomorphological maps of the world’s coral reefs. We also present preliminary results obtained for a variety of applications—namely reef surface quantification, establishment of marine protected areas in Papua New Guinea and Eastern Caribbean, reef condition assessment in the Caribbean, morphometric analyses of Maldivian reefs and geochemical budgets in French Polynesian atolls. We believe that many scientists and managers will use the data and maps in a variety of applications and that the dataset will become a lasting foundation for understanding and conserving coral reefs around the world. Keywords remote sensing, mapping, geomorphology, Landsat ETM+ Introduction Current coral reef science and management programs seek understanding or protection of ecological processes that can be presented along a continuum of complex ecological scales—from individual coral reef organisms to coral reef biogeographic regions (Hatcher 1997). Small spatial scales and lower ecological organizational levels can be studied repetitively using experimental procedures, in aquaria, mesocosms, and in the field. However, at the other end of this continuum, regional (or global) processes are often handled in a limited way, by scaling up from a small number of samples of sites within a region, without actually capturing the full range of features and processes occurring there. This can be described as a pseudo-regional or pseudo-global approach. Remote sensing technology offers the most efficient means to uniformly observe and characterize an entire region without relying on sampling and extrapolation. For many regions, one of the most basic problems yet to be solved can be summarized in three questions: “Where are the coral reefs? How large are they? How are they connected with each other and the surrounding environment?” Providing accurate answers to these simple questions is a basic requirement for successful management. Whether estimating the state of reefs, estimating the risks that land-based or marine-based threats can cause to reefs, determining the monetary value of one region’s reefs, or designing a network of marine protected areas—all require accurate descriptions of reef system spatial structures and properties including geographic location, size, connectivity, topology, etc. There are now a number of remote sensing data sources that can be used to help answer these basic questions. One preliminary approach is to use existing global data sets at low spatial resolution (1-9 km) to make analysis simple (e.g. Stumpf et al. 2003). However, we believe that trying to observe reef structures “directly” (sensu Andréfouët and Riegl 2004) at a 1-km scale is a mismatch with the spatial scale of coral reef features—like trying to study the dynamics of a city like Miami or Paris with 100 km resolution data. Loss of information on reef spatial parameters increases rapidly as spatial resolution becomes more coarse (Andréfouët et al. 2002; Andréfouët et al. 2003a). Based on these studies, we believe that high 1732 Proceedings of 10th International Coral Reef Symposium, 1732-1745 (2006)