Ground cover mapping methodology
How do we map ground cover?
We monitor ground cover by analysing thousands of satellite images using a high-performance computing facility.
Scientists at the Remote Sensing Centre follow this process below, which has been validated using approximately 1500 ground reference sites across the state:
- Download Landsat imagery from the United States Geological Survey.
- Pre-process imagery to correct atmospheric and landscape effects and remove cloud contamination.
- Apply algorithms, calibrated to our extensive field measurements of ground cover, which calculate the percentage of green cover, non-green cover and bare ground for each pixel.
- Deliver maps of digital ground cover data to landholders, extension officers, private industry and government for a range of applications and purposes.
Ground cover datasets can be provided as either individual date products or as seasonal composites. Individual date imagery is the data that is captured each time the Landsat satellites pass over. This isuseful for obtaining current information on ground cover.
Seasonal composites are made by selecting the pixels from the individual date imagery, which best represent the range of measurements taken by the satellite throughout the three-month season. These seasonal composites are useful for analysing and monitoring trends over time.
The algorithms may also be applied to newer higher resolution imagery as it becomes available. Work is currently underway using the European Space Agency’s Sentinel-2 imagery. Seasonal fractional cover composites are now available for Sentinel-2.
What products are available?
Three main products are available through the ground cover analysis process. Data is freely available through a variety of online tools and data sources.
The Remote Sensing Centre has developed this approach to measuring and monitoring vegetation cover, which measures land cover as percentages of green vegetation cover, non-green vegetation cover (litter, dead vegetation, branches) and bare ground (soil and rock). Scientists produce fractional cover images by applying the fractional cover algorithm to Landsat imagery. Fractional cover images provide a measure of total vegetation cover – they do not distinguish between over-storey ‘green’ vegetation (i.e. trees and shrubs) and ground layer ‘green’ vegetation (i.e. grasses, herbs, forbs).
Fractional cover is also produced using higher resolution, Sentinel-2 imagery.
The green cover fraction from the fractional cover product can be used to map persistent green cover in the landscape, which often represents woody vegetation. While grass and herbaceous vegetation show seasonal effects such as becoming green in response to rainfall and then drying out, woody vegetation such as trees and shrubs tend to stay green over a longer period of time. Because of this, the persistent green fraction can be separated from ground cover and herbaceous vegetation through an algorithm comparing patterns of variation over time. The persistent green product is helping to understand woody vegetation trends and processes such as encroachment and thickening. These processes can affect the grazing potential of land, influence management actions, and have impacts on water quality.
Ground cover (cover under trees)
The persistent green product is used with the fractional cover product to produce a ground cover product that does not include over-storey and mid-storey woody vegetation. The data derived from this method is called fractional ground cover or simply ground cover. The method is being applied operationally for single-date and seasonal imagery.
Scarth, P., Byrne, M., Danaher, T., Henry, B., Hassett, R., Carter, J. and Timmers, P. (2006). `State of the paddock: monitoring condition and trend in groundcover across Queensland’ in Proceedings of the 13th Australasian Remote Sensing and Photogrammetry Conference, Canberra, Australia, November 2006.
Scarth, P., Roder, A. and Schmidt, M. (2010). ‘Tracking grazing pressure and climate interaction – the role of Landsat fractional cover in time series analysis’ in Proceedings of the 15th Australasian Remote Sensing and Photogrammetry Conference, Alice Springs, Australia, September 2010.Tindall, D., Trevithick, R., Scarth, P., Collett, L., Goodwin, N., Denham, R. and Flood, N. (2014). ‘Ground cover and fire in the grazing lands: RP64G Synthesis Report’, Department of Science, Information Technology, Innovation and the Arts. Brisbane.
For more information about accessing ground cover products, contact our Principal Scientist via email at firstname.lastname@example.org.