NSW Interim Native Vegetation Extent (2008-v1)

Dataset description contents:

Note: this dataset description is metadata (data about data) which describes the actual dataset in accordance with the ANZLIC (Australia New Zealand Land Information Council) Core Metadata Guidelines Version 1


Citation

ANZLIC unique identifier: ANZNS0208000244

Title: NSW Interim Native Vegetation Extent (2008-v1)

Custodian: NSW Department of Environment and Climate Change

Jurisdiction: New South Wales


Description

Abstract:

Various vegetation Extent and Nativeness products, that include: the delineation by woody and non-woody; and maps of the uncertainty associated with assignment of extent and nativeness. The primary product is a NSW Native Vegetation Extent layer prepared for the National Land and Water Resources Audit 2008. This represents native vegetation extent at 2006. A number of other useful intermediate products are also available. Full report is available from http://maps.environment.nsw.gov.au/

VEGETATION EXTENT was derived from interim Foliage Projected Cover (FPC) data (FPC 00,02,04,06) prepared by the DECC\\\'s Woody Vegetation Change Detection Program in early 2008. FPC is generated from Statewide Landcover And Trees Study (SLATS) methodology, developed in Queensland by Qld Dept of Natural Resources and recently applied in NSW. The SLATS program applies a series of algorithms on Landsat TM data to produce FPC values ranging from 0-100 percent. A threshold (or thresholds) are applied to the FPC data to separate woody from non-woody vegetation. NATIVENESS was derived from interim NSW Land-use Mapping data (Emery et. al), and then used to ascertain whether vegetation may be native or non-native.

CAVEATE:

span:

Vegetation extent and nativeness layers generated for this project are interim in nature and are subject to ongoing refinement. While these products represent the best available estimate of vegetation extent and nativeness at this time they will be improved overtime and hence are not suitable for accurate reporting of vegetation change.

ANZLIC search words:

Spatial domain:

Geographic extent name:
NEW SOUTH WALES - State or Territory - New South Wales

Note: The format for each Geographic extent name is: Name - Identifier - Category - Jurisdiction (as appropriate)

Geographic bounding box:
The bounding box encloses the maximum extents of the dataset. There may be voids or gaps within the bounding box, depending on the defined coverage of the dataset. Outer bounding rectangle as Latitude/Longitude coordinates based on the WGS84 datum (expressed as both Degrees Minutes Seconds and as decimal Degrees):

austral

spatial-representation:

sprep-type:

sprep-raster:

cell-size:

25

num-rows:

40836

num-cols:

49416

Data currency

Beginning date: Not Known
Ending date: Current

Dataset status

Progress: In Progress
Maintenance and update frequency: As required

ds-version:

ds-vnum:

08V1

ds-vdate:

20080530

Access

Stored data format:
Digital DIGITAL ArcGIS GRID
Available format type:
Digital DIGITAL - ArcGIS GRID
Access constraints:

Publically available, all thought this data is subject to several caveats. see http://maps.environment.nsw.gov.au/


Data quality

Lineage:

Extent was derived from an interim foliage projected cover (FPC) data based on 2000,02,04,06 Landsat TM epochs. FPC was prepared by DECC's Woody Vegetation Change Detection Program with assistance from Qld Dept of Natural Resources, using the SLATS methodology. Nativeness (native or non-native) was derived from draft Land-use mapping data provided by DECC's NSW Land-use Mapping Program (Emery et al.). It is intended that updated native vegetation extent products will be generated over the next 12mths.

Positional accuracy:

Not known. Landsat TM data utilised in SLATS program were registered to NSW SPOT5 2004/5 data. Landuse data were derived from varied sources, including API, Spot5 and Landsat

Attribute accuracy:

As the FPC values used in the derivation of woody/non-woody extent are essentially continuous data, the choice of an appropriate threshold is critical to maximise accuracy in a binary classification. This issue is complicated by the fact that an appropriate threshold for one part of the state (eg; eastern) is too conservative (regarding woody vegetation) in western areas. The threshold selected for this dataset version has erred on the conservative side for woody vegetation. An indication of the uncertainty regarding the woody/non-woody threshold is demonstrated in a parallel dataset which further categorises woody and non-woody classes into definite or probable sub-classes.

The accuracy of the native/exotic sub-classes is dependent on the accuracy of the landuse information and the interpretation of landuse classes. That is, each landuse class had to be assessed as to whether vegetation in that class would most likely be native or exotic. Some landuse classes were relatively easy to assess (eg; andquot;Native forest andquot;), whilst many were more difficult. As a general rule, landuse classes that could not be assessed as exotic were assigned a native sub-class. Therefore, the classification into native or exotic classes has erred in the favour of potential nativeness.

An accuracy assessment was carried out using 300 random points across NSW. The reference data for each point was derived from the visual interpretation of SPOT5 imagery. This assessment suggested an overall accuracy of 92.3 percent, with a Kappa co-efficient of 0.77869.

Logical consistency:

Data exist as co-registered raster datasets. Only simple attribute structure, with field andquot;label andquot; commonly used to describe cell values.

Completeness:

Woody and non-woody vegetation extent covers all of NSW. Native and exotic subclasses cover almost all of NSW, except for Sydney and other small areas where landuse data are currently unavailable.

Vegetation classifications are not finalised. Refinement of SLATS FPC and watermask data is continuing. The native/exotic subclasses may change as landuse information/interpretation is refined, and other potential data sources are examined.


Contact information

Contact organisation: Department of Environment and Climate Change, NSW
Contact position: Data Exchange Officer
Postal address: PO Box 1967
Locality: Hurstville
State: NSW
Country: Australia
Postcode: 2220
Telephone: 02 95856684
Facsimile: 02 95856466

Metadata information

Metadata date: 20 June 2008 (2008-06-20)
Metadata Reference: http://canri.nsw.gov.au/nrdd/records/ANZNS0208000244.html

Additional metadata

***Key Layers and Attributes***

strong:

vegext_08v1
=

u:

Vegetation Extent
dataset: binary woody and non-woody vegetation extent classification derived from interim SLATS FPC (2000,02,04,06) data.

strong:

Codes
: 0 = other; 3 = woody; 7 = non-woody; 9 = no data. Has some data gaps due to missing nativeness (landuse) data

strong:

vegext1_08v1
=

u:

Vegetation Extent with confidence
dataset: 4-class woody and non-woody vegetation classification derived from interim SLATS FPC (2000,02,04,06) data.

strong:

Codes
: 0 = other; 1 = non-woody (most likely); 2 = non-woody (likely); 3 = woody (likely); 4 = woody (most likely). Has some data gaps due to missing nativeness (landuse) data

strong:

wdy_nat
=

u:

Woody Nativeness with confidence
dataset: woody native, non-native, and indeterminate classification derived from land use data.

strong:

Code
s: 0 = no data; 1 = native; 2 = non-native; 3 = indeterminate. Has some data gaps due to missing nativeness (landuse) data

strong:

nwdy_nat_08v1
=

u:

Non-woody Nativeness with confidence
datset: non-woody native, non-native and indeterminate classification from land use data.

strong:

Codes
: 0 = no data; 10 = native; 20 = non-native; 30 = indeterminate. Has some data gaps due to missing nativeness (landuse) data

strong:

vegtype_08v1
=

u:

Native Veg Extent
dataset: 4 class native/non-native woody and native/non-woody extent classification derived from combination of binary vegetation extent data and nativeness classifications. Has some data gaps due to missing nativeness (landuse) data

strong:

vegtype1_08v1
=

u:

Native Veg Extent with Confidence
dataset: 8 Class native/non-native woody and native/non-woody extent uncertainty classification derived from combination of 4-class vegetation extent data and 2-Class nativeness classifications. Has some data gaps due to missing nativeness (landuse) data.

strong:

Codes
: 1 = other; 2 = woody (likely) andndash; native; 3 = woody (most likely) andndash; native; 4 = non-woody (most likely) andndash; native; 5 = non-woody (likely) andndash; native; 6 = non-woody (likely) andndash; non-native; 7 = non-woody (most likely) andndash; non-native; 8 = woody (likely) andndash; non-native; 9 = woody (most likely) andndash; non-native

strong:

vegtype2_08v1
=

u:

Hybrid Native Veg Extent
dataset: Keith and Simpson (2006) native vegetation extent data used to fill in data gaps in vegtype1_08.

strong:

Codes:
0 = other; 1 = non-woody (most likely) - native; 2 = non-woody (most likely) andndash; non-native; 3 = non-woody (likely) - native; 4 = non-woody (likely) andndash; non-native; 5 = non-woody (K andS) - native; 6 = woody (most likely) - native; 7 = woody (most likely) andndash; non-native; 8 = woody (K andS) native; 9 = woody (likely) - native; 10 = woody (likely) andndash; non-native

Does not include offshore islands such as Lord Howe.



content:

attrinf:

attrib:

attrname:

value1

attrdesc:

codes for vegetation extent classes

attrib:

attrname:

label

attrdesc:

textual description of vegetation extent classes

coordref:

hdatum:

Geodetic Datum of Australia (GDA94)

proj:

Lamberts Conformal Conic

prjparam:

Central Meridian = 147 degrees False easting = 9300000 False northing = 4500000 Standard parallel 1 = -30.75 Standard parallel 2 = -35.75 Latitude of origin = -33.25 Linear metre = 1.0

This dataset description does not comply (has content errors, XML is not valid) with ANZLIC Core Metadata Guidelines Version 1 and XML DTD ANZMETA-NRIMS Page 1 v1.0 (metadata validation date: 2008-08-19)