Mapping spatial variability in shoreline change hotspots from satellite data; a case study in southeast Australia
Introduction
Quantifying changes in shoreline position remains essential for effective coastal management, including the regulation of coastal hazards, zoning and decisions around coastal protection and climate change adaptation (Wainwright et al., 2014, 2015; Jongejan et al., 2016; Vos et al., 2019). There is growing concern for the stability of shorelines. Around 24% of the world's sandy coasts have shown a significant trend of shoreline retreat in recent decades (1984–2016) (Luijendijk et al., 2018). This number is expected to increase as climate change-forced sea-level rise and shifts in wave climate alter sediment transport pathways (Zhang et al., 2004; IPCC, 2014; Ranasinghe, 2016); and there is an increased focus on the prediction of shoreline behaviour under both current and future climates (Ranasinghe et al., 2012; Hinkel et al., 2013; Ranasinghe, 2016; Toimil et al., 2017; Dastgheib et al., 2018; Le Cozannet et al., 2019). Analyses of past shoreline behaviour have the potential to highlight the primary mechanisms of forcing, in addition to identifying distinct hotspots of contrasting coastal behaviour (e.g. Addo et al., 2008; Maiti and Bhattacharya, 2009; del Río et al., 2013; Burningham and French, 2017; Rosskopf et al., 2018; Castelle et al., 2018). The initial step towards predicting future change is often, therefore, to construct robust estimates of past trends in shoreline position (e.g. Maiti and Bhattacharya, 2009; Vitousek et al., 2017; Le Cozannet et al., 2019). Such analyses should result in a better understanding of the relative importance of local, regional and global drivers of coastal change.
Many shorelines show little regional coherence in shoreline behaviour at various spatio-temporal scales (del Río et al., 2013; Burningham and French, 2017; Harley et al., 2017; Vos et al., 2019). Understanding shoreline behaviour therefore requires comprehensive regional-scale studies which consider site-specific forcing-mechanisms and local geomorphology (Ranasinghe, 2016). Identifying trends in shoreline position, however, can be challenging. Coastal environments are highly dynamic, adjusting continuously across a range of temporal and spatial scales in response to changes in natural processes or modification of the environment by humans. Hydrodynamics (sea level, wave conditions, storm surge, coastal currents, river flow) are a primary driver of coastal change (Stive et al., 2002; Hapke et al., 2016; Bird, 2018). For example, change in wave climates related to large scale climate oscillations (i.e. El Niño Southern Oscillation) are associated with inter-annual time scale shifts in embayed beach shoreline position (Barnard, et al., 2015, Ranasinghe et al., 2004, Short and Trembanis, 2004, Thomas et al., 2011); , while at longer-time scales (centuries-millennia) shoreline variability reflects changes in relative sea level. At intermediate time-scales (years – decades) these hydrodynamic forcing interacts with shoreline morphology (i.e. orientation of the coast, bathymetry), sediment supply, and geology to control shoreline behaviour (Stive et al., 2002; Houser et al., 2008; del Río et al., 2013; Hapke et al., 2016; Ranasinghe, 2016). Coastal behaviour has been additionally altered by anthropogenic interventions; including shore protection structures and decreased fluvial sediment inputs related to dam construction (Bird, 2018; Luijendijk et al., 2018; Mentaschi et al., 2018).
Trends in shoreline position are more readily documented at high spatial and temporal resolutions. In general, rates of shoreline change have been extracted by comparing a limited number of aerial photographs at relatively coarse spatial and temporal resolutions (Dolan et al., 1991; Burningham and French, 2017). Increased availability of satellite imagery now provides alternative relatively temporally dense global datasets from which spatially extensive patterns of shoreline behaviour over the last three decades can be examined. Various techniques to extract satellite derived shorelines have been developed, but typically these apply algorithms to automatically detect shorelines based on the land-water boundary (e.g. Hagenaars et al., 2018; Dai et al., 2019; Vos et al., 2019). Errors include the misclassification of shorelines due to clouds, shadows, surf, and errors arising from the large tidal range on meso-macrotidal coasts, as well as positional offsets related to georeferencing and sensor corrections and image pixel resolution (Feyisa et al., 2014; Hagenaars et al., 2018; Dai et al., 2019; Vos et al., 2019). Techniques to minimise these errors, including the use of composite images, applying cloud filters and tidal corrections to single images, and thresholding techniques for image classification (e.g., Hagenaars et al., 2018; Vos et al., 2019) allow for the analysis of shoreline change at sub-pixel resolutions. Recent applications of satellite derived shorelines to detect coastal change include analyses at single embayment and regional scales (i.e. Maiti and Bhattacharya, 2009; Anthony et al., 2015; Gomez et al., 2014; Vos et al., 2019; Tătui et al., 2019); but also global-scale assessment of yearly shoreline change rates (Luijendijk et al., 2018; Mentaschi et al., 2018). This body of work has now established methodologies that enable assessments of patterns of long-term (decadal) shoreline change, although these studies recommend conducting local and regional verification to better evaluate the competence of these methods across different types of coast. Yet studies which have applied these techniques to the assessment of regional patterns of shoreline behaviour to improve our understanding of coastal change remain few.
This paper presents the first regional analysis of shoreline movements along the 1230 km long wave-exposed coast of Victoria, south-east Australia. Areas of significant shoreline retreat or accretion (hotspots) are identified from a dataset of annual satellite derived shorelines (Luijendijk et al., 2018); with the aim of improving our understanding of the spatial patterns and principal drivers of shoreline change. Victoria has a coast typical of high-energy microtidal coasts, and provides an ideal opportunity to undertake regional-scale analyses of shoreline behaviour due to its varied geology and diverse coastal landforms (Bird, 1993). Analysis of Global Climate Models have shown changes in wave climate and wind-driven coastal currents under climate change scenarios, with the potential to alter sediment transport pathways and storm surge around the Victorian coast (Hemer et al., 2010, 2013; McInnes et al., 2009, 2013; Morim et al., 2019; O'Grady et al., 2015, 2019). In addition, the Victorian coast is highly valued recreationally, and supports many important ecological, economic and cultural assets (VEAC, 2019). The potential for increased coastal erosion resulting in shore retreat due to climate change has been identified as a significant threat to these values, and there is already demand from local communities for new investment in coastal protection works (Victorian Auditor-General’s Office, 2018; VEAC, 2019). Thus, there is an immediate need for an improved understanding of shoreline change in Victoria at a regional scale to inform coastal management decisions. Working through archival imagery provides a unique opportunity to model spatio-temporal patterns of coastal change.
Section snippets
Study region
The coast of Victoria extends from the South Australian border in the west to the New South Wales border in the east (Fig. 1). It comprises approximately 1230 km of ocean-wave-exposed coastline, and three sheltered embayments that were not included in the present study: Port Phillip Bay (262 km in coastline length), Western Port Bay (263 km), and inner coast of Corner Inlet (150 km) (Bird, 1993). The coast is generally oriented to the south, but with alternating southwest and southeast
Derivation of shoreline positions
The magnitude, rate and direction of shoreline change was determined for a 31-year period (1987–2017). Shoreline positions were extracted from the global shoreline dataset generated by Luijendijk et al. (2018) and consisted of a mean annual shoreline position derived from composite images from multiple satellite missions (Table 1). Shorelines were identified as the most probable land-water boundary determined by the Normalised Difference Water Index (NDWI), with an alongshore resolution of
Results
Spatial clustering identified 63 progradational hotspots and 59 recessional hotspots with a combined total length of 72.7 km and 76.6 km respectively (Fig. 2, Fig. 3). Collectively both types of hotspot comprised 13.3% (n = 2387) of the total number of analysed transects (n = 17908); equivalent to approximately 12.1% of the total ocean-coast of the Victorian shoreline. Hotspots were broadly distributed across Victoria but not uniformly, with variation in the proportion of impacted coast and the
Discussion
The regional analysis of shoreline change presented here demonstrates the potential for satellite-derived shoreline datasets to elucidate large scale regional patterns of shoreline behaviour. By analysing long-term trends in satellite-derived shorelines this study contributes a spatially extensive and time-efficient evaluation of shoreline change, allowing for the characterisation of regional patterns of shoreline behaviour. These results show, for the first time: that (i), significant change
Conclusions
There has been an increased focus on predicting shoreline erosion under both current and future climates as concern regarding the impacts of climate change on coastal lands has grown in recent decades. By identifying hotspots of significant change over a 30-year study period through utilising a spatially and temporally rich dataset of shoreline position, this study has shown that the coast of Victoria is generally not recessional. Instead hotspots of both recession and progradation are
Declaration of competing interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Acknowledgements
This project is funded through The Earth Systems and Climate Change Hub of the Australian Government's National Environmental Science Program, The Victorian Coastal Monitoring Program, and the Victorian Government's Department of Environment, Water, Land and Planning. RR is supported by the AXA Research fund and the Deltares Strategic Research Programme ‘Coastal and Offshore Engineering’.
References (82)
Late Interglacial and Holocene trends in sea-level maxima around Australia: implications for Modern rates
Mar. Geol.
(1992)- et al.
Understanding coastal change using shoreline trend analysis supported by cluster-based segmentation
Geomorphology
(2017) - et al.
Spatial and temporal patterns of shoreline change of a 280-km high-energy disrupted sandy coast from 1950 to 2014: SW France. Estuarine
Coast. Shelf Sci.
(2018) - et al.
Coastline extraction from repeat high resolution satellite imagery
Remote Sens. Environ.
(2019) - et al.
Automated water extraction index: a new technique for surface water mapping using Landsat imagery
Remote Sens. Environ.
(2014) - et al.
On the accuracy of automated shoreline detection derived from satellite imagery: a case study of the Sand Motor mega-scale nourishment
Coast. Eng.
(2018) - et al.
Decoupling processes and scales of shoreline morphodynamics
Mar. Geol.
(2016) - et al.
A global analysis of erosion of sandy beaches and sea-level rise: an application of DIVA
Global Planet. Change
(2013) - et al.
Controls on coastal dune morphology, shoreline erosion and barrier island response to extreme storms
Geomorphology
(2008) - et al.
National-scale wave energy resource assessment for Australia
Renew. Energy
(2010)