ANZLIC Metadata Profile: An Australian/New Zealand Profile of AS/NZS ISO 19115:2005, Geographic information - Metadata
Metadata Standard Version
Reference System Info
Reference System Identifier
ANZLIC LUCAS NZ Forest Clearing 2008 2020 v019
Areas of LUCAS LUM forests that have been cleared between 2008 and 2020 inclusive. Forest clearing is detected using mosaics of satellite imagery captured over a range of dates; therefore the year of clearing provided is nominal.
In order to produce the LUCAS Land Use Maps 1990, 2008, 2012 and 2016 and estimate annual rates of deforestation, the LUCAS mapping programme routinely detects and classifies forest clearing in NZ.
*Detection* Image mosaics used for change detection are cloud-masked, orthorectified, corrected for atmospheric conditions, and reflectance standardised using the Ecosat algorithms documented in Dymond et al, (2001), Shepherd and Dymond (2003) and Dymond and Shepherd (2004).
In order to estimate rates of deforestation for a calendar year, two mosaics of multispectral satellite imagery with shortwave-Infrared bands (such as SPOT 5, Landsat 7, Landsat 8 and Sentinel 2) are compiled (from capture as near as possible to the beginning of January of Year A and Year B respectively). Matching spectral bands from the two mosaics are compared in a stack to detect change.
Manual quality control checking is performed to remove false-positives arising from small clouds or image mis-registration (Newsome et al, 2018).
A nominal year of clearing is automatically derived, but may receive manual correction during the land use intention classification stage.
*Classification* Detection and classification of non-anthropogenic change is a by-product of attempting to find the directly anthropogenic (human-induced) change that is applicable to the LUCAS Land Use Map.
The determination that a change is non-anthropogenic is usually made by judging context and pattern visible in satellite imagery. However, when higher resolution aerial imagery becomes available following the clearing event, it can often be useful in confirming the classification.