RFI - PA2906-Coastal Hazard Assessment-R2 · MCU21/0212

                                         Internal use only




3.1         Satellite Imagery
Satellite imagery can be used to determine shoreline position and behaviour from 30+ years of satellite
imagery. Satellite platforms include Landsat 5 (1984 - 2013), 7 (1999 - present), 8 (2013 - present) and
Sentinel-2 (2015 - present), with pixel resolution ranging between 10m and 15m and image capture
varying from 5 to 16 days.

Two independent tools have been developed to analyse the satellite imagery, which are:

     1. CoastSat a Python tool that enables extraction of shoreline position from individual satellite
        images at any coastline worldwide using sub-pixel extraction methods (Vos et al., 2019a, 2019b).
        While the horizontal errors for detected shoreline range have been reported to be between 7.3m
        and 12.7m Root Mean Square Error (RMSE) considering individual images, the bias of the data is
        much lower (Vos et al., 2019a). As such, the temporal resolution of the data means that signals
        can be discerned at seasonal and greater timescales.

     2. Digital Earth Australia (DEA) is a tool to accurately map the boundary between land and water
        (the ‘waterline’) (Bishop-Taylor et al, 2019). The tool uses the same sub-pixel extraction method
        as CoastSat. However, DEA combines aerials to provide a single, annualised shoreline position
        and is therefore suited to determining long-term trends.

Outputs from both CoastSat and DEA are available online, with key outputs for the coastline adjacent to
the site are provided in Figure 9 to Figure 11.

Figure 9 from DEA indicates that the position of the shoreline in the immediate vicinity of the site has
prograded by approximately 0.5 metres annually (±0.3 metres at 95% confidence interval). However,
minimal progradation (less than ±0.1m/yr) is reported at transects nearby to the north and south. Further,
the location/s where accretion is reported (i.e. blue dots in Figure 9) generally coincides with the location
of drainage channels across the beach. The shoreline position from the DEA dataset in Figure 10
indicates that the beach has generally been stable from 1990 to present. At locations where progradation
has been reported in Figure 9, the shoreline was generally further landward in 1988 with rapid
progradation from 1988 to 1989. Given the progradation during this time was localised, it is inferred that
freshwater flows and/or rip heads caused localised erosion of the beach around 1988, which is otherwise
stable. The localised erosion at commencement of the reporting interval has resulted in the reported
accretion.

The CoastSat time series in Figure 11 indicates that the shoreline at this location was relative stable.
Figure 11 indicates the shoreline fluctuated by approximately ±30m from the mean shoreline position
annually, some of which may be due to the RSME of the dataset. However, it is likely that episodic
recession and progradation would occur due to annual beach rotation and storm events.

Care should be taken when analysing shoreline recession to ensure the data captures long term trends
rather than annual or interdecadal variability. Long-term shoreline recession is adopted as 0m/yr.




15 October 2021   66 SEAVIEW TERRACE, SUNSHINE BEACH                 PA2906-RHD-ZZ-XX-RP-Z-0001     12