Annual terrestrial ecosystem condition score map for Australia derived using the FACTE framework: 100m, 2013-2022

We mapped ecosystem condition for terrestrial ecosystems of Australia annually using the FACTE framework (Yang et al. in submission). These maps provide a way to identify the temporal change in ecosystem condition from the natural dynamics of ecosystems and the spatial change from alternative natural states of ecosystems.\n\nThe maps are annual ecosystem condition score at 100m for Australia from 2013-2022. Future maps will be updated by roughly a lag of one year. \nLineage: The method is documented in the manuscript below and in Yang et al. (in submission) and in the code provided in Yang et al. (2024). Briefly, each pixel is compared with environmentally similar training sites with known ecosystem condition. The comparison uses 16 Landsat derived ecosystem condition variables that describe the structural, compositional, and functional conditions of the ecosystem. The environmental similarity is based on ten environmental variables that are known to capture vascular plant biodiversity in Australia (Mokany, 2022). The framework uses the distributions of ecosystem condition variables from the training sites to estimate the probability of each pixel being in good condition. The probability is then converted to a condition score for each pixel based on derived relationship between the probability and condition in the training sites. The predicted ecosystem condition score ranges from 0 to 100 with 0 indicating complete loss of native biodiversity while 100 indicating full capacity to support native biodiversity. \n\nReference:\nYang et al. (in submission) A framework for dynamic assessments of terrestrial ecosystem condition\nYang et al. (2024) FACTE: A framework for dynamic assessments of terrestrial ecosystem condition. https://data.csiro.au/collection/csiro:62663\n