Benthic Classes Research Articles

Scalable semantic 3D mapping of coral reefs with deep learning

Abstract Coral reefs are among the most diverse ecosystems on our planet, and essential to the livelihood of hundreds of millions of people who depend on them for food security, income from tourism and coastal protection. Unfortunately, most coral reefs are existentially threatened by global climate change and local anthropogenic pressures. To better understand the dynamics underlying deterioration of reefs, monitoring at high spatial and temporal resolution is key. However, conventional monitoring methods for quantifying coral cover and species abundance are limited in scale due to the extensive manual labor required. Although computer vision tools have been employed to aid in this process, in particular structure‐from‐motion (SfM) photogrammetry for 3D mapping and deep neural networks for image segmentation, analysis of the data products creates a bottleneck, effectively limiting their scalability. This paper presents a new paradigm for mapping underwater environments from ego‐motion video, unifying 3D mapping systems that use machine learning to adapt to challenging conditions under water, combined with a modern approach for semantic segmentation of images. The method is exemplified on coral reefs in the northern Gulf of Aqaba, Red Sea, demonstrating high‐precision 3D semantic mapping at unprecedented scale with significantly reduced required labor costs: given a trained model, a 100 m video transect acquired within 5 min of diving with a cheap consumer‐grade camera can be fully automatically transformed into a semantic point cloud within 5 min. We demonstrate the spatial accuracy of our method and the semantic segmentation performance (of at least 80% total accuracy), and publish a large dataset of ego‐motion videos from the northern Gulf of Aqaba, along with a dataset of video frames annotated for dense semantic segmentation of benthic classes. Our approach significantly scales up coral reef monitoring by taking a leap towards fully automatic analysis of video transects. The method advances coral reef transects by reducing the labor, equipment, logistics, and computing cost. This can help to inform conservation policies more efficiently. The underlying computational method of learning‐based Structure‐from‐Motion has broad implications for fast low‐cost mapping of underwater environments other than coral reefs.

Mapping the distribution of sea urchin (Echinoidea) and benthic habitat using drone in the waters of Lancang Island

Sea urchins are valuable marine creatures that live in the waters of Indonesia, and typically found in seagrass and coral reefs. This study aimed to determine the distribution of sea urchins and the benthic characteristics that serve as their habitat using high resolution drone images. Object-based image analysis (OBIA) with a support vector machine algorithm is used to identify benthic habitats, while the distribution of sea urchins is determined by thresholding pixel values. A total of 812 aerial photos with a spatial resolution of 3.98 cm/px were obtained and mapped into five classes of benthic habitat (sand, rubble, seagrass, living coral, and dead coral algae). The dominant habitat class obtained was dead coral algae, with an area of 234,035 m2 (29.63%), whereas seagrass was the smallest, with an area of 74,149 m2 (9.39%). The overall accuracy of the benthic classes was 65.44%, with kappa coefficient value of 0.57. The study found that the total number of sea urchins on Lancang Island was 61,551. The dominant distribution of sea urchins was found in the living coral class with 18,094 individuals, but the highest density was in the rubble class with 1358 individuals/ha.

Mapping reef terrain roughness and benthic habitat complexity in the Pamegaran Island, Kepulauan Seribu National Marine Park

The formation of coral cays in the Kepulauan Seribu Marine Parks, Jakarta Special Capital Region, is the result of patch reef evolution which linked to the geology of sediment accretion, including the ecology of coral growth and competition of benthic community. This study aims to map the level of reef complexity in Pamegaran Island using Benthic Terrain Modeler. Field observation to monitor benthic communities in four stations at 3- and 5- meter depth was combined with SPOT-7 image processing to evaluate in situ rugosity, 3-dimensional bathymetry and benthic habitat classification. The results of this study revealed low degree of terrain complexity as results of habitat degradation and dominant benthic class of sand. Classification of reef benthic habitat resulted in five classes, namely sand (Sd= 126.65 ha), seagrass (Sg= 35.29 ha), bare rubble (Rb= 32.68 ha), reefs dominated by dead corals (DC=3.16 ha) and dead coral with algae (DA= 20.50 ha); with 67.21% score of overall accuracy (OA). In situ rugosity at 5 m was higher than 3 m, as result of higher coverage of live corals particularly in the east site. Similar profile was shown for rugosity index, which reflects ratio between surface to planar area and seafloor gradient along the reef slope. Finally, positive correlation between reef surface and in situ rugosity implies the importance of live coral communities in supporting complex terrain and benthic habitat in the patch reef environment.

Spatial distribution of benthic habitats in Kapota Atoll (Wakatobi National Park, Indonesia) using remote sensing imagery

Abstract. Hamuna B, Pujiyati S, Gaol JL, Hestirianoto T. 2023. Spatial distribution of benthic habitats in Kapota Atoll (Wakatobi National Park, Indonesia) using remote sensing imagery. Biodiversitas 24: 3700-3707. This study aims to classify and map benthic habitats in Kapota Atoll, Wakatobi National Park, using Sentinel-2A satellite imagery, that has undergone both atmospheric and geometric corrections. Sentinel-2A satellite imagery was processed using pixel-based classification method. This study used two machine learning algorithms: support vector machine (SVM) and random forest (RF). This study used nine classified benthic habitat classes, namely five homogeneous benthic classes (sand, live coral, dead coral, rubble, and dense seagrass) and four mixed benthic classes (mixed rubble and sand, mixed sand and rare seagrass, mixed medium seagrass and sand, and a mixture of rubble, dead coral, and seagrass). The RF algorithm produced benthic habitat maps with a higher overall accuracy than the SVM algorithm, which was 65.31% and 63.51%, respectively. These accuracy values have met the standards of accuracy for mapping benthic habitats that apply in Indonesia. The SVM algorithm classification results showed that the benthic habitat of the mixed class of medium seagrass and sand was more dominant than other benthic habitat classes. In contrast, the RF algorithm showed the class dense seagrass was more dominant. The selection of the appropriate algorithm for processing satellite imagery was found to be a significant factor in producing an accurate benthic habitat map.

Species level mapping of a seagrass bed using an unmanned aerial vehicle and deep learning technique.

BackgroundSeagrass beds are essential habitats in coastal ecosystems, providing valuable ecosystem services, but are threatened by various climate change and human activities. Seagrass monitoring by remote sensing have been conducted over past decades using satellite and aerial images, which have low resolution to analyze changes in the composition of different seagrass species in the meadows. Recently, unmanned aerial vehicles (UAVs) have allowed us to obtain much higher resolution images, which is promising in observing fine-scale changes in seagrass species composition. Furthermore, image processing techniques based on deep learning can be applied to the discrimination of seagrass species that were difficult based only on color variation. In this study, we conducted mapping of a multispecific seagrass bed in Saroma-ko Lagoon, Hokkaido, Japan, and compared the accuracy of the three discrimination methods of seagrass bed areas and species composition, i.e., pixel-based classification, object-based classification, and the application of deep neural network.MethodsWe set five benthic classes, two seagrass species (Zostera marina and Z. japonica), brown and green macroalgae, and no vegetation for creating a benthic cover map. High-resolution images by UAV photography enabled us to produce a map at fine scales (<1 cm resolution).ResultsThe application of a deep neural network successfully classified the two seagrass species. The accuracy of seagrass bed classification was the highest (82%) when the deep neural network was applied.ConclusionOur results highlighted that a combination of UAV mapping and deep learning could help monitor the spatial extent of seagrass beds and classify their species composition at very fine scales.

Machine-Learning for Mapping and Monitoring Shallow Coral Reef Habitats

Mapping and monitoring coral reef benthic composition using remotely sensed imagery provides a large-scale inference of spatial and temporal dynamics. These maps have become essential components in marine science and management, with their utility being dependent upon accuracy, scale, and repeatability. One of the primary factors that affects the utility of a coral reef benthic composition map is the choice of the machine-learning algorithm used to classify the coral reef benthic classes. Current machine-learning algorithms used to map coral reef benthic composition and detect changes over time achieve moderate to high overall accuracies yet have not demonstrated spatio-temporal generalisation. The inability to generalise limits their scalability to only those reefs where in situ reference data samples are present. This limitation is becoming more pronounced given the rapid increase in the availability of high temporal (daily) and high spatial resolution (&lt;5 m) multispectral satellite imagery. Therefore, there is presently a need to identify algorithms capable of spatio-temporal generalisation in order to increase the scalability of coral reef benthic composition mapping and change detection. This review focuses on the most commonly used machine-learning algorithms applied to map coral reef benthic composition and detect benthic changes over time using multispectral satellite imagery. The review then introduces convolutional neural networks that have recently demonstrated an ability to spatially and temporally generalise in relation to coral reef benthic mapping; and recurrent neural networks that have demonstrated spatio-temporal generalisation in the field of land cover change detection. A clear conclusion of this review is that existing convolutional neural network and recurrent neural network frameworks hold the most potential in relation to increasing the spatio-temporal scalability of coral reef benthic composition mapping and change detection due to their ability to spatially and temporally generalise.

Coral Reef Mapping with Remote Sensing and Machine Learning: A Nurture and Nature Analysis in Marine Protected Areas

Mapping habitats is essential to assist strategic decisions regarding the use and protection of coral reefs. Coupled with machine learning (ML) algorithms, remote sensing has allowed detailed mapping of reefs at meaningful scales. Here we integrated WorldView-3 and Landsat-8 imagery and ML techniques to produce a map of suitable habitats for the occurrence of a model species, the hydrocoral Millepora alcicornis, in coral reefs located inside marine protected areas in Northeast Brazil. Conservation and management efforts in the region were also analyzed, integrating human use layers to the ecological seascape. Three ML techniques were applied: two to derive base layers, namely geographically weighted regressions for bathymetry and support vector machine classifier (SVM) for habitat mapping, and one to build the species distribution model (MaxEnt) for Millepora alcicornis, a conspicuous and important reef-building species in the area. Additionally, human use was mapped based on the presence of tourists and fishers. SVM yielded 15 benthic classes (e.g., seagrass, sand, coral), with an overall accuracy of 79%. Bathymetry and its derivative layers depicted the topographical complexity of the area. The Millepora alcicornis distribution model identified distance from the shore and depth as topographical factors limiting the settling and growth of coral colonies. The most important variables were ecological, showing the importance of maintaining high biodiversity in the ecosystem. The comparison of the habitat suitability model with species absence and human use maps indicated the impact of direct human activities as potential inhibitors of coral development. Results reinforce the importance of the establishment of no-take zones and other protective measures for maintaining local biodiversity.

KLASIFIKASI HABITAT PERAIRAN DANGKAL DARI CITRA MULTISPASIAL DI PERAIRAN PULAU KAPOTA DAN PULAU KOMPOONE, KEPULAUAN WAKATOBI

Habitat perairan dangkal sangat penting dipetakan diantaranya karena: (1) mendukung perencanaan, manajemen, dan pengambilan keputusan tata ruang pemerintah; (2) mendukung dan mendesain Marine Protected Area (MPA); (3) melakukan program penelitian ilmiah yang bertujuan untuk menghasilkan pengetahuan tentang ekosistem bentik dan geologi dasar laut; (4) melakukan penilaian sumber daya dasar laut yang hidup dan tidak hidup untuk tujuan ekonomi dan menajemen, termasuk rancangan cadangan perikanan. Hingga saat ini belum ada standar untuk tingkat kedetailan peta tematik ekosistem pesisir khususnya habitat perairan dangkal sesuai kebutuhan pengelolaan wilayah pesisir dengan skema klasifikasi tertentu. Penelitian ini bertujuan untuk membandingkan akurasi peta hasil klasifikasi habitat perairan dangkal antara citra SPOT 6, Sentinel 2A, dan Landsat 8 menggunakan algoritma klasifikasi support vector machine. Lokasi penelitian terletak di Kepulauan Wakatobi, meliputi 2 lokasi yaitu Pulau Kapota dan Pulau Kompoone. Pengambilan data in-situ dilaksanakan pada tanggal 7-11 Juli 2019. Sebanyak 347 ground truth dan foto transek hasil sampling di lapangan telah dianalisis menggunakan coral point count with excel extension (CPCe). Skema klasifikasi yang dihasilkan yaitu 8 kelas habitat bentik, selanjutnya dilakukan klasifikasi dengan mengkelaskan kembali menjadi 6 dan 5 kelas. Hasil yang diperoleh pada citra SPOT-6 untuk semua kelas habitat perairan dangkal yang digunakan memiliki overall accuracy yang lebih besar. Perbedaan ukuran piksel (resolusi spasial) dan jumlah skema klasifikasi sangat memengaruhi hasil akurasi.

Monitoring benthic habitats using Lyzenga model features from Landsat multi-temporal images in Google Earth Engine

Benthic habitats have conventionally been monitored through physical site visits by diving. This requires the use of expensive resources and time to achieve the desired results. This study adopts remotely sensed Landsat multi-temporal images to monitor benthic habitats within marine protected areas in Kenya. We adopt the Lyzenga approach to compute depth invariant index features on Google Earth Engine (GEE) API to aid benthic classification. Random Forest (RF) and Support Vector Machines (SVM) machine learning algorithms were used for the supervised classification of benthic habitats based on training classes selected under the guidance of an initial K-means unsupervised classification. RF had a higher overall accuracy of 83.48% compared to SVM at 52.69% as per 2015 reference data. We, therefore, used RF to map benthic habitats between 2003 and 2018. Findings indicate that there are high conversions between corals, seagrass and sand benthic classes. Since environmental and anthropogenic pressures act synergistically in causing changes, an in-depth research would be necessary to assess the vulnerability of benthic habitats to these factors. However, the use of multi-spectral imagery and powerful GEE cloud computing platform can aid monitoring and acquisition of marine data.

Assessment of Sentinel‐2A multispectral image for benthic habitat composition mapping

Sentinel-2A accuracy for benthic habitat composition mapping was tested and compared to ALOS AVNIR-2. Aerial image acquired using custom-made unmanned aerial vehicle was used to train and validate the model. The mapping was conducted regardless of the benthic class and at individual benthic class. Benthic habitat class spatial distribution was obtained using the combination of image segmentation and classification tree analysis. The aerial image was interpreted based on the percentage of the constructed and non-constructed classes. The constructed class includes coral reefs, dead coral, seagrass, and macroalgae, while non-constructed class covers carbonate sand, rock, and rubble. Sentinel-2A produced higher accuracy (92%) than ALOS AVNIR-2 (78%) for benthic habitat spatial distribution mapping. However, in the empirical modelling of benthic habitat composition, ALOS AVNIR-2 (SE 23-24%) produced slightly better accuracy than Sentinel-2A (SE 23-27%). Several factors affected the low accuracy, which include the sub-pixel mixing of benthic habitat and constructed class, the delay between dates of acquisition, and radiometric quality of the images. Since the fundamental relationship between reflectance value and the percentage of the constructed class has been justified and consistent, given more experiments it has the potential to predict benthic habitat composition with higher accuracy in the future.

Impact of water characteristics on the discrimination of benthic cover in and around coral reefs from imaging spectrometer data

Coral reefs are the foundation of productive ecosystems in the global, tropical oceans and are under threat from a variety of local to global scale stressors. Satellite imagery provides a tool to identify and understand the processes that control coral reef degradation, however due to the dynamic nature of seawater constituents, current spaceborne multispectral sensors cannot reliably discriminate between the many coral reef benthic classes necessary to detect change. Hyperspectral imagers may provide sufficient spectral resolution to estimate water column properties and differentiate benthic classes, however, the effects of depth, seawater constituents, and classification algorithm on the accuracy of benthic classifications have not been systematically assessed. Here, we simulate the ability of a spaceborne hyperspectral imager to accurately map fractional cover of coral reef benthic classes under a variety of conditions. Benthic reflectance is simulated by combining pure reflectance spectra of coral, algae, and sand and projecting these mixed spectra through a fully crossed set of water columns. We then use a semi-analytical optimization procedure to estimate the water column properties and multiple endmember spectral mixture analysis to estimate the fractional cover of the benthic classes using many independent endmember spectra. We compare our estimated benthic class fractions to the original, actual fractions used to produce the mixed coral reef spectra to quantify several measures of error. We found that multiple endmember spectral mixture analysis decreases fractional retrieval error, which is also reduced when the first derivative of the mixed and endmember spectra is used prior to unmixing. The estimation of fractional benthic class cover is most accurate for depths ≤3 m for most water conditions. Depths ≥5 m should be classified only if chlorophyll and sediment concentration are <0.1 mg m−3 and <0.1 g m−3, respectively. Our results indicate that the fractional cover of coral and algae should be at least 25% for accurate benthic class estimates (mean relative error < 50%), however there will be many ways to leverage the repeat measurements of a hyperspectral satellite sensor, such as a stable depth retrievals and benthic cover estimates, to produce more accurate and useful fractional cover data. We show how this simulation analysis can be used to generate maps of predicted benthic cover fractional retrieval uncertainty across a coral reef system using aerial hyperspectral imagery acquired over Hawaii, USA, although reef-specific, within pixel variations in depth and benthic class complexity should be considered.

Optimising Sampling Strategies in Coral Reefs Using Large-Area Mosaics

Adequately sampling benthic cover in marine ecosystems is a challenge with most methods encompassing only a small portion of the area for which cover is estimated. Recent advances in photogrammetric techniques are providing opportunity to map expansive areas of reef. This study aimed to evaluate the adequate level of sampling for traditional quadrat-based designs at ecologically relevant scales. We used large-area (~250 m2), high-resolution (0.911 ± 0.143 mm/pixel) mosaics to assess the precision and reproducibility of quadrat-based benthic sampling and identify the most efficient strategy (size and number of quadrats). There was a strong relationship between the percent cover of benthic classes and the level of sampling effort required to adequately sample them. As expected, larger quadrats were found to be more efficient when sampling effort was expressed in number of quadrats. This study aims to identify the optimal level of sampling (least effort that would result in a given target precision) to characterize coral reef benthic communities (whatever they are) within each site. As such, the sites selected were intentionally very different and together represented the broad scale of heterogeneity found in shallow coral reef communities. Abundance data can be used in combination with the relationships presented here to determine the optimal sampling protocols for management approaches to coral reef monitoring.

Kelp-bed dynamics across scales: Enhancing mapping capability with remote sensing and GIS

Kelp are important drivers of productivity and biodiversity patterns in cold-water and nutrient-rich rocky reefs. Scuba- and boat-based methods are routinely used to study submerged kelp beds. However, these time-consuming and labor-intensive methods enable monitoring of beds or the factors and processes that control their distribution over only small spatial (few 100s of m2) and temporal (<5 years) scales. Remote sensing and geographic information system (GIS) technologies are increasingly used to compare marine species distribution over multiple spatiotemporal scales. However, there is currently no clear framework and limited demonstration of their potential for studies of broad-scale changes in completely submerged kelp beds. The present study aims to establish the foundation of a simple, accessible, and robust set of remote sensing and GIS-based methods to address key questions about the stability of subtidal kelp beds across multiple spatial and temporal scales. It tests the suitability of conventional image classification methods for mapping kelp from digital aerial (acquired on board a helicopter) and satellite (SPOT 7) imagery of ~250 ha of seabed around four islands in the Mingan Archipelago (northern Gulf of St. Lawrence, Canada). Three classification methods are compared: 1) a software-led unsupervised classification in which pixels are grouped into clusters based on similarity in spectral signature among pixels; 2) a software-led supervised classification in which pixels are assigned to categories based on similarity in the spectral signature of pixels and that of reference data from each category; and 3) a visual classification carried out by a trained observer. Supervised classification of satellite imagery and visual classification of aerial imagery were the top methods to map kelp, with overall accuracies of 89% and 90%, respectively. Unsupervised classification of both types of imagery showed poor discrimination between kelp and non-kelp benthic classes. Kelp bed edges were more difficult to identify on satellite than aerial imagery because the former presented poorer contrasts and a lower spatial resolution. Kelp bed edges identified with visual classification appeared artificially jagged for both types of imagery, mainly because of the coarse (225-m2) spatial units used for this classification. Kelp bed edges were smoother on maps created with the unsupervised and supervised classifications, which used 1-m-pixel images. The present study demonstrates that conventional remote sensing and GIS methods can accurately map submerged kelp beds over large spatial domains in the Mingan Archipelago or in other benthic systems with similar oceanic conditions and a largely dichotomous (kelp-barrens) biological makeup.

Coral habitat mapping: a comparison between maximum likelihood, Bayesian and Dempster–Shafer classifiers

This study deals with the mixed-pixel problem of detecting benthic habitat class membership and evaluates two soft classifiers for coral habitat mapping on Lang Tengah island (Malaysia). A comparison was made between the Bayesian and Dempster–Shafer (D–S) with a traditional maximum likelihood (ML). The heterogeneous pattern of reef environment, established by field observation, four classes of coral habitats containing various combinations of live coral, dead coral with algae, rubble coral and sand. Posterior probability and belief maps, generated by Bayesian and D–S, respectively, were evaluated by visual inspection and final coral habitat distribution maps were validated via accuracy assessment estimates. The accuracy validation tests agreed with the visual inspection of the probability, uncertainty and coral distribution maps. The Bayesian algorithm performed better, with a 34.7–68.5% improvement in accuracy compared to D–S and ML, respectively. Probability maps demonstrate the advantages of the soft classifier over the hard classifier for coral mapping.

Benthic Habitat Mapping Using Multispectral High-Resolution Imagery: Evaluation of Shallow Water Atmospheric Correction Techniques.

Remote multispectral data can provide valuable information for monitoring coastal water ecosystems. Specifically, high-resolution satellite-based imaging systems, as WorldView-2 (WV-2), can generate information at spatial scales needed to implement conservation actions for protected littoral zones. However, coastal water-leaving radiance arriving at the space-based sensor is often small as compared to reflected radiance. In this work, complex approaches, which usually use an accurate radiative transfer code to correct the atmospheric effects, such as FLAASH, ATCOR and 6S, have been implemented for high-resolution imagery. They have been assessed in real scenarios using field spectroradiometer data. In this context, the three approaches have achieved excellent results and a slightly superior performance of 6S model-based algorithm has been observed. Finally, for the mapping of benthic habitats in shallow-waters marine protected environments, a relevant application of the proposed atmospheric correction combined with an automatic deglinting procedure is presented. This approach is based on the integration of a linear mixing model of benthic classes within the radiative transfer model of the water. The complete methodology has been applied to selected ecosystems in the Canary Islands (Spain) but the obtained results allow the robust mapping of the spatial distribution and density of seagrass in coastal waters and the analysis of multitemporal variations related to the human activity and climate change in littoral zones.

Mapping shallow nearshore benthic features in a Caribbean marine-protected area: assessing the efficacy of using different data types (hydroacoustic versus satellite images) and classification techniques

ABSTRACTVarious benthic mapping methods exist but financing and technical capacity limit the choice of technology available to developing states to aid with natural resource management. Therefore, we assessed the efficacy of using a single-beam echosounder (SBES), satellite images (GeoEye-1 and WorldView-2) and different image (pixel-based Maximum Likelihood Classifier (MLC), and an object-based image analysis (OBIA)) and hydroacoustic classification and interpolation techniques, to map nearshore benthic features at the Bluefields Bay marine protected area in western Jamaica (13.82 km2 in size). A map with three benthic classes (submerged aquatic vegetation (SAV), bare substrate, and coral reef) produced from a radiometrically corrected, deglinted and water column-corrected WorldView-2 image had a marginally higher accuracy (3%) than that of a map classified from a similarly corrected GeoEye-1 image. However, only one of the two extra WorldView-2 image bands (coastal) was used because the yellow band was completely attenuated at depths ≥3.7 m. The coral reef class was completely misclassified by the MLC and had to be contextually edited. The contextually edited MLC map had a higher overall accuracy (OA) than the OBIA map (86.7% versus 80.4%) and maps that were not contextually edited. But, the OBIA map had a higher OA than a MLC map without edits. Maps produced from the images also had a higher accuracy than the SAV map created from the acoustic data (OAs >80% and kappa >0.67 versus 76.6% and kappa = 0.32). SAV classification was comparable among the classified SBES SAV data points and all the final maps. The total area classified as SAV was marginally larger for satellite maps; however, the total area classified as bare substrate using the images was twice as large. A substrate map with three classes (silt, sand, and coral/hard bottom) produced from the SBES data using a random forest classifier and a Markov chain interpolator had a higher accuracy than a substrate map produced using a fractal dimension classifier and an indicator krig (the default choice) (72.4% versus 53.5%). The coral reef class from the SBES, OBIA, and contextually edited maps had comparable accuracies, but covered a much smaller area in the SBES maps because data points were lost during the interpolation process. The use of images was limited by turbidity levels and cloud cover and it yielded lower benthic detail. Despite these limitations, satellite image classification was the most efficacious method. If greater benthic detail is required, the SBES is more suitable or more effort is required during image classification. Also, the SBES can be operated in areas with turbid waters and greater depths. However, it could not be used in very shallow areas. Also, processing and interpolation of data points can result in a loss of resolution and introduces spatial uncertainty.

Hyperspectral remote sensing of shallow waters: Considering environmental noise and bottom intra-class variability for modeling and inversion of water reflectance

Hyperspectral remote sensing is now an established tool to determine shallow water properties over large areas, usually by inverting a semi-analytical model of water reflectance. However, various sources of error may make the observed subsurface remote-sensing reflectance deviate from the model, resulting in an increased retrieval error when inverting the model based on classical least-squares fitting. In this paper, we propose a probabilistic forward model of shallow water reflectance variability that describes two of the main sources of error, namely, (1) the environmental noise that includes every source of above-water variability (e.g., sensor noise and rough water surface), and (2) the potentially complex inherent spectral variability of each benthic class through their associated spectral covariance matrix. Based on this probabilistic model, we derive two inversion approaches, namely, MILE (MaxImum Likelihood estimation including Environmental noise) and MILEBI (MaxImum Likelihood estimation including Environmental noise and Bottom Intra-class variability) that utilize the information contained in the proposed covariance matrices to further constrain the inversion while allowing the observation to differ from the model in the less reliable wavebands. In this paper, MILE and MILEBI are compared with the widely used least-squares (LS) criterion in terms of depth, water clarity and benthic cover retrievals. For these three approaches, we also assess the influence of constraining bottom mixture coefficients to sum to one on estimation results.The results show that the proposed probabilistic model is a valuable tool to investigate the influence of bottom intra-class variability on subsurface reflectance, e.g., as a function of optical depth or environmental noise. As expected, this influence is critical in very optically shallow waters, and decreases with increasing optical depth. The inversion results obtained from synthetic and airborne data of Quiberon Peninsula, France, show that MILE and MILEBI generally provide better performances than LS. For example, in the case of airborne data with depth ranging from 0.44 to 12.00m, the bathymetry estimation error decreases by about 32% when using MILE and MILEBI instead of LS. Estimated maps of bottom cover are also more consistent when derived using sum-to-one constrained versions of MILE and MILEBI. MILE is shown to be a simple but powerful method to map simple benthic habitats with negligible influence of intra-class variability. Alternatively, MILEBI is to be preferred if this variability cannot be neglected, since taking bottom covariance matrices into account concurrently with mean reflectance spectra may help the bottom discrimination, e.g., in the presence of overlapping classes. This study thus shows that taking potential sources of error into account through appropriate parameterizations of spectral covariance may be critical to improve the remote sensing of shallow waters, hence making MILE and MILEBI interesting alternatives to LS.

The importance of offshore origination revealed through ophiuroid phylogenomics.

Our knowledge of macro-evolutionary processes in the deep sea is poor, leading to much speculation about whether the deep sea is a source or sink of evolutionary adaptation. Here, we use a phylogenetic approach, on large molecular (688 species, 275 kbp) and distributional datasets (104 513 records) across an entire class of marine invertebrates (Ophiuroidea), to infer rates of bathymetric range shift over time between shallow and deep water biomes. Biome conservation is evident through the phylogeny, with the majority of species in most clades distributed within the same bathome. Despite this, bathymetric shifts have occurred. We inferred from ancestral reconstructions that eurybathic or intermediate distributions across both biomes were a transitional state and direct changes between shallow and deep sea did not occur. The macro-evolutionary pattern of bathome shift appeared to reflect micro-evolutionary processes of bathymetric speciation. Results suggest that most of the oldest clades have a deep-sea origin, but multiple colonization events indicate that the evolution of this group conforms neither to a simple onshore-offshore hypothesis, nor the opposite pattern. Both shallow and deep bathomes have played an important role in generating the current diversity of this major benthic class.

Developing Benthic Class Specific, Chlorophyll-a Retrieving Algorithms for Optically-Shallow Water Using SeaWiFS.

This study evaluated the ability to improve Sea-Viewing Wide Field-of-View Sensor (SeaWiFS) chl-a retrieval from optically shallow coastal waters by applying algorithms specific to the pixels’ benthic class. The form of the Ocean Color (OC) algorithm was assumed for this study. The operational atmospheric correction producing Level 2 SeaWiFS data was retained since the focus of this study was on establishing the benefit from the alternative specification of the bio-optical algorithm. Benthic class was determined through satellite image-based classification methods. Accuracy of the chl-a algorithms evaluated was determined through comparison with coincident in situ measurements of chl-a. The regionally-tuned models that were allowed to vary by benthic class produced more accurate estimates of chl-a than the single, unified regionally-tuned model. Mean absolute percent difference was approximately 70% for the regionally-tuned, benthic class-specific algorithms. Evaluation of the residuals indicated the potential for further improvement to chl-a estimation through finer characterization of benthic environments. Atmospheric correction procedures specialized to coastal environments were recognized as areas for future improvement as these procedures would improve both classification and algorithm tuning.

Corals & benthic habitat mapping using DubaiSat-2: a spectral-spatial approach applied to Dalma Island, UAE (Arabian Gulf)

ABSTRACTSatellite remote sensing-based monitoring of coastal habitats, like those in the Arabian Gulf, presents a special challenge due to the attenuation of light through the turbid atmosphere and water, as well as the spectral similarity of many benthic habitats. The present study aims to evaluate the potential of DubaiSat-2 imagery in mapping corals and benthic habitat in the vicinity of Dalma Island, United Arab Emirates (UAE). To do so, this study proposes a spectral-spatial method that uses a combination of different remote sensing techniques in order to discern benthic communities with reference to their spatial distribution. This method involves multi-level segmentation, non-linear feature analysis and ensemble learning methods for the characterization of corals, benthic classes and underlying seabeds. Maximum Likelihood and Random Forest, pixel-based methods, were also used for comparison of classification performances. Comparative data are derived from published marine habitat maps. The spectral-spatial method achieves 96.60% overall accuracy. Maximum Likelihood and Random Forest classification methods are assessed to be 92.59% and 94.52% accurate, respectively. The application of the proposed method and the use of DubaiSat-2 imagery can help in achieving well-designed coastal management plans in the region.