The map shows the likelihood of koala occurrence in
- 5-kilometre grids across north-east NSW
- 10-kilometre grids across the rest of NSW
Each grid cell is populated with
- data reflecting the proportion of koalas relative to other common mammals
- a measure of statistical confidence associated with data in the cell
Download the spatial datasets
To download the spatial datasets, please visit our open data portal and search using the keyword `Koala'.
Developing the baseline map
The EPA engaged the Office of Environment and Heritage to develop the map, using current available data from the Atlas of NSW Wildlife. The map will be updated periodically to reflect new records entered in the Atlas.
A report on the preliminary map of the likelihood of koala occurrence in NSW (PDF 3.9MB) was released in February 2015. A supplementary comparison (PDF 4.8MB) found broad agreement between the baseline map and other koala mapping available for the NSW north coast.
Updating the baseline map
The map was updated (PDF 2.4MB) following a review by koala experts, who recommended a number of modifications to the map, including filtering data for duplication and revising the data cut-off date to ensure regionally significant surveys were captured.
The EPA engaged an independent koala expert to update the baseline map. This included using new Atlas of NSW Wildlife data and an extensive set of additional koala survey data for north eastern NSW
The statistical confidence and likelihood levels of the baseline map were then used to direct additional survey effort to data deficient areas.
- filled in the more significant gaps in data
- improved the confidence level of statistical data
- identified koalas in some areas where they were not previously known, while in other areas, established confidence that koalas are not likely to occur
Validation of the revised baseline map showed that there was a strong relationship between the map and koala activity, as indicated by independent survey results.
Figure 1: Likelihood of koalas
Figure 2: Confidence level in the estimated likelihood