Mapping the spatial and temporal distribution of Melaleuca spp on the Magela floodplain between 1950 & 2004, using object-based analysis and GIS 2008
Internal Report 545
Supervising Scientist Division
Department of the Environment, Water, Heritage and the Arts, 2008
- Mapping the spatial and temporal distribution of Melaleuca spp on the Magela floodplain between 1950 & 2004, using object-based analysis and GIS (PDF 2.08 MB)
About the report
Wetland ecosystems within Kakadu National Park have been recognised for their global significance under the Convention on Wetlands of International Importance. Both natural and anthropogenic factors have been identified that may impact and alter wetland ecosystems in the region. Monitoring natural and human induced environmental change has been identified as a priority. Remote sensing technologies offer an effective way to monitor change in the natural environment at a variety of spatial and temporal scales. Aerial photographs have been identified as a useful resource in floristic studies in wetlands of northern Australia; additionally they provide an important record for assessment of historical landscape change. The spatial and temporal change in the distribution of Melaleuca spp on a portion of the Magela floodplain was investigated, using four aerial photographic datasets.
Automated classification of panchromatic and true colour aerial photographs using traditional per-pixel algorithms has been limited due to their fine resolution and reduced spectral properties. Current advances in multi-scaled object-based classification techniques have enabled the successful classification of very high resolution data. The use of a multi-scaled approach has enabled the development of classification methods resembling the way humans interpret an image. In this project, classification of the aerial photographs was undertaken using a semi-automated object-based approach. It was found that environmental conditions at the time of acquisition of the photographs influenced the success of the classification. Estimated accuracy of the four classified datasets ranged between 82% and 90%.
Change analysis was performed on the classified aerial photographs to identify both spatial and temporal change in Melaleuca spp canopy cover. The results indicate that there has been little change in canopy cover (± 3%) over the 54 year period. However, the results identified that the spatial distribution of Melaleuca spp canopy cover has been dynamic across the study area with continual increase in the lower eastern portion of the study area.