Brief description
For some time, Remote Sensing Sciences, has produced Foliage Projective Cover (FPC) using a model applied to Landsat surface reflectance imagery, calibrated by field observations. An updated model was developed which relates field measurements of FPC to 2-year time series of Normalized Difference Vegetation Index (NDVI) computed from Landsat seasonal surface reflectance composites. The model is intended to be applied to Landsat and Sentinel-2 satellite imagery, given their similar spectral characteristics. However, due to insufficient field data coincident with the Sentinel-2 satellite program, the model was fitted on Landsat imagery using a significantly expanded, national set of field data than was used for the previous Landsat FPC model fitting. The FPC model relates the field measured green fraction of mid- and over-storey foliage cover to the minimum value of NDVI calculated from 2-years of Landsat seasonal surface reflectance composites. NDVI is a standard vegetation index used in remote sensing which is highly correlated with vegetation photosynthesis. The model is then applied to analogous Sentinel-2 seasonal surface reflectance composites to produce an FPC image at Sentinel-2 spatial resolution (i.e. 10 m) using the radiometric relationships established between Sentinel-2 and Landsat in Flood (2017). This is intended to represent the FPC for that 2-year period rather than any single date, hence the date range in the dataset file name. The dataset is generally expected to provide a reasonable estimate of the range of FPC values for any given stand of woody vegetation, but it is expected there will be over- and under-estimation of absolute FPC values for any specific location (i.e. pixel) due to a range of factors. The FPC model is sensitive to fluctuations in vegetation greenness, leading to anomalies such as high FPC on irrigated pastures or locations with very green herbaceous or grass understoreys. A given pixel in the FPC image, represents the predicted FPC in the season with the least green/driest vegetation cover over the 2-year period assumed to be that with the least influence of seasonally variable herbaceous vegetation and grasses on the more seasonally stable woody FPC estimates. The two-year period was used partly because it represents a period relative to tree growth but was also constrained due to the limited availability of imagery in the early Sentinel-2 time series. The FPC dataset is constrained by the woody vegetation extent dataset for the FPC year.Lineage
The FPC model was developed using: The national data set of historic star-transect field data (Muir et al, 2011) and two-year (eight-season) time series of Landsat seasonal surface reflectance composites.
Landsat imagery was downloaded from the United States Geological Survey (USGS) as radiance and converted to surface reflectance using Flood, 2013a, followed by seasonal surface reflectance using Flood, 2013b.
The FPC model was fit between the field measured green fraction of mid- and over-storey foliage cover and the minimum value of NDVI calculated from 2-years of Landsat seasonal surface reflectance composites. This is expected to represent the driest pixel over the two-year period, where there is the least influence of green understorey on the surface reflectance.
Sentinel-2 seasonal surface reflectance mosaics were computed from the Sentinel-2 source data using the same procedures as for Landsat (Flood, 2013a; 2013b).
The FPC model was applied to a 2-year (8 season) time series of Sentinel-2 seasonal surface reflectance composites to produce an FPC image.
Finally, the 2018 Woody Extent data set was used to reset FPC values in non-woody regions to 'no data' to eliminate over-estimation of FPC in green pastures, cropping regions and other nonwoody landscapes.
Notes
CreditWe at TERN acknowledge the Traditional Owners and Custodians throughout Australia, New Zealand and all nations. We honour their profound connections to land, water, biodiversity and culture and pay our respects to their Elders past, present and emerging.
Foliage projective cover (FPC) is a metric of vegetation cover used in many Australian vegetation classification frameworks. The Statewide Land and Trees Study (SLATS) uses the FPC metric derived from Sentinel-2 imagery to provide broad estimates about the range of tree and shrub densities represented in woody vegetation across Queensland.
Data Quality Assessment Scope
local :
dataset
<br>The data set is generally expected to provide a reasonable estimate of the range of FPC values for any given stand of woody vegetation, but it is expected there will be over- and under-estimation of absolute FPC values for any specific location (i.e. pixel) due to a range of factors. Error is also expected in areas of consistent cloud cover and urban areas, due to interference from the cloud mask.</br>
<br>The FPC model is sensitive to fluctuations in vegetation greenness, leading to anomalies such as high FPC in locations with consistently green herbaceous or grass understoreys over the 2-year period.</br>
<br>The FPC is reset to 'no-data' for pixels which are non-woody in the Woody Extent data set. Woody patches smaller than the Woody Extent Minimum Mapping Unit (0.5 ha) will have FPC='No Data'. Likewise, nonwoody patches smaller than the MMU may have a non-zero predicted FPC.</br>
<br></br>
Data Quality Assessment Result
local :
Quality Result
The model Root Mean Square Error (RMSE) is 9.1 FPC points.
Created: 2018-01-01
Issued: 2024-09-03
Modified: 2024-09-23
Data time period: 2018-01-01 to 2022-12-31
text: State of Queensland.
User Contributed Tags
Login to tag this record with meaningful keywords to make it easier to discover
Point-of-truth metadata URL
Flood N. (2013a). An operational scheme for deriving standardised surface reflectance from Landsat TM/ETM+ and SPOT HRG imagery for Eastern Australia. Remote Sensing, 5(1): 83-109
doi :
https://doi.org/10.3390/rs5010083
Flood N. (2013b). Seasonal composite Landsat TM/ETM+ images using the medoid (a multi-dimensional median). Remote Sensing, 5(12): 6481-6500
doi :
https://doi.org/10.3390/rs5126481
Armston J.D., Denham R.J., Danaher T.J., Scarth P.F. and Moffiet T. (2008). Prediction and validation of foliage projective cover from Landsat-5 TM and Landsat-7 ETM+ imagery for Queensland, Australia. Journal of Applied Remote Sensing, 3: 033540-28
doi :
https://doi.org/10.1117/1.3216031
Muir J., Schmidt M., Tindall D., Trevithick R., Scarth P., and Stewart J. (2011). Field measurement of fractional ground cover: a technical handbook supporting ground cover monitoring for Australia. Technical report, Australian Bureau of Agricultural and Resource Economics and Sciences, Canberra, ACT
uri :
https://www.researchgate.net/publication/236022381_Field_measurement_of_fractional_ground_cover
- URI : geonetwork.tern.org.au/geonetwork/srv/eng/catalog.search#/metadata/c65ad708-e270-431a-bb5b-13f1a4ec13db
- global : c65ad708-e270-431a-bb5b-13f1a4ec13db