da Silveira, C.B.L.; Strenzel, G.M.R.; Maida, M.; Araújo, T.C.M., and Ferreira, B.P., 2020. Multiresolution satellite-derived bathymetry in shallow coral reefs: Improving linear algorithms with geographical analysis. Journal of Coastal Research, 36(6), 1247–1265. Coconut Creek (Florida), ISSN 0749-0208.
Bathymetric maps are one of the first steps for most hydrological and ecological studies of the seascape, as depth is a determinant factor in the distribution of organisms, patterns of wave exposure, and coastal circulation. The coral reefs of Tamandaré-Brazil, located at Costa dos Corais marine protected area (MPA), encompass a mosaic of interconnected habitats of complex geomorphology, including coral reefs, algal, and seagrass beds. These coastal habitats are subjected to chronic impacts such as sedimentation, reef erosion, and increasing human use, leading to habitat loss. Despite their social and ecological importance and the conservation measures in place, bathymetric and habitat maps of this coast are lacking. Indeed, in situ surveys are not always feasible in shallow coral reef areas. The present study offers a detailed bathymetric mapping of the area using multiresolution satellite imagery. One Landsat-8 (December 2016) and one WorldView-03 (February 2017) imagery were used to derive medium (30-m) and high-resolution (2-m) bathymetry of the study area. Single-beam echo sounder surveys were performed to obtain field data to calibrate the depth-retrieving algorithms: linear model, ratio band model, principal component analyses of the transformed bands, and geographically weighted regressions (GWR). For both resolution datasets, results showed that the algorithms' RMS were significantly improved by the GWR technique (RMS > 0.9) because of adaptability to the bottom heterogeneity found in complex areas such as coral reefs. Specific geomorphological reef zones were recognizable in the resulting bathymetric maps, such as intrareef lagoon, reef crest, fore-reef, and reef flat. This research concluded that affordable methods such as single-beam data coupled with satellite imagery through GWR can provide the required inputs for mapping shallow areas with complex relief. Such results may be further used to habitat mapping, necessary to inform the multiple-use zonation foreseen in MPA management plans.