Biomass Plot Library

Biomass Plot Library - National collation of tree and shrub inventory data, allometric model predictions of above and below-ground biomass, Australia

Figure 1: Location of stem inventory sites across Australia with allometric estimates of biomass

Link to the data

These datasets are not static - they are directly created from PostGIS database records and will change with ongoing data collection and improvements to the biomass plot library.

These data are currently made available via a field data portal. Information on the details of the attribute fields for the site and tree level attributes are located in this document in the section titled 'Observation attributes'.

Only the site level mean biomass estimates are currently available. The tree level data and error propagation results will be made available by May 2016.

Data licence and Access rights

Overview

The Biomass Plot Library is a collation of stem inventory data across federal, state and local government departments, universities, private companies and other agencies. It was motivated to the need for calibration/validation data to underpin national mapping of above-ground biomass from integration of Landsat time-series, ICESat/GLAS lidar, and ALOS PALSAR bacscatter data under the auspices of the JAXA Kyoto & Carbon (K&C) Initiative (Armston et al., 2016). At the time of Version 1.0 publication 1,073,837 hugs of 839,866 trees across 1,467 species had been collated. This has resulted from 16,391 visits to 12,663 sites across most of Australia's bioregions. Data provided for each project by the various source organisation were imported to a PostGIS database in their native form and then translated to a common set of tree, plot and site level observations with explicit plot footprints where available (Figure 2).

Figure 2: Plot measurement spatial database

The above-ground, below-ground and total biomass (live+dead) estimation was implemented using the generic allometric models for Australian plant functional types developed by Paul et al. (2016a,b) and decay correction factors implemented by Lucas et al. (2010). Site level estimates of biomass error were derived through comprehensive analysis of measurement and allometric model prediction errors (Roxburgh & Paul, 2016), propagating the random error (precision) associated with the development of the allometric models, all the way through to the prediction of total site biomass (Figure 3). In total eight sources of error are included, with four of these relating to errors encountered during allometric model development, with additional errors arising from diameter measurements during field inventory, allometric model prediction errors, and plot sampling and areal extent errors during field inventory. The tree level error propagation and biomass predictions were undertaken using the Carbon Analysis Tool (CAT), a software system developed by CSIRO for the Department of the Environment in support of vegetation-based greenhouse-gas mitigation activities under the Emissions Reduction Fund.

Figure 3: Biomass plot library workflow

Spatial and Temporal extents

Point of contact

Credit

The following organisations contributed to the Biomass Plot Library:

• Commonwealth Scientific and Industrial Research Organisation (CSIRO)
• Flinders University, South Australia
• Victorian Department of Environment, Land, Water and Planning (DELWP)
• Queensland Herbarium, Department of Science, Information Technology and Innovation (DSITI)
• Western Regeneration Pty. Ltd.
• Department of the Environment, Australia
• Charles Darwin University, Northern Territory
• Darwin Centre for Bushfire Research
• NSW Office of Environment and Heritage (OEH)
• TERN Ausplots
• TERN Supersite Network      
• University of New South Wales
• University of Queensland
• Queensland Department of Agriculture and Fisheries (DAF)
• Forestry Tasmania
• Queensland Remote Sensing Centre, DSITI
• Department of Parks and Wildlife, Western Australia
• New South Wales Forestry Corporation
• Joint Remote Sensing Research Program

Funding for development of Biomass Plot Library and Field Data Portal was largely provided by the Queensland Department of Science, Information Technology, and Innovation through the Joint Remote Sensing Research Program.

Funding for development of the functional type allometric models was largely provided by the Department of the Environment, Australia.

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

Site level

• Spatial domain - preliminary analysis indicates a relative paucity of observations in the Great Western Woodlands and Kimberley regions in Western Australia, and the savanna woodlands of Cape York in Queensland. An assessment will be presented in a forthcoming publication.
• Temporal domain - the majority of sites were surveyed between 1995 and 2015, which is coincident with most satellite sensor data used to support regional to global above-ground biomass mapping. Site prior to 1995 are present, however additional sites will be added to the Biomass Plot Library in future in collation with state agencies that maintain long-term plot monitoring networks.
• Attribute accuracy - A number of site observation locations were not measured using GPS units that could achieve positional accuracy <10 m. This is generally only the case for sites last surveyed prior to 2000 and is continuously improving as sites are revisited. Projects affected include long-term monitoring programs:
  • NSW Forestry Corporation commercial assets
  • WA Parks and Wildlife South-West Forests
  • Qld DSITI Herbarium Native Forest Permanent Plot Network (NFPP)
• The Biomass Plot Library is a post-hoc collation of stem inventory data, therefore the emergent spatial and temporal distribution of data is not designed to meet the specific needs of many applications. Caveats on use include and are not limited to:
  • The Biomass Plot Library data is not the result of a stratified random sampling design, therefore should not be used directly for estimation of regional carbon stocks. This is especially the case for some projects that are biased towards large trees (e.g. sites established by Forestry Tasmania).
  • Many plots are very small in size (<0.1 ha) and should be considered in methods used to link these field data with remote sensing data.

Tree level

• The functional type allometrics from Paul et al. (2016a,b) are not applicable to fern, palm, cycad, grass tree, and pandanus species. These are presently excluded from the Biomass Plot Library.
• Trees less than 5 cm diameter were not included in the site level observations, so have been excluded from the tree level observations.
• A summary of the individual tree/shrub error sources, and their relative contributions as summarised over the whole dataset, are given in Table 1. Overall, three of the error sources accounted for >98% of the total variability, with the dominant error source being the allometric model prediction error.

Table 1: The relative tree level error contributions summarised over the whole dataset   

References

Related products

Product version history

Observation attributes

Site level

Tree level

Metadata history