Acoustic seabed classification using QTC IMPACT on single-beam echo sounder data from the Norwegian Channel, northern North Sea

Abstract

Sediment mapping is important for understanding the physical processes, the impact of human activity, and the conditions for marine life on the seabed. For this purpose, the seabed classification tool QTC IMPACT analyses statistical variations in single-beam echo sounder data. QTC was applied in a large and physically diverse area of the Norwegian Channel, between 59°30′N and 61°N, to produce a new sediment map and to verify the QTC algorithm. The results were interpreted using ground truth (grain size analyses of 40 gravity cores and five grab samples), multi-beam echo sounder bathymetry (MBES), and seismo-acoustic profiles. Surficial sediments were divided into five classes: (1) mud and silt, (2) a variety of clay, silt and sand, (3) sandy mud with gravel, (4) sand with gravel, and (5) clay and sandy clay. Along the Norwegian coast, where MBES imagery shows evidence of glacial erosion, the surficial sediment distribution is variable. The echo shape analysis of QTC did not produce a natural partition of the data, and statistical assumptions did not always hold. Sediment classification was therefore sensitive to the choice of cluster algorithm. However, QTC produced the most physically plausible results on a large scale compared to other cluster algorithms. Class boundaries were consistent with supporting data. One exception is a transition from muddy to sandy sediments not visible in seismo-acoustic data. A possible explanation is that seabed fluid seepage and water current erosion cause sand particle transport into the western part of the channel. The study confirms the capability of QTC in a complex environment, but there are some possible improvements.