Creative desktop methods for plot-based monitoring of Australia’s ecosystems—the next best thing to fieldwork in a time when fieldwork is not the best thing.
Enter digital innovations to traditional methods
In true innovative TERN spirit, we quickly looked to digital desktop solutions that would enable us to continue monitoring and delivering open data on changes in the environment.
Through partnership with the Food and Agriculture Organisation of the United Nations (FAO), the team has begun integrating traditional field-collected data with the visual assessment of recent and historical earth observation images using Open Foris’ Collect Earth software.
The Collect Earth tool enables land use and vegetation assessment through freely available, very high-resolution satellite imagery, using the interface of Google Earth, and access to 40 years of satellite data through Google Earth Engine. It is allowing TERN to remotely collect information on major disturbance, such as from fire and flood, and changes in vegetation cover.
“The researcher can ‘fly’ to the sites, all thanks to high-resolution satellite imagery. Once at a ‘site’, the researcher can then move back through the time-series satellite image archive to assess the date-stamped current estimate of cover—a measure of the rapidity of change occurring—if and when there is any evidence of fire in the image archive, and visually assess fire severity.
The researcher can also link to TERN field-collected data to quantify the accuracy of the visual interpretation of images method and what and when other major disturbance has occurred at TERN’s plots.”
Fire history data from >750 TERN plots
Using a multi-evidence-based approach consisting of historical aerial images, Sentinel-2, Landsat 7/8 and MODIS spectral imagery, as well as Normalised Difference Visualisation Index (NDVI) graphs, TERN’s Ecosystem Surveillance team has been recording fire history across the national network of TERN sites, including the determination of fire severity based on level of canopy burnt.
Using the imagery available in Google Earth and Bing Maps, vegetation cover and disturbance over time are also being assessed based on abundance of trees versus shrubs, grass, and bare ground or other.
The work is set to provide a really valuable add-on to our field-collected datasets. In addition to tracking fire or drought recovery over time, the information provides an extra filtering tool to researchers and land managers looking to use TERN data and samples.
It will also be a valuable supplement to our site selection process for new TERN plots and site revisits.
The process is not without its challenges, though. The Australian landscape, vegetation type and growth form, as well as soil colour, potentially present some difficulties due to their effect on the spectral images. However, the multiple sources of evidence enable the team to effectively overcome these challenges and continue TERN’s critical data collection and environmental monitoring.
For more information, see: www.tern.org.au/news-collect-earth/