Home > Ningaloo Collaboration Cluster: Hyperspectral mapping of bathymetry and benthic cover

Ningaloo Collaboration Cluster: Hyperspectral mapping of bathymetry and benthic cover

Posted on 05 June 2012

TitleNingaloo Collaboration Cluster: Hyperspectral mapping of bathymetry and benthic cover
Publication TypeReport
Year of Publication2011
AuthorsKlonowski, W, Lynch M, Fearns P, Majewski L, Gray M
Date Published2011
InstitutionCSIRO’s Flagship Collaboration Fund
ISBN Number978-0-86905-959-3
Keywordsbathymetry, Cluster, hyperspectral
AbstractA detailed bathymetry and a benthic cover classification map of the Ningaloo Reef was derived from airborne remotely sensed imagery on a scale of 3.5 m , one of the most extensive, fine scale mapping programs ever attempted. At this fine scale, benthic cover may easily be verified by divers. In situ measurement can easily be used to validate the bathymetry and cover predicted from remote sensing. This study included a verification program conducted by a research team from Murdoch University at a number of selected locations within the Reef. Additionally, Curtin scientists undertook an independent validation of bathymetry using a small craft equipped with an echo sounder. These verification studies found that the bathymetry product was accurate to 20 cm within the 0-20 m depth range and will be of great benefit to the community of users. Accurate bathymetry of the shallow water lagoon is critical and provides the foundation needed for assimilation into numerical hydrodynamic models of coastal currents and near-shore circulation. These calculations are of great importance when predicting coastal sediment transport and in environmental management studies where the impact of engineering structures [pipelines, jetties, marinas, etc] on coastal systems are to be investigated. Bathymetry is a prime variable in such studies where it needs to be spatially well sampled [e.g. a regular grid] and the data of known accuracy. Accurate bathymetry also provides a baseline data set for future management. Any physical forcing on coastal systems [e.g. severe storms, tropical cyclones, storm surge, Tsunami] has the potential to significantly modify the bathymetry. A reference dataset enables any future impacts to be assessed with good accuracy. Such information on system vulnerability clearly feeds into Reef management practice. Ningaloo Reef is impacted by local currents, primarily the Leeuwin Current which appears to entrain Reef waters on its poleward flow down the mid-West Coast. However, the inshore Ningaloo Current and the Holloway Currents also appear quite relevant in that they can transport biological materials such as larvae, plankton, hazardous algal blooms (HABs) and numerous other biological species originating from remote locations into Ningaloo waters. These, when combined with potential global change, for example, water temperature rise, changed wind regimes, changed wave energy forcing, have the potential to impact significantly over time the indigenous marine flora and fauna. A validated baseline dataset comprising a detailed knowledge of the range of benthic species on the Reef along with their spatial distribution at pixel scale of 3.5 m will enable the monitoring and, ideally, the detection, of any such impacts on Ningaloo Reef. Curtin captured underwater video imagery of the benthos from a small craft. A preliminary description of the validation of the remotely sensed benthic cover classification has been reported [see Section 1.5.2 of this Report]. The Curtin Team produced two additional products. The first, requested by the Murdoch University Team, was the spectral reflectance of the reef bottom. This was requested by the Murdoch University researchers because this is a variable that they measure in situ with an underwater spectrometer. The other product was the in-water diffuse attenuation coefficient at 490 nm [K490] which indicated the reduction in light intensity per metre as light travels down to the seabed. This is a difficult variable to estimate in shallow water regions and is not successfully measured by on-orbit satellite sensors. It is a “noisy” product when measured in the current airborne program because the overall attenuation at 490 nm in Ningaloo shallow water is very small. However, some spatial averaging from 3.5 m pixels to 10 m pixels will produce a more spatially homogeneous product which can be compared with in situ measurements. A key parameter in the deterioration of the condition of benthic flora is the reduction in light reaching the seabed. K490 is the relevant physical variable that monitors this process. In summary, we have produced a quantitative set of data at high spatial resolution that covers the entire Ningaloo Reef system for a given period in time (April 2006). The data have been measured with a single instrument so the quality of data across the whole Reef is standardised. The high spatial resolution, sampling on a uniform spatial grid and the characterisation of measurement errors for the bathymetry is particularly beneficial for numerical modelling of coastal processes. It also provides a very accurate baseline for assessing major changes in the system due to geophysical disturbances such as Tsunamis, cyclones and severe storms. The benthic cover mapping also maps the whole Reef at an instant in time and as stated previously, using a single instrument. These data are comparable to data collected using divers and visual assessments where some diver-dependent subjectivity always enters into the monitoring process. As the Ningaloo Reef covers such a vast spatial area, in situ sampling frequently is neither comprehensive in coverage or accuracy nor collected at the same time as other physical data. However, it is clearly the case that these remote sensing methods do not replace the need for in situ assessment. The remote sensing and the diver-based sampling enhance each other. Remote sensing provides total spatial coverage maps or benthic substrate at a coarse level of classification, and the diver sampling provides validation data at discrete points and classification at species level. We have identified several management implications that such data will have. These include the frequency of occurrence and areal coverage of specific marine habitats. This information is important in guiding the establishment of marine protected areas. The vulnerability of particular habitats to weather system will relate to their location. Further, high value marine assets, such as coral reefs, need to be identified and their locations mapped accurately if there is a requirement to be able to protect them from a hazardous event such as an oil spill. The hydrodynamic modelling that will be undertaken using shallow water bathymetric data provided by this project will provide information both on reef currents and within-reef circulation systems which are important variables in developing strategies for the Reef’s management. For example, with such information, the impact of sea level rise on the Reef may be accurately modelled. The WA State agency closely concerned with these above matters is the Department of Environment and Conservation. Wave climate is monitored by the Marine Branch of the Department of Transport. We believe that once the Cluster datasets are released and made publicly available there will be many more benefits that emerge and these in turn will find their way into impacting management practices.
Refereed DesignationUnknown
Hyperspectral mapping of bathymetry and benthic cover.pdf2.02 MB