Continental shelves are areas of extreme environmental and economic importance and priority areas for management and planning efforts. However, there is a knowledge gap regarding many tropical shelves, mainly due to high costs associated with surveying large marine areas with traditional techniques. Carbonatic, sediment-starved shelves are distributed across tropical and temperate regions and display characteristics that may allow the application of remote sensing methods, albeit the depth range found in these areas. Thus, this study tested the applicability of Landsat-8 imagery to detect bottom features in a sediment-starved shelf in North-eastern Brazil. The Brazilian Northeast Continental Shelf harbours a variety of geohabitats, including coral and algal reefs, beachrocks, mesophotic reefs, canyons, and paleochannels. Nine scenes were processed following the same steps: radiometric and atmospheric correction, masking, glint removal, water column correction and individual image adjustments. Two scenes were deemed best to portray targeted features. We detected 14 paleochannels in a 9700 km2 area, as well as the shelf break. A pixel-by-pixel analysis using all available bathymetric field data (n = 42,373 pixels) showed that 92.38% of the pixels were accurately classified. Misinterpreted pixels were either due to bordering regions incorrectly identified as paleochannels (3.99%) or paleochannels’ sections undetected in our final mosaic (3.62%). The final mosaic allowed a detailed view of submerged valleys (including previously unmapped channels) distributed up to areas deeper than 100 m using free of charge, medium-resolution satellite imagery. Some of these channels are known fishing grounds and fish spawning aggregation areas, making them priority zones for management and conservation efforts. To the best of our knowledge, this represents the first study applying Landsat imagery to map paleochannels in submerged areas, and most importantly, in areas deeper than 30 m. We believe this research can help focusing future mapping efforts and may advance the general knowledge of the area.

Depth range
0- 90 m

Mesophotic “mentions”
5 x (total of 5923 words)

Classification
* Presents original data
* Focused on 'mesophotic' depth range

Fields
Geomorphology
Methods and Technology

Locations
Brazil - Eastern Brazil