Abstract for presentation at The 13th Australasian Remote Sensing and Photogrammetry Conference

Seagrass species monitoring in a South Australian shallow coastal site using remote sensing

  • Janet Anstee, CSIRO, Division of Land and Water, Australia
  • Dr Arnold Dekker, CSIRO Land and Water, Australia
  • Dr Vittorio Brando, CSIRO Land and Water, Australia
  • Dr Magnus Wettle, CSIRO Land and Water, Australia
  • Mr Paul Daniel, CSIRO Land and Water, Australia

    • Monitoring coastal and estuarine seagrass meadows, macro-algae and the physical substrata, over a range of spatial and temporal scales, is vital in the assessment of coastal systems. Benthic species can be discriminated on the basis of their spectral, as illustrated by a library of in situ measured spectral signatures of many Australian seagrasses and macro-algae species. We present an approach that estimates the single or combined species composition in each pixel of a remote sensing image. The method involves simulation, based on radiative transfer principles, of the ‘passage’ of a seagrass spectrum through the water column, the air/water interface and the atmosphere to a remote sensor. Subsequently, this process is inverted using optimisation techniques to retrieve the species composition within a pixel, depth of the water column and the amount of bare substrate. In developing a hyperspectral remote sensing based detection and monitoring method, species discrimination is dependent on the accuracy with which the atmosphere, the air/water interface and the water column effects are accounted for. At a South Australia coastal site, airborne hyperspectral data collected in November 2003 which was used for the detection of seagrass and macro-algae. This data will eventually be compared with airborne hyperspectral data derived results from 2001 to illustrate the seagrass change over a period of 2 years. The results form the basis of a spatially referenced database to be used in the long term monitoring of this site and provides a solid basis for determining the operational applications of remote sensing. The results of this study will lead to the improved understanding of the benthic vegetation changes over a period of 2 years, depicting the ecosystem response to environmental stress. These results will provide the basis for the ongoing monitoring of this ecosystem’s modification and the adaptive and sustainable management of the site.

    Conference Organiser - ICMS Pty Ltd