TY - JOUR
T1 - Seafloor classification in a sand wave environment on the Dutch continental shelf using multibeam echosounder backscatter data
AU - Koop, Leo
AU - Amiri-Simkooei, Alireza
AU - van der Reijden, Karin J.
AU - O’Flynn, Sarah
AU - Snellen, Mirjam
AU - Simons, Dick G.
PY - 2019/3/1
Y1 - 2019/3/1
N2 - High resolution maps of sandy seafloors are valuable to understand seafloor dynamics, plan engineering projects, and create detailed benthic habitat maps. This paper presents multibeam echosounder backscatter classification results of the Brown Bank area of the North Sea. We apply the Bayesian classification method in a megaripple and sand wave area with significant slopes. Prior to the classification, corrections are implemented to account for the slopes. This includes corrections on the backscatter value and its corresponding incident angle. A trade-off in classification resolutions is found. A higher geo-acoustic resolution is obtained at the price of losing spatial resolution, however, the Bayesian classification method remains robust with respect to these trade-off decisions. The classification results are compared to grab sample particle size analysis and classified video footage. In non-distinctive sedimentary environments, the acoustic classes are not attributed to only the mean grain size of the grab samples but to the full spectrum of the grain sizes. Finally, we show the Bayesian classification results can be used to characterize the sedimentary composition of megaripples. Coarser sediments were found in the troughs and on the crests, finer sediments on the stoss slopes and a mixture of sediments on the lee slopes.
AB - High resolution maps of sandy seafloors are valuable to understand seafloor dynamics, plan engineering projects, and create detailed benthic habitat maps. This paper presents multibeam echosounder backscatter classification results of the Brown Bank area of the North Sea. We apply the Bayesian classification method in a megaripple and sand wave area with significant slopes. Prior to the classification, corrections are implemented to account for the slopes. This includes corrections on the backscatter value and its corresponding incident angle. A trade-off in classification resolutions is found. A higher geo-acoustic resolution is obtained at the price of losing spatial resolution, however, the Bayesian classification method remains robust with respect to these trade-off decisions. The classification results are compared to grab sample particle size analysis and classified video footage. In non-distinctive sedimentary environments, the acoustic classes are not attributed to only the mean grain size of the grab samples but to the full spectrum of the grain sizes. Finally, we show the Bayesian classification results can be used to characterize the sedimentary composition of megaripples. Coarser sediments were found in the troughs and on the crests, finer sediments on the stoss slopes and a mixture of sediments on the lee slopes.
KW - Acoustic backscatter
KW - Benthic habitats
KW - Marine geology
KW - Megaripples
KW - Multibeam echosounder
KW - Sand waves
KW - Seafloor mapping
KW - Sediment
UR - http://www.scopus.com/inward/record.url?scp=85065288527&partnerID=8YFLogxK
U2 - 10.3390/geosciences9030142
DO - 10.3390/geosciences9030142
M3 - Article
AN - SCOPUS:85065288527
VL - 9
JO - Geosciences (Switzerland)
JF - Geosciences (Switzerland)
IS - 3
M1 - 142
ER -