Analysis of vegetation condition using remote sensing technologies
Article first published online: 11 MAY 2006
Ecological Management & Restoration
Volume 7, Issue Supplement s1, page S77, June 2006
How to Cite
Sheffield, K. (2006), Analysis of vegetation condition using remote sensing technologies. Ecological Management & Restoration, 7: S77. doi: 10.1111/j.1442-8903.2006.298_1.x
- Issue published online: 11 MAY 2006
- Article first published online: 11 MAY 2006
Analysis of vegetation condition using remote sensing technologies. Kathryn Sheffield. School of Mathematical and Geospatial Sciences, RMIT University, GPO Box 2476V, Melbourne, Vic. 3001, Australia. Email: firstname.lastname@example.org.
Key words: ground data, remote sensing, spatial resolution, vegetation attributes.
In Australia, several methods of assessing vegetation condition, within a biodiversity context, have been developed (e.g. Habitat Hectares (Parkes et al. 2003) and BioMetric (Gibbons et al. 2004)). These metrics are site-based assessments and use a suite of vegetation attributes to derive a measure of vegetation condition. This research investigates the potential use of remote sensing technologies to enhance site assessment of vegetation, by providing information at the landscape-regional scale.
Research overview. This research uses vegetation attributes such as those outlined in the BioMetric assessment methodology (Gibbons et al. 2004), and explores a variety of classification methods to measure these attributes using remotely sensed imagery. This project is based in southern New South Wales and incorporates agricultural land, travelling stock reserves and National Park. The study area is approximately 60 km × 60 km and is located between the Murray River and Holbrook. This research incorporates two key objectives:
- • Development of tools to derive measures of vegetation attributes from remotely sensed imagery
- • Investigation of the effect of spatial resolution on remotely sensed image classification results
Spatial resolution is an indication of the spatial detail (features) able to be detected in an image and is linked to the pixel size of the imagery. To explore the effect of spatial resolution on attributes of vegetation, a number of data sources, representing different levels of spatial detail, will be used in this research including ground data, Ikonos imagery, SPOT 5 imagery and Landsat 7 + ETM imagery. Pixel sizes will range from 4 m to 30 m.
Ground data. Measurements of selected vegetation attributes will be collected in the field to underpin the classification of remotely sensed imagery within the study area. A key research question that will be addressed is how ground data can be collected in the field to maximize the likelihood of accurately measuring vegetation attributes from remotely sensed imagery.
To explore the relationship between spatial resolution of ground data and remotely sensed imagery, ground data will be collected at a range of spatial detail levels from individual objects (trees) to larger areas (sites). A suite of site sizes will be assessed, which are determined by the pixel sizes of the remotely sensed imagery. Rather than assessing individual pixels, clusters of pixels will be used to address issues such as positional accuracy. These data will ultimately allow an investigation of effective methods to integrate ground data and remotely sensed data, and methods to derive vegetation condition attributes from remotely sensed imagery.
- 2004) BioMetric Decision Support Tool. Version 1.4. Biodiversity Assessment Tool to Support the NSW Property Vegetation Plan Developer. DRAFT User Manual. NSW Department of Environment and Conservation, Canberra, ACT. , , , and (
- 2003) Assessing the quality of native vegetation: The ‘habitat hectares’ approach. Ecological Management & Restoration 4, S29–S38. , and (