The focus of this chapter lies in describing digital multispectral and hyperspectral remote sensing developments and applications in the mapping and monitoring of seagrass ecosystems. Multispectral refers to a sensor that registers light in a limited number of relatively broad spectral bands (band-widths of20-60 nm); hyperspectral (also referred to as imaging spectrometry) is defined for sensors that measure the entire spectrum under consideration in contiguous narrow spectral bands (bandwidths between 2 and 20 nm).
Currently, seagrass maps are still predominantly being produced from the interpretation of aerial photography although it is likely that airborne and spaceborne remote sensing methods will rapidly take over this role given the advantages they present in terms of accuracy, repeatability, versatility, and information content. Nevertheless, retrospective studies of sea-grass change using the more modern methodologies will still need to make use of results generated by the more traditional methods since aerial photographs are the dominant archival source of historical spatial
* Author for correspondence, email: [email protected]
information on seagrass meadows. The subject of aerial photography and the assessment of seagrass vegetation has been extensively dealt with in Global Seagrass Research Methods (Short and Coles, 2001) and the reader is referred to reviews in that publication, as well as to relevant sections of Green and Short (2003).
Traditional satellite sensors offer a cost-effective approach for mapping seagrass over large areas and in remote locations (Ferguson and Korfmacher, 1997; Mumby et al., 1999) particularly where the meadows are large, monospecific, and continuous. The value of the repeated temporal cover provided by satellite sensors has been shown by Jensen and coworkers who investigated the application of Landsat MSS and SPOT multitemporal data to successfully evaluate change in cattail and seagrass species in the Everglades (Jensen et al., 1995). Multi-date satellite remote sensing is geometrically highly repeat-able and a cost-effective method for detecting large changes in seagrass distribution or extent over time (Robbleeetal., 1991;Zainaletal., 1993; Ward etal., 1996; Macleod and Congalton, 1998). Anstee et al. (2004) detected seagrass and macro-algae change in a shallow coastal tidal lake in Australia using archival Landsat satellite image data from 1988 to 2002, by applying hyperspectral measurement and modeling techniques to the multispectral Landsat images. They were able to determine that Posidonia
A. WD. Larkum et al. (eds.), Seagrasses: Biology, Ecology and Conservation, pp. 347-359. © 2006 Springer. Printed in the Netherlands.
australis cover remained stable whilst Ruppia sp. and Halophila sp. cover varied slightly. The most significant change had occurred in the cover of Zostera capricorni, which had been replaced by cover of macro-algae species like Chara and Nitella or the Zostera had been overgrown by dense epiphytes.
Was this article helpful?