Saltmarsh ecosystems in Australia provide essential ecosystem services, yet despite being one of Australia’s most widespread coastal ecosystems, their spatial extent remains largely unquantified. Despite significant recent advances in cloud-based geospatial analysis platforms and free access to immense earth observation archives, Australia-wide monitoring of saltmarsh ecosystems has remained unfeasible. This is largely due to the considerable effort required to develop large, analysis-ready, reference datasets suitable for training and testing remote sensing classification models.
This project aims to progress our ability to develop nationwide, highly accurate maps of the distribution of saltmarsh ecosystems. We have developed a set of occurrence records (more than 10,000-point records across Australia's coastline) that are suitable to deploy as training and validation data in newly developed remote sensing classification models [1-3].
Our remote sensing classification approach uses up to 100 nationwide remote sensing-derived covariate layers to support a suite of machine-learning classification models. The classification models are tasked with estimating the coastal ecosystem type of every 30-m pixel that occurs around Australia's coastal zone. This mapping framework will be implemented as a collaborative effort between the Clean Energy Regulator, JCU, UNSW and Digital Earth Australia.