The geometric accuracy of tens of millions of scenes of medium-resolution remote sensing (RS) images collected in the past 45 years has been systematically evaluated for land scenes, but the accuracy of ocean scenes is poorly known due to the lack of ground control points (GCPs). In this study, the locations of offshore platforms are first derived from time-series of Landsat-8 OLI images, and are then used as offshore reference points to systematically assess the geometric performance of RS images covering offshore oil/gas development areas. An inventory of 16,131 offshore platforms at the global scale is established, and then a novel method using the position-invariant characteristic of offshore platforms and the coherent characteristic of the geometric shift among tie-points (i.e. between sensed points from to-be-assessed images and the corresponding OLI-derived reference points) is developed for assessing the geometric accuracy of Landsat and other RS images. The method has been applied to 112,935 Landsat scenes (~1.87% of the entire archive) over oceans. The results indicate an optimal performance of Landsat OLI images (both pre-collection and Collection-1) but a less reliable performance of Landsat TM/ETM+ L1TP images. Approximately 50% of TM L1GS and ETM+ L1GT images have at least 2 pixels of geometric error. The new reference points inventory and the developed method were also applied to many other low-resolution and finer-resolution imagery (e.g. VIIRS Night-fire product, Terra/Aqua MODIS active fire product, ENVISAT ASAR, ALOS-1 PALSAR, Sentinel-1 SAR, Sentinel-2 MSI, the National Agriculture Imagery Program (NAIP) aerial images, and images from several Chinese satellites), and a quantitative description of the geometric accuracy of these sensors is also presented. The findings suggest that the new offshore reference point inventory is probably useful to help establish more robust offshore GCPs for U.S. Geological Survey (USGS) GCP library and further improve the ongoing USGS Global GCP improvement plan and European Space Agency Global Reference Image plan.
Remote Sensing and GIS
Coral reefs are among the most biodiverse ecosystems on Earth in large part owing to their unique three-dimensional (3D) structure, which provides niches for a variety of species. Metrics of structural complexity have been shown to correlate with the abundance and diversity of fish and other marine organisms, but they are imperfect representations of a surface that can oversimplify key structural elements and bias discoveries. Moreover, they require researchers to make relatively uninformed guesses about the features and spatial scales relevant to species of interest. This paper introduces a machine-learning method for automating inferences about fish abundance from reef 3D models. It demonstrates the capacity of a convolutional neural network (ConvNet) to learn ecological patterns that are extremely subtle, if not invisible, to the human eye. It is the first time in the literature that no a priori assumptions are made about the bathymetry–fish relationship.
Seagrass meadows are one of the most important coastal habitats across the globe. These are mainly constituted by the marine plants of the genus Posidonia and Thalassia. In the Mediterranean Sea, Posidonia oceanica is the dominant endemic plant that affects physical, biogeochemical, and biological processes. The decline in the spatial distribution has been attributed to excessive anthropic pressures and other large-scale environmental changes. The monitoring of the spatial distribution requires an update and accurate seagrass meadows delineation, i.e. meadow edge marking with a replicable method. The present study aims to present an approach to support the coastal marine habitat mapping, under the scheme of the Natura 2000 network using very high resolution Earth observation data and to prove that satellite images can be used for the mapping of the deep limits of the seagrass meadows. Pixel-based classification and object-oriented image analysis have been implemented for the image classification. Pixel-based Support Vector Machines and object-based Nearest Neighbor classifiers provided the best results with an overall accuracy of more than 90%, while deep limits have been successfully identified and separated from the deep waters.
The aim of this research was to propose and evaluate a methodological approach to integration and spatial data analysis in order to generate information towards a participatory site selection for bivalve marine aquaculture in the Baía Sul, Florianópolis, Santa Catarina, Brazil. For this purpose, the Baía Sul was investigated considering an ecosystem approach for aquaculture leading to an assessment of its potential for marine aquaculture. The planning of the aquaculture parks was made through a participatory process to incorporate both environmental carrying capacity and social carrying capacity. Experts and modellers developed a GIS model to assess the potential for marine aquaculture in Baía Sul. Continuous (unclassified) maps were used to provide spatial information about the variation of the potential for marine aquaculture in the Baía Sul. The maps were used to plan 53 aquaculture parks over the Baía Sul. The site selection of the parks was made in six public hearings attended by 403 stakeholders from 38 institutions representing different sectors with diverse interests in coastal zone. The results showed that although the Baía Sul is suitable for the growth of bivalve molluscs, some hydrodynamic characteristics and the influence of urbanization constitute a sanitary risk for the activity. Experts, modellers and stakeholders had a different perception about the importance of criteria in the aquaculture parks site selection. While the experts and modellers considered the environmental criteria as the most important aspect to locate the aquaculture parks, the stakeholders took into account mainly the logistics. The final result of the aquaculture parks location, approved by the Brazilian Ministry of Fisheries and Aquaculture (MPA), adopted the site selection by the stakeholders, providing aquaculture parks in areas with sanitary risk for the bivalve cultivation. The main advantage of the adopted assessment strategy was to identify the divergence between experts, modellers and the stakeholders and the distance that still exist between scientist and decision makers in Brazil.
Increased loads of land-based pollutants are a major threat to coastal-marine ecosystems. Identifying the affected marine areas and the scale of influence on ecosystems is critical to assess the impacts of degraded water quality and to inform planning for catchment management and marine conservation. Studies using remotely-sensed data have contributed to our understanding of the occurrence and influence of river plumes, and to our ability to assess exposure of marine ecosystems to land-based pollutants. However, refinement of plume modeling techniques is required to improve risk assessments. We developed a novel, complementary, approach to model exposure of coastal-marine ecosystems to land-based pollutants. We used supervised classification of MODIS-Aqua true-color satellite imagery to map the extent of plumes and to qualitatively assess the dispersal of pollutants in plumes. We used the Great Barrier Reef (GBR), the world's largest coral reef system, to test our approach. We combined frequency of plume occurrence with spatially distributed loads (based on a cost-distance function) to create maps of exposure to suspended sediment and dissolved inorganic nitrogen. We then compared annual exposure maps (2007–2011) to assess inter-annual variability in the exposure of coral reefs and seagrass beds to these pollutants. We found this method useful to map plumes and qualitatively assess exposure to land-based pollutants. We observed inter-annual variation in exposure of ecosystems to pollutants in the GBR, stressing the need to incorporate a temporal component into plume exposure/risk models. Our study contributes to our understanding of plume spatial–temporal dynamics of the GBR and offers a method that can also be applied to monitor exposure of coastal-marine ecosystems to plumes and explore their ecological influences.
The increasing demand for protein from aquaculture will trigger a global expansion of the sector in coastal and offshore waters. While contributing to food security, potential conflicts with other traditional activities such as fisheries or tourism are inevitable, thus calling for decision-support tools to assess aquaculture planning scenarios in a multi-use context. Here we introduce the AquaSpace tool, one of the first Geographic Information System (GIS)-based planning tools empowering an integrated assessment and mapping of 30 indicators reflecting economic, environmental, inter-sectorial and socio-cultural risks and opportunities for proposed aquaculture systems in a marine environment. A bottom-up process consulting more than 350 stakeholders from 10 countries across southern and northern Europe enabled the direct consideration of stakeholder needs when developing the GIS AddIn. The AquaSpace tool is an open source product and builds in the prospective use of open source datasets at a European scale, hence aiming to improve reproducibility and collaboration in aquaculture science and research. Tool outputs comprise detailed reports and graphics allowing key stakeholders such as planners or licensing authorities to evaluate and communicate alternative planning scenarios and to take more informed decisions. With the help of the German North Sea case study we demonstrate here the tool application at multiple spatial scales with different aquaculture systems and under a range of space-related development constraints. The computation of these aquaculture planning scenarios and the assessment of their trade-offs showed that it is entirely possible to identify aquaculture sites, that correspondent to multifarious potential challenges, for instance by a low conflict potential, a low risk of disease spread, a comparable high economic profit and a low impact on touristic attractions. We believe that a transparent visualisation of risks and opportunities of aquaculture planning scenarios helps an effective Marine Spatial Planning (MSP) process, supports the licensing process and simplifies investments.
Remote sensing techniques are currently the main methods providing elevation data used to produce Digital Terrain Models (DTM). Terrain attributes (e.g. slope, orientation, rugosity) derived from DTMs are commonly used as surrogates of species or habitat distribution in ecological studies. While DTMs’ errors are known to propagate to terrain attributes, their impact on ecological analyses is however rarely documented. This study assessed the impact of data acquisition artefacts on habitat maps and species distribution models. DTMs of German Bank (off Nova Scotia, Canada) at five different spatial scales were altered to artificially introduce different levels of common data acquisition artefacts. These data were used in 615 unsupervised classifications to map potential habitat types based on biophysical characteristics of the area, and in 615 supervised classifications (MaxEnt) to predict sea scallop distribution across the area. Differences between maps and models built from altered data and reference maps and models were assessed. Roll artefacts decreased map accuracy (up to 14% lower) and artificially increased models’ performances. Impacts from other types of artefacts were not consistent, either decreasing or increasing accuracy and performance measures. The spatial distribution of habitats and spatial predictions of sea scallop distributions were always affected by data quality (i.e. artefacts), spatial scale of the data, and the selection of variables used in the classifications. This research demonstrates the importance of these three factors in building a study design, and highlights the need for error quantification protocols that can assist when maps and models are used in decision-making, for instance in conservation and management.
Fishery–independent surface density and abundance estimates for the swordfish were obtained through aerial surveys carried out over a large portion of the Central Mediterranean, implementing distance sampling methodologies. Both design- and model-based abundance and density showed an uneven occurrence of the species throughout the study area, with clusters of higher density occurring near converging fronts, strong thermoclines and/or underwater features. The surface abundance was estimated for the Pelagos Sanctuary for Mediterranean Marine Mammals in the summer of 2009 (n=1152; 95%CI=669.0–1981.0; %CV=27.64), the Sea of Sardinia, the Pelagos Sanctuary and the Central Tyrrhenian Sea for the summer of 2010 (n=3401; 95%CI=2067.0–5596.0; %CV=25.51), and for the Southern Tyrrhenian Sea during the winter months of 2010–2011 ( n=1228; 95%CI=578–2605; %CV=38.59). The Mediterranean swordfish stock deserves special attention in light of the heavy fishing pressures. Furthermore, the unreliability of fishery–related data has, to date, hampered our ability to effectively inform long-term conservation in the Mediterranean Region. Considering that the European countries have committed to protect the resources and all the marine-related economic and social dynamics upon which they depend, the information presented here constitute useful data towards the international legal requirements under the Marine Strategy Framework Directory, the Common Fisheries Policy, the Habitats and Species Directive and the Directive on Maritime Spatial Planning, among the others.
Environmental sensitivity analysis provides a framework for systematically and objectively determining the potential for significant environmental impacts. The higher the natural or acquired sensitivity of the receiving environment, the less capable it is to cope with human-induced change. Given that sensitivity is context- and spatially-specific, Geographic Information Systems have been applied to develop an operational Webtool to analyse it. The Webtool enables a rapid and replicable spatial examination of environmental sensitivities and potential for land-use conflicts that supports Strategic Environmental Assessment and, ultimately, informed planning and decision-making. The novelty is on the provision of an online geoprocessing Widget that enables creation of context-specific maps. Pilot testing the Webtool in land-use and renewable energy planning through stakeholder engagement has validated its applicability. Stakeholders confirmed that it enables replicating and, in some cases, improving in-house SEA mapping processes while saving time and effort. However, its full reliance on publicly available spatial datasets renders completeness and resolution issues. The Webtool provides a critical starting-point for sectoral planning discussions and for developing plan/programme alternatives that avoid or minimise potentially incompatible or unsustainable zonings, while promoting consistency and transparency in impact assessment.
Mangrove forests grow in intertidal zones in tropical and subtropical regions and have suffered a dramatic decline globally over the past few decades. Remote sensing data, collected at various spatial resolutions, provide an effective way to map the spatial distribution of mangrove forests over time. However, the spectral signatures of mangrove forests are significantly affected by tide levels. Therefore, mangrove forests may not be accurately mapped with remote sensing data collected during a single-tidal event, especially if not acquired at low tide. This research reports how a decision-tree −based procedure was developed to map mangrove forests using multi-tidal Landsat 5 Thematic Mapper (TM) data and a Digital Elevation Model (DEM). Three indices, including the Normalized Difference Moisture Index (NDMI), the Normalized Difference Vegetation Index (NDVI) and NDVIL·NDMIH (the multiplication of NDVIL by NDMIH, L: low tide level, H: high tide level) were used in this algorithm to differentiate mangrove forests from other land-cover and land-use types in Fangchenggang City, China. Additionally, the recent Landsat 8 OLI (Operational Land Imager) data were selected to validate the results and compare if the methodology is reliable. The results demonstrate that short-term multi-tidal remotely-sensed data better represent the unique nearshore coastal wetland habitats of mangrove forests than single-tidal data. Furthermore, multi-tidal remotely-sensed data has led to improved accuracies using two classification approaches: i.e. decision trees and the maximum likelihood classification (MLC). Since mangrove forests are typically found at low elevations, the inclusion of elevation data in the two classification procedures was tested. Given the decision-tree method does not assume strict data distribution parameters, it was able to optimize the application of multi-tidal and elevation data, resulting in higher classification accuracies of mangrove forests. When using multi-source data of differing types and distributions to map mangrove forests, a decision-tree method appears to be superior to traditional statistical classifiers.