Seasonal fractional cover - Landsat, JRSRP algorithm, Australia coverage

Seasonal fractional cover - Landsat, JRSRP algorithm, Australia coverage

Link to the data

Data licence and Access rights

Alternate title

Seasonal fractional vegetation cover for Queensland derived from USGS Landsat images

Abstract or Summary

Land cover fractions representing the proportions of green, non-green and bare cover retrieved by inverting multiple linear regression estimates and using synthetic endmembers in a constrained non-negative least squares unmixing model. 

The opening of the Landsat archive has provided an opportunity to composite imagery into representative seasonal images. The benefits of compositing in this manner are the creation of a regular time-series capturing seasonal variability, and the minimisation of missing data and contamination present in single date imagery (Flood, 2013).

Spatial and Temporal extents

Point of contact

Credit

• Joint Remote Sensing Research Program.
• Landsat 5 TM, Landsat 7 ETM+ and Landsat 8 OLI images were acquired from United States Geologic Survey.

Keywords

There are three main thesauri that AusCover recommends:

1. Global Change Master Directory (http://gcmd.nasa.gov)
2. Climate and Forecast (CF) convention standard names (http://cfconventions.org/standard-names.html).
3. Fields of Research codes (http://www.abs.gov.au/ausstats/abs@.nsf/0/6BB427AB9696C225CA2574180004463E?
   opendocument).

Data quality

Horizontal Positional Accuracy

All the data described here has been generated from the analysis of Landsat TM, ETM+ and OLI data, which has a spatial resolution of approximately 30 m. The imagery is rectified using control points measured with a differential GPS ensuring a maximum root mean square (RMS) error of 20 m at these control points. However, it is possible that errors up to ±50 m occur between these control points. The imagery has been corrected for height displacement using a 3" digital elevation model (DEM) based on the National Aeronautics and Space Administration (NASA), Shuttle Radar Topography Mission (SRTM). It is not recommended that these data sets be used at scales more detailed than 1:100,000.

Vertical Positional Accuracy

All the data described here has been generated from the analysis of Landsat TM, ETM+ and OLI data, which has a spatial resolution of approximately 30 m.

The imagery is rectified using control points measured with a differential GPS ensuring a maximum root mean square (RMS) error of 20 m at these control points. However, it is possible that errors up to ±50 m occur between these control points. The imagery has been corrected for height displacement using a 3" digital elevation model (DEM) based on the National Aeronautics and Space Administration (NASA), Shuttle Radar Topography Mission (SRTM). It is not recommended that these data sets be used at scales more detailed than 1:100,000.

Filenaming convention

Filenames for the seasonal fractional cover deciles product conforms to the AusCover standard naming convention. The standard form of this convention is:

<satellite category code><instrument code><product code>_<where>_<when>_<processing stage code>_<additional dataset specific tags>

Details for the unique codes used for this dataset can be found in the following table.

Related products

Landsat Fractional Cover

Persistent Green-Vegetation Fraction and Wooded Mask - Landsat, Australia coverage

SLATS Star Transects

References

Algorithm summary

Image Pre-Processing

All input Landsat TM/ETM+ imagery was downloaded from the USGS EarthExplorer website as level L1T imagery. Images which the EarthExplorer site rated as having greater than 80% cloud cover were not downloaded, on the assumption that they would contribute little, and would add extra noise in the form of undetected cloud, shadow, and mis-registration (which is a greater risk in very cloudy images). The imagery has been corrected for atmospheric effects, and bi-directional reflectance and topographic effects, using the methods detailed by Flood et al (2013).This is a preprocessing scheme for minimising atmospheric, topographic and bi-directional reflectance effects on Landsat-5 TM, Landsat-7 ETM+ and SPOT-5 HRG imagery. The approach involves atmospheric correction to compute surface-leaving radiance, and bi-directional reflectance modelling to remove the effects of topography and angular variation in reflectance. The bi-directional reflectance model has been parameterised for eastern Australia, but the general approach is more widely applicable. The result is surface reflectance standardised to a fixed viewing and illumination geometry. The method can be applied to the entire record for these instruments, without intervention. The resulting imagery is expressed as surface reflectance. Cloud, cloud shadow and snow have been masked out using the Fmask automatic cloud mask algorithm. Topographic shadowing has been masked using the Shuttle Radar Topographic Mission DEM at 30 m resolution. Water has been masked out using the methods outlines in Danaher & Collett, (2006).

Fractional Cover Model

The bare soil, green vegetation and non-green vegetation endmembers for the combination of Landsat-5 TM and Landsat-7 are calculated using models linked to an intensive field sampling program whereby more than 1500 sites covering a wide variety of vegetation, soil and climate types were sampled to measure overstorey and ground cover following the procedure outlined in Muir et al (2011).A constrained linear spectral unmixing is applied to the image archive using the derived endmembers and has an overall model Root Mean Squared Error (RMSE) of 11.6%. Values are reported as percentages of cover plus 100. The fractions stored in the 4 image layers are:  Band1 - bare (bare ground, rock, disturbed), Band2 - green vegetation, Band3 - non green vegetation (litter, dead leaf and branches), Band4 - Model fitting error.

Seasonal Compositing

The method of compositing used in the creation of the seasonal fractional cover product is the selection of representative pixels through the determination of the medoid of three months (a season) of fractional cover imagery.The medoid is a multi-dimensional equivalent of the median. Within the three-dimensional space of the cover fractions, the medoid is the point which minimises the total distance between the selected point and all other points. Thus the selected point is “in the middle” of the set of points, in terms of the cover fractions. Like the median, the value selected is a specific data point and not an averaged or blended value made up of different image layers.The pixel selected is therefore a true representative pixel. In addition, because it selects the centrally located point in the multi-dimensional space, it is robust against extreme values, inherently avoiding the selection of outliers, such as occurs when cloud or cloud shadow goes undetected. For further details on this method see Flood (2013). The nature of the medoid means that for each pixel in the representative image created at least three pixels from the time-series of imagery for the season must be available. Unfortunately, due to the high level of cloud cover in some areas, often three cloud free pixels are not available, resulting in data gaps in the resulting seasonal fractional cover image.

Data storage
A 4 band (byte) image is produced:

• band 1 – bare ground fraction (in percent) + 100
• band 2 - green vegetation fraction (in percent) +100
• band 3 – non-green vegetation fraction (in percent) + 100
• band 4 – Error Layer representing the RMSE between the predicted pixel value and the actual pixel value on a nominal scale of 100 (no error) to 200 (very large error).

Product version history

Metadata history