Explore the vast range of titles available.

Detailed seabed substrate maps are increasingly in demand for effective planning and management of marine ecosystems and resources. It has become common to use remotely sensed multibeam echosounder data in the form of bathymetry and acoustic backscatter in conjunction with ground-truth sampling data to inform the mapping of seabed substrates. Whilst, until recently, such data sets have typically been classified by expert interpretation, it is now obvious that more objective, faster and repeatable methods of seabed classification are required. This study compares the performances of a range of supervised classification techniques for predicting substrate type from multibeam echosounder data. The study area is located in the North Sea, off the north-east coast of England. A total of 258 ground-truth samples were classified into four substrate classes. Multibeam bathymetry and backscatter data, and a range of secondary features derived from these datasets were used in this study. Six supervised classification techniques were tested: Classification Trees, Support Vector Machines, k-Nearest Neighbour, Neural Networks, Random Forest and Naive Bayes. Each classifier was trained multiple times using different input features, including i) the two primary features of bathymetry and backscatter, ii) a subset of the features chosen by a feature selection process and iii) all of the input features. The predictive performances of the models were validated using a separate test set of ground-truth samples. The statistical significance of model performances relative to a simple baseline model (Nearest Neighbour predictions on bathymetry and backscatter) were tested to assess the benefits of using more sophisticated approaches. The best performing models were tree based methods and Naive Bayes which achieved accuracies of around 0.8 and kappa coefficients of up to 0.5 on the test set. The models that used all input features didn't generally perform well, highlighting the need for some means of feature selection.

Although the deep-sea floor accounts for approximately 60 % of Earth's surface, there has been little progress in relation to deriving maps of seafloor sediment distribution based on transparent, repeatable, and automated methods such as machine learning. A new digital map of the spatial distribution of seafloor lithologies below 500 m water depth is presented to address this shortcoming. The lithology map is accompanied by estimates of the probability of the most probable class, which may be interpreted as a spatially explicit measure of confidence in the predictions, and probabilities for the occurrence of five lithology classes (calcareous sediment, clay, diatom ooze, lithogenous sediment, and radiolarian ooze). These map products were derived by the application of the random-forest machine-learning algorithm to a homogenised dataset of seafloor lithology samples and global environmental predictor variables that were selected based on the current understanding of the controls on the spatial distribution of deep-sea sediments. It is expected that the map products are useful for various purposes including, but not limited to, teaching, management, spatial planning, design of marine protected areas, and as input for global spatial predictions of marine species distributions and seafloor sediment properties. The map products are available at https://doi.org/10.1594/PANGAEA.911692 (Diesing, 2020).

There is a need for fit-for-purpose maps for accurately depicting the types of seabed substrate and habitat and the properties of the seabed for the benefits of research, resource management, conservation and spatial planning. The aim of this study is to determine whether it is possible to predict substrate composition across a large area of seabed using legacy grain-size data and environmental predictors. The study area includes the North Sea up to approximately 58.44°N and the United Kingdom's parts of the English Channel and the Celtic Seas. The analysis combines outputs from hydrodynamic models as well as optical remote sensing data from satellite platforms and bathymetric variables, which are mainly derived from acoustic remote sensing. We build a statistical regression model to make quantitative predictions of sediment composition (fractions of mud, sand and gravel) using the random forest algorithm. The compositional data is analysed on the additive log-ratio scale. An independent test set indicates that approximately 66% and 71% of the variability of the two log-ratio variables are explained by the predictive models. A EUNIS substrate model, derived from the predicted sediment composition, achieved an overall accuracy of 83% and a kappa coefficient of 0.60. We demonstrate that it is feasible to spatially predict the seabed sediment composition across a large area of continental shelf in a repeatable and validated way. We also highlight the potential for further improvements to the method.

Cold-water coral reefs are rich, yet fragile ecosystems found in colder oceanic waters. Knowledge of their spatial distribution on continental shelves, slopes, seamounts and ridge systems is vital for marine spatial planning and conservation. Cold-water corals frequently form conspicuous carbonate mounds of varying sizes, which are identifiable from multibeam echosounder bathymetry and derived geomorphometric attributes. However, the often-large number of mounds makes manual interpretation and mapping a tedious process. We present a methodology that combines image segmentation and random forest spatial prediction with the aim to derive maps of carbonate mounds and an associated measure of confidence. We demonstrate our method based on multibeam echosounder data from Iverryggen on the mid-Norwegian shelf. We identified the image-object mean planar curvature as the most important predictor. The presence and absence of carbonate mounds is mapped with high accuracy. Spatially-explicit confidence in the predictions is derived from the predicted probability and whether the predictions are within or outside the modelled range of values and is generally high. We plan to apply the showcased method to other areas of the Norwegian continental shelf and slope where multibeam echosounder data have been collected with the aim to provide crucial information for marine spatial planning.

High alkalinity values observed in coastal seas promote the uptake of CO 2 from the atmosphere. However, the alkalinity budget of coastal areas and marginal seas is poorly understood, even though some of the recently observed alkalinity enhancement can be ascribed to riverine fluxes and anaerobic processes in shelf sediments. Here, we investigate the alkalinity budget of the Baltic Sea to identify previously unrecognized alkalinity sources. We quantify the generation of alkalinity and dissolved calcium (Ca) in this marginal sea applying simple mass balance calculations. Using this approach, we identify alkalinity and Ca sources of approximately 324 Gmol yr -1 and 122 Gmol yr -1 , respectively, that cannot be ascribed to the riverine input. The magnitude of the Ca source suggests that a major fraction of the excess alkalinity (244 Gmol yr -1 ) is induced by the dissolution of calcium carbonate (CaCO 3 ). A review of available field data shows that carbonate-bearing rocks at the coast and the seabed of the Baltic Sea are rapidly eroded and may provide sufficient CaCO 3 to close the Ca budget. Hence, dissolution of eroded CaCO 3 is the most likely source for the Ca enrichment observed in Baltic Sea water. This hypothesis is supported by mass accumulation rates of sediments derived from radioisotope data that are evaluated to derive a basin-wide rate of mud to muddy sand accumulation at the bottom of the Baltic Sea. The resulting value (139 Tg yr -1 ) exceeds current estimates of riverine particle fluxes into the Baltic Sea by more than one order of magnitude and confirms that rates of till erosion are sufficiently high to account for the Ca and most of the alkalinity excess in Baltic Sea water. Finally, we show that deliberate addition of CaCO 3 to sediments deposited in the Baltic Sea could neutralize significant amounts of CO 2 and help to achieve net-zero greenhouse gas emissions in the Baltic region.

Sediment maps developed from categorical data are widely applied to support marine spatial planning across various fields. However, deriving maps independently of sediment classification potentially improves our understanding of environmental gradients and reduces issues of harmonising data across jurisdictional boundaries. As the groundtruth samples are often measured for the fractions of mud, sand and gravel, this data can be utilised more effectively to produce quantitative maps of sediment composition. Using harmonised data products from a range of sources including the European Marine Observation and Data Network (EMODnet), spatial predictions of these three sediment fractions were generated for the north-west European continental shelf using the random forest algorithm. Once modelled these sediment fraction maps were classified using a range of schemes to show the versatility of such an approach, and spatial accuracy maps were generated to support their interpretation. The maps produced in this study are to date the highest resolution quantitative sediment composition maps that have been produced for a study area of this extent and are likely to be of interest for a wide range of applications such as ecological and biophysical studies.

Continental margin sediments are important sinks for organic carbon, thus providing valuable climate regulating functions. Human activities such as mobile bottom fishing might, however, compromise the effectiveness of the seabed in accumulating and storing organic carbon through a net increase of organic carbon remineralisation. While there currently is no consensus on the magnitude of this effect, it can be postulated that seabed areas characterised by high rates of organic carbon accumulation, high organic carbon stocks and a large fraction of labile organic matter might be most vulnerable to mobile bottom fishing. Conversely, areas characterised by low or no organic carbon accumulation, low stocks and a small fraction of labile organic matter might be of least concern. Based on reactivity measurements we spatially predict the thermal reactivity of sedimentary organic matter on the Norwegian continental margin. Surface sediments contain between 13% and 34% labile organic matter, with the highest values attained in the central Barents Sea. Using these results and previously developed maps of organic carbon stocks and accumulation rates, we then carry out a regionalisation using unsupervised classification with the aim to identify distinct seabed areas. We identify hotspots with regard to accumulation, storage and lability in the Skagerrak and the central Barents Sea and coldspots on the North Sea plateau, shelf banks in the Norwegian and southern Barents Seas and along the continental shelf break. Our results have the potential to be used in the ecosystem-based management of marine areas in Norway, and our methodology provides a template for similar analyses in other sea areas.

Seabed sediment composition is an important component of benthic habitat and there are many approaches for producing maps that convey sediment information to marine managers. Random Forest is a popular statistical method for thematic seabed sediment mapping using both categorical and quantitative supervised modelling approaches. This study compares the performance and qualities of these Random Forest approaches to predict the distribution of fine-grained sediments from grab samples as one component of a multi-model map of sediment classes in Frobisher Bay, Nunavut, Canada. The second component predicts the presence of coarse substrates from underwater video. Spatial and non-spatial cross-validations were conducted to evaluate the performance of categorical and quantitative Random Forest models and maps were compared to determine differences in predictions. While both approaches seemed highly accurate, the non-spatial cross-validation suggested greater accuracy using the categorical approach. Using a spatial cross-validation, there was little difference between approaches—both showed poor extrapolative performance. Spatial cross-validation methods also suggested evidence of overfitting in the coarse sediment model caused by the spatial dependence of transect samples. The quantitative modelling approach was able to predict rare and unsampled sediment classes but the flexibility of probabilistic predictions from the categorical approach allowed for tuning to maximize extrapolative performance. Results demonstrate that the apparent accuracies of these models failed to convey important differences between map predictions and that spatially explicit evaluation strategies may be necessary for evaluating extrapolative performance. Differentiating extrapolative from interpolative prediction can aid in selecting appropriate modelling methods.

A carbon budget for the northwest European continental shelf seas (NWES) was synthesised using available estimates for coastal, pelagic and benthic carbon stocks and flows. Key uncertainties were identified and the effect of future impacts on the carbon budget were assessed. The water of the shelf seas contains between 210 and 230 Tmol of carbon and absorbs between 1.3 and 3.3 Tmol from the atmosphere annually. Off-shelf transport and burial in the sediments account for 60-100% and 0-40% of carbon outputs from the NWES, respectively. Both of these fluxes remain poorly constrained by observations and resolving their magnitudes and relative importance is a key research priority. Pelagic and benthic carbon stocks are dominated by inorganic carbon. Shelf sediments contain the largest stock of carbon, with between 520 and 1600 Tmol stored in the top 0.1 m of the sea bed. Coastal habitats such as salt marshes and mud flats contain large amounts of carbon per unit area but their total carbon stocks are small compared to pelagic and benthic stocks due to their smaller spatial extent. The large pelagic stock of carbon will continue to increase due to the rising concentration of atmospheric CO2, with associated pH decrease. Pelagic carbon stocks and flows are also likely to be significantly affected by increasing acidity and temperature, and circulation changes but the net impact is uncertain. Benthic carbon stocks will be affected by increasing temperature and acidity, and decreasing oxygen concentrations, although the net impact of these interrelated changes on carbon stocks is uncertain and a major knowledge gap. The impact of bottom trawling on benthic carbon stocks is unique amongst the impacts we consider in that it is widespread and also directly manageable, although its net effect on the carbon budget is uncertain. Coastal habitats are vulnerable to sea level rise and are strongly impacted by management decisions. Local, national and regional actions have the potential to protect or enhance carbon storage, but ultimately global governance, via controls on emissions, has the greatest potential to influence the long-term fate of carbon stocks in the northwestern European continental shelf.

The role of continental margin sediments in the carbon cycle and the associated management potential for climate mitigation are currently poorly understood. Previous work has indicated that margin sediments store significant amounts of organic carbon, but few studies have quantified the rates at which organic carbon is accumulated. Here, we use machine learning to make spatial predictions of the organic carbon stocks and accumulation rates of sediments on the Norwegian continental margin. We show that surface sediments (upper 10 cm) store 814 Tg and accumulate 6 Tg yr−1 of organic carbon. Shelf-incised glacial troughs account for 39% of the stocks and 48% of the accumulation, with the main accumulation hotspot located in the Skagerrak. Continental margin sediments accumulate organic carbon at scales much larger than vegetated coastal ecosystems in Norway because of their larger extent. Future studies should explore to what extent management interventions could increase accumulation rates, e.g., by minimising anthropogenic disturbance of seafloor sediments.