Geomorphic insights into Australia's coastal change using a national dataset derived from the multi-decadal Landsat archive

Australia has a vast, diverse and dynamic shoreline, and national perspectives on the mechanisms, magnitude and trajectories of coastal adjustment are required to underpin informed management. Geoscience Australia's national Digital Earth Australia Coastlines (DEA Coastlines) was developed to help address this need. These datasets were derived by the application of sub-pixel analysis techniques to the Landsat archive to provide annual coastal contours for the years 1988–2019, and accompanying national point datasets provide change statistics at two scales (30 m and 5 km alongshore spacing). This study examines the utility of these three datasets for understanding temporal and spatial patterns of geomorphic change for representative case examples around Australia. These examples have been selected to demonstrate DEA Coastlines capability for examining the dynamics of Australia's diverse shorelines, from the mangrove-fringed, macrotidal systems of north-western Australia to the microtidal clastic barrier systems of the south-east. We examine this capability at tertiary to sub-tertiary coastal compartment (TCC; sediment cell) spatial scales. While some TCC are strongly and consistently net-progradational or net-erosional, most TCC demonstrate alongshore variations in trajectories and rates of change. We use these patterns of change to infer sub-TCC sediment pathways, and examine the temporal trends contained within representative point datasets to link changes to known process drivers. DEA Coastlines also captures diminishing widths of coastal barriers and has the potential to be combined with more detailed site analyses to forecast future trajectories of change. There are, however, some limitations to DEA Coastlines. Owing to the annual scale of the analysis used to develop the datasets, rapidly moving intertidal bars can introduce considerable error, and an uncertainty field successfully flags most of these locations. Other overlapping and complex coastline contours can sometimes be unpacked into more discrete phases to provide a richer understanding of non-linear coastal trends, and these examples highlight the need to simultaneously consider both of the DEA Coastlines 30 m point statistic and contour datasets. The Australian DEA Coastlines dataset thus provides nationally consistent perspectives for improved coastal management, and has been made freely available online to support a broad range of end users. While these datasets are presently exclusive to the Australian coast, the approach also lends itself to future application to the global Landsat archive.