Title of article :
Spatially quantitative seafloor habitat mapping: example from the northern South Carolina inner continental shelf
Author/Authors :
Germ?n Y. Ojeda، نويسنده , , Paul T. Gayes، نويسنده , , Robert F. Van Dolah، نويسنده , , William C. Schwab، نويسنده ,
Issue Information :
روزنامه با شماره پیاپی سال 2004
Abstract :
Naturally occurring hard bottom areas provide the geological substrate that can support diverse assemblages of sessile benthic
organisms, which in turn, attract many reef-dwelling fish species. Alternatively, defining the location and extent of bottom sand
bodies is relevant for potential nourishment projects as well as to ensure that transient sediment does not affect reef habitats,
particularly in sediment-starved continental margins. Furthermore, defining sediment transport pathways documents the effects
these mobile bedforms have on proximal reef habitats. Thematic mapping of these substrates is therefore crucial in safeguarding
critical habitats and offshore resources of coastal nations. This study presents the results of a spatially quantitative mapping
approach based on classification of sidescan-sonar imagery. By using bottom video for image-to-ground control, digital image
textural features for pattern recognition, and an artificial neural network for rapid, quantitative, multivariable decision-making, this
approach resulted in recognition rates of hard bottom as high as 87%. The recognition of sand bottom was less successful (31%).
This approach was applied to a large (686 km2), high-quality, 2-m resolution sidescan-sonar mosaic of the northern South Carolina
inner continental shelf. Results of this analysis indicate that both surficial sand and hard bottoms of variable extent are present over
the study area. In total, 59% of the imaged area was covered by hard bottom, while 41% was covered by sand. Qualitative spatial
correlation between bottom type and bathymetry appears possible from comparison of our interpretive map and available
bathymetry. Hard bottom areas tend to be located on flat, low-lying areas, and sandy bottoms tend to reside on areas of positive
relief. Published bio-erosion rates were used to calculate the potential sediment input from the mapped hard bottom areas rendering
sediment volumes that may be as high as 0.8 million m3/yr for this portion of the South Carolina coast.
Keywords :
Habitat mapping , seafloor habitat , inner continental shelf , sidescan-sonar neural networks , South Carolina
Journal title :
Estuarine, Coastal and Shelf Science
Journal title :
Estuarine, Coastal and Shelf Science