Boulders provide ecologically important hard grounds in shelf seas, and form protected habitats under the European Habitats Directive. Boulders on the seafloor can usually be recognized in backscatter mosaics due to a characteristic pattern of high backscatter intensity followed by an acoustic shadow. The manual identification of boulders on mosaics is tedious and subjective, and thus could benefit from automation. In this study, we train an object detection framework, RetinaNet, based on a neural network backbone, ResNet, to detect boulders in backscatter mosaics derived from a sidescan-sonar operating at 384 kHz. A training dataset comprising 4617 boulders and 2005 negative examples similar to boulders was used to train RetinaNet. The trained model was applied to a test area located in the Kriegers Flak area (Baltic Sea), and the results compared to mosaic interpretation by expert analysis. Some misclassification of water column noise and boundaries of artificial plough marks occurs, but the results of the trained model are comparable to the human interpretation. While the trained model correctly identified a higher number of boulders, the human interpreter had an advantage at recognizing smaller objects comprising a bounding box of less than 7 × 7 pixels. Almost identical performance between the best model and expert analysis was found when classifying boulder density into three classes (0, 1–5, more than 5) over 10,000 m2 areas, with the best performing model reaching an agreement with the human interpretation of 90%.
No abstract is provided for this article.
Subtidal hard substrate habitats are unique habitats in the marine environment. They provide crucial ecosystem services that are socially relevant, such as water clearance or as nursery space for fishes. With increasing marine usage and changing environmental conditions, pressure on reefs is increasing. All relevant directives and conventions around Europe include sublittoral hard substrate habitats in any manner. However, detailed specifications and specific advices about acquisition or delineation of these habitats are internationally rare although the demand for single object detection for e.g., ensuring safe navigation or to understand ecosystem functioning is increasing. To figure out the needs for area wide hard substrate mapping supported by automatic detection routines this paper reviews existing delineation rules and definitions relevant for hard substrate mapping. We focus on progress reached in German approval process resulting in first hydroacoustic mapping advices. In detail, we summarize present knowledge of hard substrate occurrence in the German North Sea and Baltic Sea, describes the development of hard substrate investigations and state of the art mapping techniques as well as automated analysis routines.
. Tsunami, storm and flash event layers, which have been deposited over the last century on the shelf offshore from Khao Lak (Thailand, Andaman Sea), are identified in sediment cores based on sedimentary structures, grain size compositions, Ti / Ca ratios and 210Pb activity. Individual offshore tsunami deposits are 12 to 30 cm in thickness and originate from the 2004 Indian Ocean tsunami. They are characterized by (1) the appearance of sand layers enriched in shells and shell debris, (2) cross lamination and (3) the appearance of rip-up clasts. Storm deposits found in core depths between 5 and 82 cm could be attributed to individual storm events by using 210Pb dating in conjunction with historical data of typhoons and tropical storms and could thus be securely differentiated from tsunami deposits. Massive sand layers enriched in shells and shell debris characterize the storm deposits. The last classified type of event layer represents flash floods, which is characterized by a fining-upward sequence of muddy sediment. The most distinct difference between the storm and tsunami deposits is the lack of rip-up clasts, mud, and terrigenous material within the storm deposits. Terrigenous material transported offshore during the tsunami backwash is therefore an important indicator to distinguish between offshore storm and tsunami deposits.
No abstract is provided for this article.
Dataset for the study "Can Anthipatella wollastoni be detected in Multibeam Echo Sounder multi-detect data?", currently under review at Frontiers in Remote Sensing. The files include: Photos.zip: GoPro Photos with ground truthing of two ridges with the occurrence of Black Corals. Photos have been geo-located using the coordinates of the onboard Multibeam System and correcting a time offset. Refer to the paper for details. Sound velocities: Sound velocity casts using a Base-X shallow water profiler used to correct the multibeam echo sounder. MD_manual_edit: Shape file including the position of Multi-Detects after the manual cleaning. The MD point objects have been joined with information from the local bathymetry and slope. Note that associated intensity values are erroneous due to a bug in the recording MBES firmware. MBES: Includes Norbit s7k (version 3) raw files of the multibeam echo sounder data (Norbit iwbms-e, Serial number #12). Files in folders with "MD" were used for the Multi-Detect study. Files with "MFE" include multifrequency data. We failed to locate the black corals in multi-frequency backscatter data. Multi-detect data were recorded with a swath width of 100° split into 512 beams. The multifrequency data contains 190 and 370 kHz information. Spreading correction was set to 40, absorption set to 107 dB/km (calculated for the 390 kHz frequency with a mean water temperature of 16°C, a salinity of 35. Actual temperature according to diver information was 24°C at the surface, 23° until 50 m and 21° until 80 m). All offsets have been accounted for during the survey. The data is located using an RTK correction (refer to paper for further details).
No abstract is provided for this article.
<p>The identification of marine cobbles and boulders (stones) based on acoustic remote sensing is important for the detection, delineation and for an ecological assessment of important seafloor habitats. Due to the large areas involved and the required high-resolution data, a manual interpretation is not feasible. In recent years, automated methods for stone detection were developed. However, these developments were only applied in comparatively small proof of concept areas, and a common barrier to practical implementation by authorities is the required upscaling. This case study aims to apply automated methods for boulder detection based on convolutional neural networks to larger areas, by identifying and validating boulder densities over several hundred km<sup>2 </sup>in the western Baltic Sea in acoustic backscatter data and derived datasets. The use of distributed training sites of less than 0.5 km<sup>2  </sup>in size is proposed to improve the model capacity to adapt to variations of boulder appearance in remote sensing data related to local geological variation and survey conditions. Distributed validation sites of similar size are suggested to provide quality control during reprocessing with adapted models. Current limitations for the automated identification of individual boulders in backscatter data are demonstrated, which can be caused by survey geometry, data quality or obstacles and seafloor with similar acoustic characteristics.</p>
No abstract is provided for this article.
<p>Sublittoral hard substrates, for example formed by blocks and boulders, are hotspots for marine biodiversity, especially for benthic communities. Knowledge on boulder occurrence is also important for marine and coastal management, including offshore wind parks and safety of navigation. The occurrence of boulders have to be reported by member states to the European Union. Typically, boulders are located by acoustic surveys with multibeam echo sounders and side scan sonars. The manual interpretation of these data is subjective and time consuming. This presentation reports on recent work concerned with the detection of boulders in different acoustic datasets by convolutional neural networks, highlighting current approaches, challenges and future opportunities.</p>
<p>Bottom trawling is a fishing technique in which a net held open by otter boards is dragged across the seafloor to harvest bottom living resources. This action induces high levels of stress to ecosystems by overturning boulders, disturbing and resuspending surface sediment, and plowing scars into the seabed. In the long term the trawling impact on benthic habitats becomes problematic when the time between trawls is shorter than the time it takes for the ecosystem to recover. Since quantitative information on the intensity of bottom fishing is particularly important but rarely available, our study is crucial to reveal the extent and magnitude of the anthropogenic impacts to the seafloor. As part of the MGF Baltic Sea project, a multibeam-echosounder was used to record high-resolution bathymetric data in a small, heavily fished focus area at a 1-year interval. Based on bathymetric data, we present an automated workflow for extracting trawlmark features from seafloor morphology and deriving parameters that qualitatively characterize trawlmark intensity. We also demonstrate how the seafloor surface of an exploited area develops within a year and what can be derived from this for regeneration indicators.</p>
No abstract is provided for this article.
Agadir Canyon is one of the largest submarine canyons in the World, supplying giant submarine sediment gravity flows to the Agadir Basin and the wider Moroccan Turbidite System. While the Moroccan Turbidite System is extremely well investigated, almost no data from...
The detection of boulders in hydroacoustic data is essential for a range of environmental, economic and marine planning applications. The manual interpretation of hydroacoustic data for object detection is a non-trivial, tedious and subjective task. Using the conventional means accessible to hydrographic professionals, it is nearly impossible to locate all boulders or rule out their presence for extended areas of interest. Although it has been shown that AI can do the job quickly and reproducibly, earlier work has not progressed beyond scientific experiments. As a result, AI software have not been routinely integrated into the workflows of institutions involved in hydrographic data acquisition and processing, or oceanographic analysis. This paper presents a workflow for fully automated boulder detection in hydroacoustic data. A graphical user interface enables training and evaluation of detection models, boulder detection model execution, and post-processing of detection results without programming. The workflow is demonstrated on data from the southern Baltic Sea. Validation results of the detection for various data inputs include a mAP-50 of 77.83 % for raster images of backscatter intensities based on side-scan sonar, a mAP-50 of 70.46 % for raster images of slope angles based on multibeam echosounder and a mAP-50 of 44.02 % for backscatter and bathymetric data given as 3D point clouds.
