Acoustic analysis of seabed at the Ten Noorden van de Waddeneilanden wind farm zone
In this project the seabed of the Ten Noorden van de Waddeneilanden (TNW) wind farm zone (North Sea, Netherlands Continental Shelf) has been investigated towards indicators for presence of biogenic reef structures and the distribution of sediments based on acoustic data. The area is located about 60 km north from the Wadden Sea Islands and covers an area of around 25 km by 6 km with a relatively constant water depth ranging between 34 and 37 m (LAT). This area of TNW has been indicated as area for offshore wind. The actual wind farm is not built yet and the project aims for an analysis of the seabed state before the wind farm will be operational. This provides a reference of the seabed state for future studies.
The multibeam echosounder (MBES) bathymetry and backscatter data as well as the side-scan sonar (SSS) backscatter data were acquired between July and November 2019 by the company MMT Sweden AB. In addition, Van-Veen grab samples were taken in August 2019 to ground truth the seabed sediments. This has been executed by order of RVO.
The MBES bathymetry data revealed in general a flat seabed with a gentle decrease in the water depth from the west towards the east. The slope gradients are pre-dominantly below 1°. The majority of the seabed possess no bedforms and only in the most eastern part widely-spaced sediment banks and mega ripples on the western flanks of the banks are observed. The banks have widths of approximately 400 m and ranges in height between 0.3 to 0.8 m. They are situated unevenly across the area with distances between 400 and 1800 m (MMT 2020). On the broad banks in the south-east of the study area, small valleys are present reaching depths of around 40 cm. These valleys are also the most interesting location in terms of biogenic reefs. The MBES and SSS backscatter data show high backscatter within the valleys. In addition, the high-resolution SSS images present an alternating pattern of high and low backscatter values, which could be an indicator for the presence of reefs. Another possible cause for the alternating pattern could be small ripples. However, angular response curves (ARC) calculated from the MBES backscatter indicate very rough material, for example the accumulation of shells, gravel or even biogenic reefs. That means, it would be rather unusual that ripples are formed extensively on such a seabed. In order to proof the presence of biogenic reefs located within the valleys on the broad banks and eventually verify the acoustic indicators, a ground truthing campaign with video recordings is recommended. Another interesting observation west of the sand banks, are small patches of high acoustic classes surrounded by low acoustic classes. This is an acoustic indication for the accumulation of coarser material within an area of finer material. It is likely that the coarser material corresponds to an increased accumulation of shell fragments.
However, the presence of reef-building species could also be a possible explanation. Ground truthing is recommend to prove.
The majority of the seabed samples were classified as clayey to silty fine to medium sand with predominantly sand as the major constitute. However, at two locations clay was found as the major constitute and at one location medium to coarse sand was observed. The d50-value of the samples range from 80 to 400 μm. A general observation is that the seabed samples are coarser and contain less mud in the west of the study area. Considering the MBES backscatter range of 13 dB and backscatter models, showing a difference of 10 dB between clay and coarse sand (APL (1994), Gaida (2020)), the measured MBES backscatter reflects the variation observed in the sediment samples very well. Based on the MBES measurements at the sample locations, the MBES backscatter allows to distinguish medium to coarse sand from fine to medium sand, but does not clearly distinguish between clay and fine to medium sand. The MBES bathymetry has in general a better correlation with the sediment samples. Undetected shell fragments in the sample, or the small content of gravel in the fine sediment samples, can have a high influence on the backscatter, and could results in the low correlation between backscatter and the fine sediment samples. Therefore, we still argue that the MBES backscatter and the acoustic classification maps are still suitable as an indicator for the sediment distribution.
In general, two distinctive areas can be distinguished in the TNW wind farm zone. The east and middle part consist of finer sediments (clay to fine, medium sand) indicated by low to medium backscatter and the broad banks covered with coarser sand indicated by the high backscatter. As mentioned above the valleys on the broad banks contain even coarser sediments, where a chance of the presence of biogenic reefs exists. Based purely on the backscatter based classification, the western part of the study area is further divided into a circular patch of low acoustic backscatter, indicative for more clayish sediments, and the area surrounding the circular patch with medium backscatter, indicating more finer sands.
Since the bathymetry showed a better correlation with the sediment samples, we further used both layers as an input in the acoustic classification. Based on the resulting map, the circular patch in the west is much smaller and the entire area appears to be even more homogeneous. To finally assess the accuracy of both maps, more grab samples would be needed.
The full report can be read here.