Heather Holden

Measuring and modeling water column effects on hyperspectral reflectance in a coral reef environment

Abstract Much attention has been given to hyperspectral remote sensing of benthic habitat recently to quantify spectral signatures, examine linear mixing, map geomorphic zonation, or identify temporal change with varying degrees of confidence and success. Relatively less attention has been given to the effects of the water column on the hyperspectral signal given various water depths and bottom types. Hyperspectral in situ reflectance was measured at both the top and bottom of the water column to examine the effects of the intervening water layer. A radiative transfer model was used to predict the top-of-the-water column reflectance from a large number of close-range measured bottom spectra. The measured and modeled hyperspectral reflectance spectra were examined separately to compare the degree to which different substrate types can be discriminated once the water column is “added” to the spectra. The classification accuracy assessment indicated that the ability to discriminate benthic habitat based on hyperspectral characteristics is limited when the effects of the water column are included as the kappa statistic drops from 0.70 to 0.49.

Change detection of satellite imagery for reconnaissance of stressed tropical corals

Considering the vast area of coral reefs, and the remote nature of many of them, a simple but effective procedure to quickly identify the effect of ecosystem stress from satellite imagery is needed. The stresses may be related to El Nino induced temperature and radiation changes, pollution and situation, or unsustainable fishing practices. The procedure must not be dependent upon in situ measurements that typically have been used to correct beam attenuation through the water column, yet must identify signature changes at a variety of depths. The authors have selected Savusavu Bay of Fiji as a test region. A gold mine was recently recommissioned on the west side of the bay. During an anomalous winter storm, the retaining walls of some effluent storage facilities failed and a variety of chemicals were released through the adjacent river system and dispersed over a large coral reef. The authors have SPOT satellite imagery from before and after the event. Preliminary rsconnaissance indicates severe damage to the corals with colonization by algae. The authors assess the change detected from the multidate SPOT data. The Getis statistic, which is based solely on image spatial characteristics, is evaluated as a tool for change detection. Preliminary examination suggests that it meets the requirements for rapid assessment for environmental change without the need for individual image calibration based upon in situ information.

Hyperspectral versus multispectral imaging for submerged coral detection

Analysis of hyperspectral data has produced encouraging results in the discrimination of common and optically similar coral reef substrates such as healthy corals, bleached corals, sea grass, and algae-covered surfaces, but at the present time, such high spectral resolution data is unavailable from a satellite platform. If currently available satellite imagery is to be used to map and monitor changes in coral reef geographic extent and health, a quantitative procedure must be developed to discriminate healthy coral from other optically similar benthic substrates with coarse spectral resolution. The primary goal of this study is to evaluate the feasibility of using coarse spectral resolution data to map the geographic extent and monitor the changes in coral reef ecosystems. While previous studies have based analysis upon reflectance values extracted from images, an attempt is made here to discriminate common coral reef features using in situ spectral reflectance measurements with spectral resolution equivalent to SPOT HRV data. Results of a one-way analysis of variance suggest that the broad categories of in situ reflectance measurements (n equals 596) can be considered separate populations with respect to broadband reflectance characteristics.

Characterisation of Optical Water Quality in Bunaken National Marine Park, Indonesia

Effective passive optical remote sensing of submerged coral reef ecosystems requires not only appropriate atmospheric correction, but also water column correction. Algorithms accounting for atmospheric effects are fairly well established and readily available, but water column correction algorithms are still under development. Many approaches to water column correction assume horizontal homogeneity and strict adherence to Beer’s Law of logarithmic vertical attenuation, which may not be the case in many coral reef ecosystems. Water column optical properties were measured using a multispectral dropsonde radiometer in Bunaken National Marine Park, North Sulawesi, Indonesia, to examine the vertical and horizontal variability of light in a typical coral reef environment. This largely descriptive case study demonstrates the complexity of the interaction of light in shallow coastal environments with often highly reflective substrata and serves to warn against assumptions of water optical property homogeneity. Downwelling attenuation coefficients are provided for use in water column correction of future remote sensing missions.

Optical water column properties of a coral reef environment: towards correction of remotely sensed imagery

Multispectral vertical water column profiles collected in the field with a dropsonde radiometer provide the necessary information to examine the wavelength-specific extinction of light with increased water depth in a tropical coral reef environment with variable water depths and bottom types. A thorough and descriptive investigation of the anomalous behaviour of light in shallow coral reef and related environments is presented with the objective of evaluating the utility of accurate and reliable remote mapping and monitoring of changes in submerged coral reefs. Cluster analysis results indicate that on a wavelength-specific basis, the majority (no less than 70%) of both the downwelling irradiance and upwelling radiance profiles are similar enough to be grouped into one cluster, but that the anomalous profiles are more than simply outliers. Analysis of variance reveals that a significant amount of variability in both downwelling irradiance and upwelling radiance attenuation coefficients can be attributed to water depth when bottom type remains constant.

Multi-vs. hyperspectral characteristics of common coral reef features

Analysis of hyperspectral data has produced encouraging results in the discrimination of common and optically similar coral reef substrates such as healthy corals, bleached corals, sea grass, and algae-covered surfaces, but at the present time, such high spectral resolution data is unavailable from a satellite platform. If currently available satellite imagery is to be used to map and monitor changes in coral reef geographic extent and health, a quantitative procedure must be developed to discriminate healthy coral from other optically similar benthic substrates with coarse spectral resolution. The primary goal of this study is to evaluate the feasibility of using coarse spectral resolution data to discriminate common coral reef ecosystem features such as healthy coral, macroalgae, and seagrass. Principal components analysis is used to reduce the in situ reflectance dataset to representative spectra and cluster analysis is used to determine the degree of separability by class at different spectral resolutions.