Hyperspectral Shallow-Water Remote Sensing with an Enhanced Benthic Classifier

Semi-analytical model for deep-water hyperspectral imaging

We propose a semi-analytical (SA) model for relating seafloor reflectance to measured radiance in deep-water hyperspectral imaging in artificially illuminated scenes. Using accurate sensor-seafloor geometry from photogrammetry and the principle of two-viewpoint observation, we estimate the inherent optical properties (IOPs) of the water column. We demonstrate the SA model and estimation of IOPs for hyperspectral imaging of a deep-water coral reef from a remotely operated vehicle. For the calibrated SA model, evaluation of across-viewpoint similarity demonstrates the model's ability to compensate for water column and light source effects.

Modelling hydrological strength and alteration in moribund deltaic India

The Ganga-Brahmaputra moribund deltaic floodplain region hosted many socio-ecologically precious freshwater wetland ecosystems experiencing hydrological alteration. The present study aimed to model hydrological strength (HS) to show the spatial difference and account for the degree and direction of hydrological alteration of Indian moribund deltaic wetland in three phases e.g. (1) phase I (1988-1997), (2) phase II (1998-2007) and phase III (2008-2017). Three key hydrological parameters, such as Water Presence Frequency (WPF), water depth, and hydro-period were considered for hydrological strength modelling using two ensemble Machine Learning (ML) techniques (Random Forest (RF) and XGBoost). Image algebra was employed for phasal change detection. Hydrological strength models show that around 75% of the wetland area was lost in-between phases I to III and the loss was found more intensive in moderate and weak HS zones. Existing wetland shows a clear spatial difference of HS between wetland core and periphery and river linked and delinked or not linked wetlands. Regarding the suitability of the ML models, both are acceptable, however, the XGBoost outperformed in reference to applied 15 statistical validation techniques and field evidence. HS models based on change detection clarified that more than 22% and 55% of the weak HS zone in phases II and III respectively were turned into non-wetland. The degree of alteration revealed that about 40% of wetland areas experienced a negative alteration during phases I to II, and this proportion increased to 63% in between phases II to III. Since the study figured out the spatial nature of HS, degree and direction of alteration at a spatial scale, these findings would be instrumental for adopting rational planning towards wetland conservation and restoration.

The Development of A Rigorous Model for Bathymetric Mapping from Multispectral Satellite-Images

Models for bathymetry retrieval from multispectral images have not considered the errors caused by tidal fluctuation. A rigorous bathymetric model that considers the variation in tide height time series, including the tide height calculation and instantaneous tide height correction at the epoch of satellite flight into the bathymetric retrieval model, is proposed in this paper. The model was applied on Weizhou Island, located in Guangxi Province, China, and its accuracy verificated with four check lines and seven checkpoints. A scene from the Landsat 8 satellite image was used as experimental data. The reference (“true”) water depth data collected by a RESON SeaBat 7125 multibeam instrument was used for comparison analysis. When satellite-derived bathymetry is compared, it is found that maximum absolute error, mean absolute error, and RMSE have decreased 54, 45, and 30% relative to that of the traditional model in the entire test field. The accuracy of the water depths retrieved by our model increased 30 and 56% when validated using four check lines and seven checkpoints, respectively. Therefore, it can be concluded that the model proposed in this paper can effectively improve the accuracy of bathymetry retrieved from Landsat 8 images.

Fusion of Drone-Based RGB and Multi-Spectral Imagery for Shallow Water Bathymetry Inversion

Shallow bathymetry inversion algorithms have long been applied in various types of remote sensing imagery with relative success. However, this approach requires that imagery with increased radiometric resolution in the visible spectrum be available. The recent developments in drones and camera sensors allow for testing current inversion techniques on new types of datasets with centimeter resolution. This study explores the bathymetric mapping capabilities of fused RGB and multispectral imagery as an alternative to costly hyperspectral sensors for drones. Combining drone-based RGB and multispectral imagery into a single cube dataset provides the necessary radiometric detail for shallow bathymetry inversion applications. This technique is based on commercial and open-source software and does not require the input of reference depth measurements in contrast to other approaches. The robustness of this method was tested on three different coastal sites with contrasting seafloor types with a maximum depth of six meters. The use of suitable end-member spectra, which are representative of the seafloor types of the study area, are important parameters in model tuning. The results of this study are promising, showing good correlation (R2 > 0.75 and Lin’s coefficient > 0.80) and less than half a meter average error when they are compared with sonar depth measurements. Consequently, the integration of imagery from various drone-based sensors (visible range) assists in producing detailed bathymetry maps for small-scale shallow areas based on optical modelling.

Underwater Hyperspectral Imaging (UHI): A Review of Systems and Applications for Proximal Seafloor Ecosystem Studies

Marine ecosystem monitoring requires observations of its attributes at different spatial and temporal scales that traditional sampling methods (e.g., RGB imaging, sediment cores) struggle to efficiently provide. Proximal optical sensing methods can fill this observational gap by providing observations of, and tracking changes in, the functional features of marine ecosystems non-invasively. Underwater hyperspectral imaging (UHI) employed in proximity to the seafloor has shown a further potential to monitor pigmentation in benthic and sympagic phototrophic organisms at small spatial scales (mm–cm) and for the identification of minerals and taxa through their finely resolved spectral signatures. Despite the increasing number of studies applying UHI, a review of its applications, capabilities, and challenges for seafloor ecosystem research is overdue. In this review, we first detail how the limited band availability inherent to standard underwater cameras has led to a data analysis “bottleneck” in seafloor ecosystem research, in part due to the widespread implementation of underwater imaging platforms (e.g., remotely operated vehicles, time-lapse stations, towed cameras) that can acquire large image datasets. We discuss how hyperspectral technology brings unique opportunities to address the known limitations of RGB cameras for surveying marine environments. The review concludes by comparing how different studies harness the capacities of hyperspectral imaging, the types of methods required to validate observations, and the current challenges for accurate and replicable UHI research.

Bringing Bathymetry LiDAR to Coastal Zone Assessment: A Case Study in the Southern Baltic

One of the major tasks in environmental protection is monitoring the coast for negative impacts due to climate change and anthropopressure. Remote sensing techniques are often used in studies of impact assessment. Topographic and bathymetric procedures are treated as separate measurement methods, while methods that combine coastal zone analysis with underwater impacts are rarely used in geotechnical analyses. This study presents an assessment of the bathymetry airborne system used for coastal monitoring, taking into account environmental conditions and providing a comparison with other monitoring methods. The tests were carried out on a section of the Baltic Sea where, despite successful monitoring, coastal degradation continues. This technology is able to determine the threat of coastal cliff erosion (based on the geotechnical analyses). Shallow depths have been reported to be a challenge for bathymetric Light Detection and Ranging (LiDAR), due to the difficulty in separating surface, water column and bottom reflections from each other. This challenge was overcome by describing the classification method used which was the CANUPO classification method as the most suitable for the point cloud processing. This study presents an innovative approach to identifying natural hazards, by combining analyses of coastal features with underwater factors. The main goal of this manuscript is to assess the suitability of using bathymetry scanning in the Baltic Sea to determine the factors causing coastal erosion. Furthermore, a geotechnical analysis was conducted, taking into account geometrical ground change underwater. This is the first study which uses a coastal monitoring approach, combining geotechnical computations with remote sensing data. This interdisciplinary scientific research can increase the awareness of the environmental processes.

Evaluation of forward reflectance models and empirical algorithms for chlorophyll concentration of stratified waters

For waters with stratified chlorophyll concentration (Chl), numerical simulations were carried out to gain insight into the forward models of subsurface reflectance and empirical algorithms for Chl from the ocean color. It is found that the Gordon and Clark (1980) forward model for reflectance using an equivalent homogeneous water with a weighted average Chl (⟨Chl⟩) as the input works well, but depending on the contribution of gelbstoff, the difference in reflectance between stratified and the equivalent homogeneous water can be more than 10%. Further, the attenuation of upward light is better approximated as ∼1.5times that of the diffuse attenuation coefficient of downwelling irradiance. On the other hand, although the forward model for reflectance developed in Zaneveld et al. [Opt. Express13, 9052 (2005)] using equivalent homogeneous water with a weighted average of the backscattering to absorption ratio as the input also works well, this model cannot be used to obtain equivalent ⟨Chl⟩ for reflectance. Further, for empirical Chl algorithms designed for "Case 1" waters, it has been discovered that, for surface Chl in a range of ∼0.06-22.0mg/m³, the predictability of surface Chl is basically the same as that of ⟨Chl⟩ from the blue-green band ratio or the band difference of reflectance. Because ⟨Chl⟩ is wavelength and weighting-formula dependent, and it is required to have profiles of both Chl and the optical properties, these results emphasize that for empirical Chl algorithms, it is easier, less ambiguous, and certainly more straightforward and simple to use surface Chl for algorithm development and then its evaluation, rather than to use ⟨Chl⟩, regardless of whether or not the water is stratified.

Confidence Levels, Sensitivity, and the Role of Bathymetry in Coral Reef Remote Sensing

Remote sensing is playing an increasingly important role in the monitoring and management of coastal regions, coral reefs, inland lakes, waterways, and other shallow aquatic environments. Ongoing advances in algorithm development, sensor technology, computing capabilities, and data availability are continuing to improve our ability to accurately derive information on water properties, water depth, benthic habitat composition, and ecosystem health. However, given the physical complexity and inherent variability of the aquatic environment, most of the remote sensing models used to address these challenges require localized input parameters to be effective and are thereby limited in geographic scope. Additionally, since the parameters in these models are interconnected, particularly with respect to bathymetry, errors in deriving one parameter can significantly impact the accuracy of other derived parameters and products. This study utilizes hyperspectral data acquired in Hawaii in 2000–2001 and 2017–2018 using NASA’s Classic Airborne Visible/Infrared Imaging Spectrometer to evaluate performance and sensitivity of a well-established semi-analytical inversion model used in the assessment of coral reefs. Analysis is performed at several modeled spatial resolutions to emulate characteristics of different feasible moderate resolution hyperspectral satellites, and data processing is approached with the objective of developing a generalized, scalable, automated workflow. Accuracy of derived water depth is evaluated using bathymetric lidar data, which serves to both validate model performance and underscore the importance of image quality on achieving optimal model output. Data are then used to perform a sensitivity analysis and develop confidence levels for model validity and accuracy. Analysis indicates that derived benthic reflectance is most sensitive to errors in bathymetry at shallower depths, yet remains significant at all depths. The confidence levels provide a first-order method for internal quality assessment to determine the physical extent of where and to what degree model output is considered valid. Consistent results were found across different study sites and different spatial resolutions, confirming the suitability of the model for deriving water depth in complex coral reef environments, and expanding our ability to achieve automated widespread mapping and monitoring of global coral reefs.

Water Column Optical Properties of Pacific Coral Reefs Across Geomorphic Zones and in Comparison to Offshore Waters

Despite the traditional view of coral reefs occurring in oligotrophic tropical conditions, water optical properties over coral reefs differ substantially from nearby clear oceanic waters. Through an extensive set of optical measurements across the tropical Pacific, our results suggest that coral reefs themselves exert a high degree of influence over water column optics, primarily through release of colored dissolved organic matter (CDOM). The relative contributions of phytoplankton, non-algal particles, and CDOM were estimated from measurements of absorption and scattering across different geomorphic shallow-water reef zones (<10 m) in Hawaii, the Great Barrier Reef, Guam, and Palau (n = 172). Absorption was dominated at the majority of stations by CDOM, with mixtures of phytoplankton and CDOM more prevalent at the protected back reef and lagoon zones. Absorption could be dominated by sediments and phytoplankton at fringing reefs and terrestrially impacted sites where particulate backscattering was significantly higher than in the other zones. Scattering at three angles in the backward direction followed recent measurements of the particulate phase function. Optical properties derived from satellite imagery indicate that offshore waters are consistently lower in absorption and backscattering than reef waters. Therefore, the use of satellite-derived offshore parameters in modeling reef optics could lead to significant underestimation of absorption and scattering, and overestimation of benthic light availability. If local measurements are not available, average optical properties based on the general reef zone could provide a more accurate means of assessing light conditions on coral reefs than using offshore water as a proxy.

Pushing the Limits of Seagrass Remote Sensing in the Turbid Waters of Elkhorn Slough, California

Remote sensing imagery has been successfully used to map seagrass in clear waters, but here we evaluate the advantages and limitations of different remote sensing techniques to detect eelgrass in the tidal embayment of Elkhorn Slough, CA. Pseudo true-color imagery from Google Earth and broadband satellite imagery from Sentinel-2 allowed for detection of the various beds, but retrievals particularly in the deeper Vierra bed proved unreliable over time due to variable image quality and environmental conditions. Calibrated water-leaving reflectance spectrum from airborne hyperspectral imagery at 1-m resolution from the Portable Remote Imaging SpectroMeter (PRISM) revealed the extent of both shallow and deep eelgrass beds using the HOPE semi-analytical inversion model. The model was able to reveal subtle differences in spectral shape, even when remote sensing reflectance over the Vierra bed was not visibly distinguishable. Empirical methods exploiting the red edge of reflectance to differentiate submerged vegetation only retrieved the extent of shallow alongshore beds. The HOPE model also accurately retrieved the water column absorption properties, chlorophyll-a, and bathymetry but underestimated the particulate backscattering and suspended matter when benthic reflectance was represented as a horizontal eelgrass leaf. More accurate water column backscattering could be achieved by the use of a darker bottom spectrum representing an eelgrass canopy. These results illustrate how high quality atmospherically-corrected hyperspectral imagery can be used to map eelgrass beds, even in regions prone to sediment resuspension, and to quantify bathymetry and water quality.

Shallow-Water Habitat Mapping using Underwater Hyperspectral Imaging from an Unmanned Surface Vehicle: A Pilot Study

The impacts of human activity on coastal ecosystems are becoming increasingly evident across the world. Consequently, there is a growing need to map, monitor, and manage these regions in a sustainable manner. In this pilot study, we present what we believe to be a novel mapping technique for shallow-water seafloor habitats: Underwater hyperspectral imaging (UHI) from an unmanned surface vehicle (USV). A USV-based UHI survey was carried out in a sheltered bay close to Trondheim, Norway. In the survey, an area of 176 m2 was covered, and the depth of the surveyed area was approximately 1.5 m. UHI data were initially recorded at a 1-nm spectral resolution within the range of 380–800 nm, but this was reduced to 86 spectral bands between 400-700 nm (3.5-nm spectral resolution) during post-processing. The hyperspectral image acquisition was synchronized with navigation data from the USV, which permitted georeferencing and mosaicking of the imagery at a 0.5-cm spatial resolution. Six spectral classes, including coralline algae, the wrack Fucus serratus, green algal films, and invertebrates, were identified in the georeferenced imagery, and chosen as targets for support vector machine (SVM) classification. Based on confusion matrix analyses, the overall classification accuracy was estimated to be 89%–91%, which suggests that USV-based UHI may serve as a useful tool for high-resolution mapping of shallow-water habitats in the future.

A Low-Rate Video Approach to Hyperspectral Imaging of Dynamic Scenes

The increased sensitivity of modern hyperspectral line-scanning systems has led to the development of imaging systems that can acquire each line of hyperspectral pixels at very high data rates (in the 200–400 Hz range). These data acquisition rates present an opportunity to acquire full hyperspectral scenes at rapid rates, enabling the use of traditional push-broom imaging systems as low-rate video hyperspectral imaging systems. This paper provides an overview of the design of an integrated system that produces low-rate video hyperspectral image sequences by merging a hyperspectral line scanner, operating in the visible and near infra-red, with a high-speed pan-tilt system and an integrated IMU-GPS that provides system pointing. The integrated unit is operated from atop a telescopic mast, which also allows imaging of the same surface area or objects from multiple view zenith directions, useful for bi-directional reflectance data acquisition and analysis. The telescopic mast platform also enables stereo hyperspectral image acquisition, and therefore, the ability to construct a digital elevation model of the surface. Imaging near the shoreline in a coastal setting, we provide an example of hyperspectral imagery time series acquired during a field experiment in July 2017 with our integrated system, which produced hyperspectral image sequences with 371 spectral bands, spatial dimensions of 1600 × 212, and 16 bits per pixel, every 0.67 s. A second example times series acquired during a rooftop experiment conducted on the Rochester Institute of Technology campus in August 2017 illustrates a second application, moving vehicle imaging, with 371 spectral bands, 16 bit dynamic range, and 1600 × 300 spatial dimensions every second.

Influence of Three-Dimensional Coral Structures on Hyperspectral Benthic Reflectance and Water-Leaving Reflectance

Shading and inter-reflections created by the three-dimensional coral canopy structure play an important role on benthic reflectance and its propagation above the water. Here, a plane parallel model was coupled with a three-dimensional radiative transfer canopy model, incorporating measured coral shapes and hyperspectral benthic reflectances, to investigate this question under different illumination and water column conditions. Results indicated that a Lambertian treatment of the bottom reflectance can be a reasonable assumption if a variable shading factor is included. Without flexibility in the shading treatment, nadir view bottom reflectances can vary by as much as ±20% (or ±9% in above-water remote sensing reflectance) under solar zenith angles (SZAs) up to 50°. Spectrally-independent shading factors are developed for benthic coral reflectance measurements based on the rugosity of the coral. In remote sensing applications, where the rugosity is unknown, a shading factor could be incorporated as an endmember for retrieval in the inversion scheme. In dense coral canopies in clear shallow waters, the benthos cannot always be treated as Lambertian, and for large solar-view angles the bi-directional reflectance distribution functions (BRDF) hotspot propagated to above water reflectances can create up to a 50% or more difference in water-leaving reflectances, and discrepancies of 20% even for nadir-view geometries.

Seabed Mapping in Coastal Shallow Waters Using High Resolution Multispectral and Hyperspectral Imagery

Coastal ecosystems experience multiple anthropogenic and climate change pressures. To monitor the variability of the benthic habitats in shallow waters, the implementation of effective strategies is required to support coastal planning. In this context, high-resolution remote sensing data can be of fundamental importance to generate precise seabed maps in coastal shallow water areas. In this work, satellite and airborne multispectral and hyperspectral imagery were used to map benthic habitats in a complex ecosystem. In it, submerged green aquatic vegetation meadows have low density, are located at depths up to 20 m, and the sea surface is regularly affected by persistent local winds. A robust mapping methodology has been identified after a comprehensive analysis of different corrections, feature extraction, and classification approaches. In particular, atmospheric, sunglint, and water column corrections were tested. In addition, to increase the mapping accuracy, we assessed the use of derived information from rotation transforms, texture parameters, and abundance maps produced by linear unmixing algorithms. Finally, maximum likelihood (ML), spectral angle mapper (SAM), and support vector machine (SVM) classification algorithms were considered at the pixel and object levels. In summary, a complete processing methodology was implemented, and results demonstrate the better performance of SVM but the higher robustness of ML to the nature of information and the number of bands considered. Hyperspectral data increases the overall accuracy with respect to the multispectral bands (4.7% for ML and 9.5% for SVM) but the inclusion of additional features, in general, did not significantly improve the seabed map quality.