|
|
Project
|
|
|
Land Cover Type Mapping Using Satellite Remote Senses Information
|
|
|
Delivery Organisation
|
|
|
CSIRO
|
|
|
South Coast NRM Project Liaison
|
|
|
Justin Bellanger 9845 8537
|
South Coast NRM and the CSIRO have worked on a project to map land cover using satellite remote sensing information.
This exciting project demonstrated the application of Landstat satellite imagery to the identification and mapping of agricultural crops and pastures at a paddock scale. The project has looked at the ability of Landsat satellite imagery to differentiate between perennial pastures, annual pastures, wheat, canola and Lucerne. In addition to identifying the crops in each paddock and the areas of particular crops in the landscape, the system also provided an indication of the variation in crop health within and between paddocks that may be associated with underlying issues such as salinity and water logging.
Preliminary results have indicated an 85% reliability rate in distinguishing between these land uses at trial sites in Oyster Harbour, Fitzgerald River Catchment and Lake Warden in Esperance. The resolution of the information is at a 30 metre scale (that is 9 pixels per hectare).
This technology could be very useful for broad-acre farmers and other landholders to get a sense of how well their paddocks are fairing and also for government agencies to get a sense of what is being planted and where, and potential yields. Further applications of this technology are likely to allow an assessment of the impact of such issues as changing climate on farm productivity and land degradation.
While the technology is not new and has been used by the sugar industry in Queensland to determine yields for some time, the CSIRO has increased its capacity and found new uses for the data.
South Coast NRM has invested over $55,000 in the project to date.
Project Contact:
Dr. Waqar Ahmad - Principal Research Scientist and Group Leader Spatial Informatics, Centre for Environment and Life Sciences, CSIRO Floreat Laboratories Perth
