No abstract

1. Intertidal macroalgal communities are economically and ecologically important and, with a likely increase in anthropogenic pressures, there is need to evaluate and monitor these diverse habitats. Efforts to conserve and sustainably manage these habitats must be underpinned by accurate, cost-effective, and efficient data collection methods. The high spatial and temporal resolution of unmanned aerial vehicles (UAVs), compared with satellites and aircraft, combined with the development of lightweight sensors, provides researchers with a valuable set of tools to research intertidal macroalgal communities.2. The ability of multispectral sensors, mounted on a satellite, an aircraft, and a UAV, to identify and accurately map the intertidal brown fucoid Ascophyllum nodosum (Fucales, Ochrophyta) at a site with a low species diversity of macroalgae were compared.3. Visual analysis confirmed that the spatial resolution of satellite imagery was too coarse to map intertidal macroalgae as it could not capture the fine spatial patterns of the macroalgal community. High-resolution RGB (colour) imagery, taken during the aircraft and UAV surveys, was used to collect training and reference data through the visual identification and digital delineation of species. Classes were determined based on the level of taxonomic detail that could be observed, with higher levels of taxonomic detail observed in the UAV imagery over the aircraft imagery. Data from both were used to train a maximum-likelihood classifier (MLC).4. The UAV imagery was able to more accurately classify a distinct A. nodosum class, along with other macroalgal and substratum classes (overall accuracy, OA, 92%), than the aerial imagery, which could only identify a lower taxonomic resolution of mixed A. nodosum and fucoid class, achieving a lower OA (78.9%). This study has demonstrated that in a coastal site with low macroalgal species diversity, and despite the spectral similarity of macroalgal species, UAV-mounted multispectral sensors proved the most accurate for focused assessments of individual canopyforming species.

ABSTRACT:The use of remote sensing has increased greatly in recent years due to technological advances and its advantages in comparison with traditional methods. In the case of Ireland however the use of these techniques is not well established and only 17% of remote sensing studies are related to marine and coastal environments. As a first step, and taking into account that fisheries and aquaculture plays an important economic and social role in Ireland, a database of Sea Surface Temperature (SST) and Chlorophyll-a (Chl-a) relating to the ICES fisheries management areas is being generated. Up to now three different products have been produced. These products correspond to the annual SST Climatology and annual SST Anomalies from 1982 to 2014, as well as the annual Chl-a Climatology taking into account the different life span

In this study we applied for the first time Fully Convolutional Neural Networks (FCNNs) to a marine bathymetric dataset to derive morphological classes over the entire Irish continental shelf. FCNNs are a set of algorithms within Deep Learning that produce pixel-wise classifications in order to create semantically segmented maps. While they have been extensively utilised on imagery for ecological mapping, their application on elevation data is still limited, especially in the marine geomorphology realm. We employed a high-resolution bathymetric dataset to create a set of normalised derivatives commonly utilised in seabed morphology and habitat mapping that include three bathymetric position indexes (BPIs), the vector ruggedness measurement (VRM), the aspect functions and three types of hillshades. The class domains cover ten or twelve semantically distinct surface textures and submarine landforms present on the shelf, with our definitions aiming for simplicity, prevalence and distinctiveness. Sets of 50 or 100 labelled samples for each class were used to train several U-Net architectures with ResNet-50 and VGG-13 encoders. Our results show a maximum model precision of 0.84 and recall of 0.85, with some classes reaching as high as 0.99 in both. A simple majority (modal) voting combining the ten best models produced an excellent map with overall F1 score of 0.96 and class precisions and recalls superior to 0.87. For target classes exhibiting high recall (proportion of positives identified), models also show high precision (proportion of correct identifications) in predictions which confirms that the underlying class boundary has been learnt. Derivative choice plays an important part in the performance of the networks, with hillshades combined with bathymetry providing the best results and aspect functions and VRM leading to an overall deterioration of prediction accuracies. The results show that FCNNs can be successfully applied to the seabed for a morphological exploration of the dataset and as a baseline for more in-depth habitat mapping studies. For example, prediction of semantically distinct classes as “submarine dune” and “bedrock outcrop” can be precise and reliable. Nonetheless, at present state FCNNs are not suitable for tasks that require more refined geomorphological classifications, as for the recognition of detailed morphogenetic processes.

Through Ireland's national seabed mapping programme, Integrated Mapping for the Sustainable Development of Ireland's Marine Resource (INFOMAR), the collaboration between Geological Survey Ireland and the Marine Institute continues to comprehensively map Ireland's marine territory in high resolution. Through its work, the programme builds on earlier Irish seabed mapping efforts, including the Irish National Seabed Survey project in producing seabed mapping products that support Ireland's blue economy, European marine policy and international efforts to understand our global oceans. INFOMAR uses a variety of marine technologies to deliver accurate bathymetric maps and useful data products to end users through a free and open source licensing agreement. To reflect the diversity of applications these data products serve, a series of four case studies are presented here focusing on marine geophysical and geological data from locations within Ireland's marine territories. The case studies illustrate how data generated through seabed mapping may be interpreted to directly impact the generation of blue knowledge across a variety of marine environments ranging from shallow coastal and shelf waters to the deep oceanic depths of the continental slope of Ireland's marine area. The impact of Ireland's seabed mapping efforts is further considered in the context of national, European and international initiatives where Ireland's marine knowledge resource is leveraged to deliver positive benefit to the programme's stakeholders.