Printed from the TERN Data Discovery portal
The maximum number of records allowed to be saved is 20.You have exceeded the limit.


153.7687,-27.9675 153.7687,-37.6423 140.6947,-37.6423 140.6947,-27.9675 153.7687,-27.9675

TERN/5e2a56b2-45be-41da-b193-9157ff02bd49 22029

Woody Extent and Foliage Projective Cover - SPOT, OEH algorithm, NSW


This dataset contains maps of woody vegetation extent and woody foliage projective cover (FPC) for New South Wales at 5 metre resolution.

Woody vegetation is a key feature of our landscape and an integral part of our society. We value it because it contributes to the economy, protects the land, provides us with recreation, and gives refuge to the unique and diverse range of fauna that we regard so highly. Yet it poses a significant threat to us in times of fire and storm. So information about trees is vital for a range of business, property planning, monitoring, risk assessment, and conservation activities.

The datasets are:
Woody vegetation extent. A presence/absence map showing areas of trees and shrubs, taller than two metres, that are visible at the resolution of the imagery used in the analysis. This shows the location, extent, and density of foliage cover for stands of woody vegetation, enabling identification of small features such as trees in paddocks and scattered woodlands through to the largest expanses of forest in the State. Woody extent products contain 'bcu' in the file name.

Woody foliage projective cover (FPC). FPC is a measure of the proportion of the ground area covered by foliage (or photosynthetic tissue) held in a vertical plane and is a measure of canopy density. Woody FPC products contain 'bcv' in the file name.

Both mosaics and tiles are available, along with a shape file that identifies the location of the tiles.

We owe a debt of gratitude to the numerous Science Division staff and volunteers who edited the maps. Thanks too, to the following organisations: Airbus Defence and Space for SPOT imagery data NSW Land and Property Information for ADS40 data NSW Land and Property Information and a number of commercial vendors for Lidar data Joint Remote Sensing Research Program

What can the maps be used for? The maps are intended for use in rural landscapes and are suited to many applications including:
- property planning
- vegetation mask for topographic maps
- local government planning
- risk assessment, such as in fire-prone areas
- native vegetation mapping
- habitat identification and mapping

Temporal coverage

From 2011-01-01 To 2011-12-31


  • Date modified


Citation information

How to cite this collection:

Office of Environment and Heritage (2011-2019), New South Wales (2015): Woody Extent and Foliage Projective Cover - SPOT, OEH algorithm, NSW. Version 1.0.0. Terrestrial Ecosystem Research Network (TERN). (Dataset).

Access data

This data can be accessed from the following websites

Access metadata

Source Metadata URL

Rights and Licensing

Creative Commons Attribution 4.0 International Licence

  • Spatial coverage Hide

  • Additional information Show


      Research Groups

      Distributed by


      Image Data
      The source data was SPOT5 High Resolution Geometric (HRG) satellite imagery. It consists of 4 multispectral bands (10 m pixels), and a panchromatic band (2.5 m pixels). A time series of one image per year for the period 2008 to 2011 was acquired during dry periods where the contrast between woody vegetation and the ground cover is high. A total 1256 images were used. The images were registered with ground control. The multispectral imagery was corrected for atmospheric and bi-directional reflectance effects and sharpened to 5 m pixels using the panchromatic imagery. The images were masked for cloud, cloud shadow, topographic shadow, and water.
      Foliage Projective Cover (FPC)
      An estimate of FPC was derived for every clear pixel in every image. This required a multiple linear regression model that related the multi-spectral reflectance to a reference data set of FPC. Each pixel contained up to 5 observations of FPC and reflectance over time. The probability of a pixel containing woody vegetation was determined using a binomial logistic regression model. The model parameters were the mean FPC, mean red reflectance, variation in FPC over time, and the climate variable vapour pressure deficit. The model was trained using 25930 observations of woody vegetation presence or absence. These points were interpreted from ADS40 aerial imagery where available (0.5 m pixels) and SPOT5 HRG panchromatic images (2.5 m pixels).
      Mapping woody vegetation
      Woody vegetation extent was mapped by applying a threshold to the probability images, with further editing by trained analysts. The mean FPC value over time was used to attribute each woody pixel.
      Accuracy assessment
      Two comparisons with independently-derived datasets of woody vegetation extent were performed as described in the Data Quality section.

      Data Quality

      • OEH staff conducted two comparisons with independent observations of woody vegetation extent. The first comparison used fine-detailed maps of woody-vegetation extent derived from airborne Lidar surveys. The state-wide map of extent had an overall accuracy of 90.1%.

        The second comparison used 6670 image-interpreted points of woody vegetation presence or absence. The points were gathered from images with 2.5 m pixels. The overall accuracy was 88%. The spatial variation in accuracy across the state, reported by Local Land Service region, is listed in the table below.

        Care should be taken when interpreting the maps. Incorrect classification is most likely to occur where it is difficult to distinguish trees greater than two metres in height from other types of vegetation. Such vegetation includes sparse woodlands, low shrubs, chenopods, heath, wetlands, and irrigated pastures and crops. Also, woody vegetation is only detected about half of the time when the fol

        Accuracy assessment results are provided in the table below.

        Local Land
        Points Lidar Local Land
        Points Lidar
        North Coast 95.8% 93.6% Riverina 89.0% 93.0%
        Northern Tablelands 91.8% 89.0% Hunter 88.7% 85.3%
        South East 91.6% 94.5% North West 88.3% 89.0%
        Central Tablelands 91.0% 86.8% Murray 84.8% 90.3%
        Greater Sydney 90.6% 89.1% Western 77.5% 88.6%
        Central West 89.8% 88.3%
    • Keywords Show


      • foliage projective cover (Unitless)


      • SPOT-5

      Temporal Resolution

      • Annual

      GCMD Science


      ANZSRC - FOR

      • Wildlife and Habitat Management
      • Crop and Pasture Biomass and Bioproducts
      • Landscape Ecology
      • Environmental Management
      • Environmental Monitoring

      User Defined

      • environment
      • geoscientificInformation
      • farming
      • imageryBaseMapsEarthCover
      • trees
    • 235 Collections with matching subjects