Variations in wave climate as a driver of decadal scale shoreline change at the Inskip Peninsula, southeast Queensland, Australia

Highlights

• Storms in SEQ are delineated into two types: ex-tropical and East Coast Lows.

• The Southern Oscillation Index is positively correlated to Hs and storm frequency.

• Periods of sustained La Nina increase Hs by 0.10 m and shift mean wave direction 6° anticlockwise.

• Shoreline erosion is tied to variability in wave height and direction, modulated by underlying ENSO signals.

• Clusters of storms in rapid succession are a major driver of coastal erosion.

Introduction

Waves provide an important process of energy transfer at the ocean-land interface. The transfer of energy from deep-water to the nearshore is controlled by the offshore wave height, direction, and period, as well as the underlying coastal bathymetry. Wave energy is a key driver of morphological change along global coastlines and understanding the temporal and spatial variability within a wave climate is essential for informed coastal management (Gurran, 2008; Harvey and Woodroffe, 2008; Hugo, 2008; Hemer et al., 2013). While sea level fluctuations have received widespread global attention as a driver of shoreline change, variability in wave climate is expected to be the main process influencing coastal morphodynamics on moderate to high-energy sandy coasts globally (Coelho et al., 2009; Hemer et al., 2012; Mortlock and Goodwin, 2015). Changes in both the height and direction of future storm waves have potential to act as drivers in large-scale coastal reorganisation.

A regional wave climate consists of both modal conditions, and conditions specifically related to storm events. While storms provide the energy to mobilise sediment and initiate rapid coastal change, the modal conditions are responsible for beach recovery and the redistribution of sediment onshore (Ranasinghe et al., 2004; Short and Trembanis, 2004). Additionally, a regional storm wave climate may be comprised of several sub-climates originating from a range of directions and synoptic weather systems (Goodwin, 2005; Mortlock and Goodwin, 2015). Differentiating between the sub-types of storm wave climates provides a mechanism of classifying storms as based on their relative frequency and intensity, and ultimately their potential to modify the coast. Although higher energy storms generally tend to induce more substantial beach erosion, other parameters have the capacity to influence the morphodynamic response of the receiving coastline. These include the storm duration, timing between storm events, wave direction, wave period, and coastal orientation (Short et al., 2001; Cooper et al., 2004). For example, higher incident wave power can increase shoreline erosion rates (Sanford and Gao, 2017) and deep-water waves of a moderate intensity but an anomalous direction can drive substantial beach erosion (Harley et al., 2017). The local planform of a coast can also determine the response to storm impacts (Goodwin et al., 2006; Thomas et al., 2010. Headlands can refract waves to alter the nearshore wave direction, as well as change the total energy reaching the shoreline and the proportion of cross vs alongshore transport (Harley et al., 2011; Thomas et al., 2011; Nichol et al., 2016; Davidson et al., 2017). Storm impacts will therefore be determined by the characteristic wave climate of each storm type (i.e. height, direction, period, and duration), and the morphology of each individual coastline.

In this study, the role of variability in the seasonal and decadal wave climate is examined as a driver of shoreline change on the open Fraser coast of southeast Queensland, Australia. The study region provides a proxy for open sandy, drift-dominated coastlines globally with similar counterparts described in New Zealand (Kasper-Zubillaga et al., 2007; Tribe and Kennedy, 20101), Brazil (e.g. Santa Catarina coast: Siegle and Asp, 2007), and the U.S.A (Allen, 1981). While a growing body of literature has focused on classifying the wave climate of southeast Australia, most work has focused on New South Wales (NSW) (Harley et al., 2010; Shand et al., 2011; You, 2011; Mortlock and Goodwin, 2015; Pender et al., 2015) and the Gold Coast (Allen and Callaghan, 1999; Strauss et al., 2007; Splinter et al., 2012). In southeast Australia, three distinct modal wave climates are recognised: (1) E-ESE (direction of 85–105°N, short wave periods of 8–9 s); (2) ESE-SSE (direction of 110–150°N, long periods of 11–12 s); and (3) SE-SSE (direction of 140–160°N, moderate periods of 9–10 s) (Shand et al., 2011; Mortlock and Goodwin, 2015; Pender et al., 2015). Storms waves are generated by: (1) easterly trough lows, also known as ‘east coast lows’; (2) extratropical cyclones; (3) southern secondary lows; (4) inland troughs; and (5) continental lows, with storm types 3–5 increasing in dominance further south along the Australian coast (Splinter et al., 2012; Browning and Goodwin, 2013). Due to a lack of long-term directional wave data, our understanding and classification of these wave climates is often applied to other sectors of the southeast Queensland coast. A shortcoming of this is that for regions located north of Brisbane (−27.45°S, 153.03°E), latitudinal differences result in a shift in regional synoptic conditions that are not accounted for. For example, the Queensland coast north of Brisbane fundamentally differs from NSW as it is more exposed to wave trains propagating from tropical cyclones generated in the Coral Sea (Mortlock and Goodwin, 2015) with the potential to cause major episodes of coastal erosion (Splinter et al., 2012; Nott et al., 2013).

A further underlying control on the variability in wave climate and storm frequency is the El Nino Southern Oscillation (ENSO). In southeast Queensland, it has been suggested that during El Nino events, increased jetstream activity may help trigger more east coast lows, reduce the number of tropical cyclones, and alter the mean wave direction (Allen and Callaghan, 1999; Short et al., 2001; You and Lord, 2008). In northern NSW, El Nino years (Southern Oscillation Index (SOI) ≤ −7) have been linked to periods of lower wave height and an increase in the southerly wave component (i.e. a clockwise rotation in wave direction), while La Nina (SOI ≥ 7) tends to result in higher waves with a dominant easterly direction (i.e. an anticlockwise shift) (Ranasinghe et al., 2004). The change in wave height and direction resulting from ENSO variability in southeast Australia has been linked to decadal scale beach rotation with alternating accretion (erosion) occurring at opposite ends of beaches (Ranasinghe et al., 2004; Short and Trembanis, 2004). ENSO impacts on wave climate variability have not yet been investigated north of Brisbane where the impact on storm frequency and wave height, particularly as associated with ex-tropical storms, could be expected to be equally if not more strongly correlated. In terms of translating the effects of wave climate variability to the morphological response of the shoreline, most prior work in Australia has been undertaken where sediment transport occurs largely within an embayed cell (Short et al., 2001; Short and Trembanis, 2004; Daly et al., 2015). This is a similar trend internationally (Ojeda and Guillén, 2008; Loureiro et al., 2009; Pinto et al., 2009). As many beaches in southeast Queensland are located along open coastlines (e.g. Noosa, Sunshine Coast, and the majority of beaches on Fraser and Stradbroke Islands), it is logical that shoreline response to wave climate variability be determined from an open coast analogue. The Interdecadal Pacific Oscillation (IPO) is a further long-term (15–30 and 50–70 years) climatic oscillation which interacts with ENSO related climate variability (Grant and Walsh, 2001; Salinger et al., 2001; Power et al., 2006; Barnard et al., 2015). Specifically, negative phases of the IPO increase sea-surface temperatures off Queensland and enhance La Nina events, whereas positive phases are associated with cooler water and reduced extra-tropical storm activity.

The present study aims to: (1) identify the wave climate for southeast Queensland based on a 31 year hindcast wave dataset; (2) delineate between different storm climates; (3) consider the role of ENSO as a driver of variation in wave climate; and (4) identify rates and trends of decadal scale shoreline change in response to temporal variability in wave conditions. The identification of different storm wave climates will enable a better understanding of events which most strongly impact upon the shoreline and provides a baseline for future comparison. For instance, small changes in the directional wave height will have implications for the coastal sediment budget and consequentially beach morphodynamics. An important consideration is the change that may occur under projected shifts in global climate, such as an increase in the magnitude of extra-tropical cyclones and the frequency of storm events (Hughes, 2003; Harvey and Woodroffe, 2008; IPCC, 2013).

The open coast of southeast Queensland, Australia, is wave-dominated and microtidal with a spring tidal range of 1.35–1.86 m (Harris et al., 2002). The coastal climate is classified as humid subtropical, consisting of warm, humid summers and mild winters (Peel et al., 2007). The present-day storm wave climate is influenced by the occurrence of tropical cyclones during November–April, most of which develop in the Coral Sea and track southward. On average, about three cyclones per year are observed in the Coral Sea with wave fields impacting the southeast Queensland coast (Allen and Callaghan, 1999), although the number of cyclones which actually make landfall is typically <1 per year (Flay and Nott, 2007). East coast lows are a further storm type influencing the coastline and result from trough intensification over eastern Australia. The interaction of east coast lows with developing high pressure systems to the south can increase the severity and duration of coastal storms (Short and Trembanis, 2004; Callaghan and Power, 2014).

While the wave data in this study are representative of southeast Queensland as a whole, a specific compartment of the coast was used to map decadal scale shoreline change in close proximity to where the wave data was extracted from (Fig. 1). The shoreline study area consists of a 15 km stretch of sandy beach along the Inskip Peninsula (Fig. 1). The study coastline is unmodified and representative of the open drift-aligned southeast Queensland coast. It is bounded by the Great Sandy Strait to the north, a significant tidal channel which separates mainland Queensland from Fraser Island, and the Double Island Point headland to the south. Tides are semidiurnal with a mean spring tidal range of 1.40 m (at Rainbow Beach) and a HAT of 2.28 m (Queensland Government, 2017a). The east Australian longshore drift system carries approximately 500,000 m³ of sand per year from the Gold Coast north towards Fraser Island where the subaqueous Breaksea Spit represents the northern terminus (Boyd et al., 2008). The net longshore drift direction is to the north with sediment being supplied from NSW coastal catchments (Roy and Crawford, 1977; Roy and Thom, 1981). Ebb-tidal flows through the Great Sandy Strait rework and transport sediment seaward from the adjacent Hervey Bay where it is then moved offshore and northward to Fraser Island (Boyd et al., 2008). The East Australian Current flows south from the Coral Sea along the edge of the continental shelf, until it reaches central NSW (Cresswell et al., 1983; Church, 1987). The East Australian Current is located 10 km offshore near Fraser Island to the north and approximately 20–30 km offshore of the Inskip Peninsular (Boyd et al., 2008).