Shoreline and beach volume change between 1967 and 2007 at Raine Island, Great Barrier Reef, Australia

Shoreline and beach volume change between 1967 and 2007 at Raine Island, Great Barrier Reef, Australia

Global and Planetary Change 72 (2010) 141–154 Contents lists available at ScienceDirect Global and Planetary Change j o u r n a l h o m e p a g e : ...

2MB Sizes 0 Downloads 47 Views

Global and Planetary Change 72 (2010) 141–154

Contents lists available at ScienceDirect

Global and Planetary Change j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / g l o p l a c h a

Shoreline and beach volume change between 1967 and 2007 at Raine Island, Great Barrier Reef, Australia John L. Dawson ⁎, Scott G. Smithers School of Earth and Environmental Sciences, James Cook University, Townsville, Queensland, 4811, Australia

a r t i c l e

i n f o

Article history: Received 11 June 2009 Accepted 22 January 2010 Available online 1 February 2010 Keywords: sand cay erosion sediment transport GIS Great Barrier Reef Raine Island

a b s t r a c t Raine Island is a vegetated coral cay located on the far northern outer Great Barrier Reef (GBR), recognised as a globally significant turtle rookery. Cay geomorphology, specifically the morphology of the beach and swale, dictate the availability of nesting sites and influence nesting success. Understanding short and long-term shoreline change is critical for managers charged with protecting the nesting habitat, particularly as climate change progresses. Historical topographic surveys, a simple numerical model and geographic information system (GIS) techniques were used to reconstruct a 40-year (1967–2007) shoreline history of Raine Island. Results show that significant shoreline change has occurred on 78% of the island's shoreline between 1967 and 2007; 34% experienced net retreat and 44% net progradation during the study interval. Shoreline retreat is mainly concentrated on the east–southeast section of the shoreline (average annual rate of − 0.3 ± 0.3 m/yr), while the shore on the western side of the island prograded at a similar rate (0.4 ± 0.2 m/yr). A seasonal signal was detected relating to oscillations in wind direction and intensity, with the southeast and west–southwest shorelines migrating an average of ∼ 17 m from season to season. The volume of sediment deposited on Raine Island between 1967 and 2007 increased by ∼ 68,000 m3 net, but accretion rates varied significantly seasonally and from year to year. The largest volumetric changes have typically occurred over the last 23 years (1984–2007). Despite the recent concern that Raine Island is rapidly eroding, our data demonstrate net island growth (6% area, 4% volume) between 1967 and 2007. Perceptions of erosion probably reflect large morphological changes arising from seasonal, inter-annual and inter-decadal patterns of sediment redistribution rather than net loss from the island's sediment budget. Crown Copyright © 2010 Published by Elsevier B.V. All rights reserved.

1. Introduction It is projected that global sea level will rise ∼0.03 m in the next decade and between 0.18 and 0.79 m by 2100, and that tropical cyclones will become more intense (IPCC, 2007; Church et al., 2008). Low-lying reef islands are widely perceived to be particularly sensitive to these changes (McLean and Tsyban, 2001; Mimura et al., 2007; Woodroffe, 2007), although a number of geomorphologists have argued that rising sea levels do not always cause reef islands to erode. For example, a rise in sea level may promote reef island growth by: i) increasing accommodation space for new sediment; ii) reinvigorating carbonate production on reef flats where further reef growth has been inhibited by a stable sea level; and iii) increasing the efficiency of waves to transport new and stored sediment to an island depocentre (Hopley, 1993; Hopley et al., 2007; Smithers et al., 2007; Woodroffe, 2007). However, many calcifying organisms that produce sediments eventually deposited on reef islands survive within narrow environmental limits, and global ⁎ Corresponding author. Tel.: +61 7 4781 6932; fax: +61 7 4781 4020. E-mail addresses: [email protected] (J.L. Dawson), [email protected] (S.G. Smithers).

climate changes are likely to raise mortality rates (Baker et al., 2008; Veron et al., 2009). At least initially, this may promote island accretion by raising the production of detrital sediments and increasing sediment availability (Smithers et al., 2007), but if community recovery is slow and the frequency of mortality events remains high, reef island sediment budgets will eventually go into deficit. Furthermore, significant and sustained reductions in coral growth may increase the wave energy reaching island shorelines as frictional wave attenuation is reduced, potentially changing shoreline dynamics (Sheppard et al., 2005). Coral cays are low-lying reef islands formed from sediments derived from the reef on which they sit and swept by refracted waves to a focal point on the reef flat where they are deposited. Seasonal fluctuations in the intensity and dominant direction of wind and wind-driven waves/currents can produce significant changes in island size, shape and position on a reef platform (Flood, 1986; Aston, 1995; Kench and Brander, 2006a). Higher-energy waves and currents produced by tropical cyclones may markedly modify reef island morphology over shorter periods (Scoffin, 1993); cyclones affect much of the Great Barrier Reef (GBR) about once every two and a half years on average (Australian Bureau of Meteorology, 2009). Cyclones can produce major depositional features such as beach ridges

0921-8181/$ – see front matter. Crown Copyright © 2010 Published by Elsevier B.V. All rights reserved. doi:10.1016/j.gloplacha.2010.01.026

142

J.L. Dawson, S.G. Smithers / Global and Planetary Change 72 (2010) 141–154

(Maragos et al., 1973; Chivas et al., 1986; Hayne and Chappell, 2005). Alternatively, cyclonic waves and surge can severely erode shorelines and even completely remove islands (Nott, 2006). Raine Island is a low reef island located on the outer edge of the far northern GBR, Australia. It is one of the world's most important nesting sites for marine turtles (Limpus et al., 2003). The availability of nesting sites is determined by cay geomorphology, specifically the morphology of the beach and adjacent swale. Understanding short and long-term shoreline change is critical for managers charged with protecting the nesting habitat, particularly as climate change progresses. Raine Island is also geomorphically significant in that it is characterised by geomorphic features that, although not particularly unique in themselves, are rarely found in combination on reef islands (e.g. a phosphate rock cap and intertidal beachrock pavements — see Baker et al., 1998). Previous geomorphic investigations at Raine Island have focused on wave dynamics and lateral shoreline change (Gourlay and Hacker, 1991), phosphate rock formation (Baker et al., 1998), geomorphological description (Jukes, 1847; Stoddart et al., 1981), and the implications of geomorphology on green turtle nesting (Neil et al., 2000; Limpus et al., 2003). Changes in island size and volume have been estimated using photogrammetry in several studies (Gourlay and Hacker, 1991; Gourlay, 1997), but these analyses focus on a relatively short time period (1984–1990) for which quality imagery is available. This study extends the timescale of investigation to reconstruct 40 years of change (1967–2007) using both historical topographic surveys and additional field data collection. Our results improve the understanding of shoreline dynamics and reef island geomorphology at Raine Island and provide important data on shoreline variability necessary to assess the potential impacts of projected environmental changes.

2. Island setting Raine Island (11°35′28″S 144°02′17″E) is located at the northwest end of a planar reef on the outer edge of the Great Barrier Reef (GBR) (Fig. 1). The reef crest is exposed to high energy Pacific Ocean swells generated by both the southeast tradewinds that prevail during the austral winter (April–October,) and more rarely by episodic cyclones during the summer monsoon (November to March). The typical diurnal mesotidal range is about 1.8 m and much of the reef-flat is ∼0.5 m above lowest astronomical tide (LAT), corresponding with the mean low water level. Highest Astronomical Tide (HAT) is 2.75 m above LAT. The island is approximately 820 m long, 440 m wide and 27.5 ha in area. Maximum elevation is ∼ 8 m above LAT (Fig. 1). The island's long axis is aligned parallel to the southeast tradewinds, as is common for cays of the GBR (Flood, 1986; Frank and Jell, 2006). Beach width ranges from 10 to 25 m and most beaches have relatively steeply sloping beach faces (7–8°) that rise to a berm crest at ∼4 m LAT (Fig. 1). Landward of the berm crest is a wide sandy berm and swale zone (hereafter collectively referred to as the swale) that terminates to landward at a prominent 0.5–1.5 m high phosphate rock cliff (Figs. 1, 2A, C). The swale is partially vegetated by a sparse cover of Lepturus sp. (thintail grass), which is seasonal but is also disturbed during turtle nesting (November–February). Beachrock outcrops were exposed on the eastern and northeastern shorelines during a field visit in November 2007 (Fig. 1). These outcrops cover ∼ 6800 m2, and include clear bedding structures that dip seawards at a similar angle to the present day beach (Fig. 2B). Phosphates leached from avian guano have reprecipitated and cemented island sediments

Fig. 1. Location map of Raine Island on the outer northern Great Barrier Reef, Australia. A cross section (A–B) is also included, illustrating beach and central island morphology. Small letters a–d indicate the locations of photographs illustrated in Fig. 2. Note the island's northwest location on the reef-flat as well as the general beach and central island geomorphology. The swale region is defined as the portion of beach between the base of the cliff and the elevated berm crest. MHW = mean high water level (1.8 m above LAT); LAT = lowest astronomical tide; HAT = highest astronomical tide (2.75 m above LAT).

J.L. Dawson, S.G. Smithers / Global and Planetary Change 72 (2010) 141–154

143

Fig. 2. Beach features of Raine Island, northern Great Barrier Reef, including: (A) phosphate rock cliff (left of photo), seasonally vegetated swale zone and elevated sandy berm (right of photo); (B) exposed beachrock to the NE sloping at a similar angle to the beach; (C) the phosphate cap/cliff showing underlying cavernous structure. Photograph (D) is taken from the top of the navigation tower in December 2006 by the Environmental Protection Agency (EPA) and illustrates the large numbers of green turtles (Chelonia mydas) coming onto the beach during nesting season. Approximate scales of foregrounds are included. Locations of photographs are indicated on Fig. 1.

(Baker et al., 1998), outcropping on the centre of the island and forming the phosphate rock cliff at the back of the swale. A cavernous region with a complex internal structure can be observed below the cliff surface documenting the downward leaching of the cements (Fig. 2C), clearly differentiating this material from the intertidal beachrock. Much of the island interior is flat, sparsely vegetated and capped with a relatively thin (0.1–0.2 m) phosphate rock layer. Beach sediments at Raine Island are well-sorted carbonate sands (0.5–1 mm), dominated by foraminiferal tests (43%), predominantly

from Baculogypsina (30%), Amphistegina (5%), and Marginopora (4%). Fragments of molluscs (25%), coralline algae (11%), coral (10%) and Halimeda (7%) are also present (Fig. 3). With the exception of a few bone fragments, egg shell, wood and pumice, non-reefal material is absent. Relatively little sand is stored over most of the reef-flat, with only thin (b5–10 cm thick) and sporadic patches of mostly well sorted medium sand (0.25–0.5 mm) occurring, mainly in topographic depressions over the reef flat. However, at various times, significant sand deposits occur closer to the cay as sediments are periodically

Fig. 3. Average composition of sediments from A) the beach and B) the adjacent reef-flat of Raine Island. The four dominant components are foraminifera, molluscs, coral, and coralline algae (predominantly red algae).

144

J.L. Dawson, S.G. Smithers / Global and Planetary Change 72 (2010) 141–154

transported on and offshore. Coral rubble is more common, particularly towards the reef edge. Reef sediments are dominated by Baculogypsina (22%), of which the majority of tests have fully intact radiating spines — indicative of recent production. Coral (21%) and coralline algal (17%) clasts are also abundant but concentrations tend to diminish towards the island, where molluscan shell fragments constitute 19% of samples on average. Similarities between components found in both beach and reef-flat sediments suggest sediment exchange between the two environments (Fig. 3; Calhoun and Field, 2008). 3. Materials and methods 3.1. Beach profiles (1967–2007) A total of 3 historic survey maps and 5 topographic survey datasets surveyed by earlier researchers were analysed and supplemented with digital elevation data collected in 1998, 2006 and 2007, to reconstruct a shoreline history for Raine Island (Table 1). Eight common profile locations (profiles 1–8) were selected on the basis of where the most data from the previous surveys were available (i.e. where the previous surveys occurred on common transects). This gave a total of 74 beach profiles (24 derived from digital elevation data between 1998 and 2007). Beach parameters such as berm elevation and width of the swale were carefully measured directly off historical survey maps and cross-section profiles from 1967 to 1987 (Pritchard, 1967; Stoddart et al., 1981; Gourlay and Hacker, 1991; Table 1). Berm heights could not be determined from the planimetric survey maps of 1967 and 1973. These heights were predicted using best-fit regression models (nonlinear regression models where a straight line could not be fitted) of nearby berm elevations (i.e. where no height for the transect was available it was predicted by using a regression that included the nearest appropriate data points). R-squared values ranged from 0.66 to 0.92 at a confidence level of N80%. We are confident that the assigned berm heights are realistic. Elevation data were collected at 10,000 points across Raine Island during the summers of 1998 using a digital theodolite, and 2006 and 2007 using a Sokkia GSR 2700 ISX real-time kinematic global positioning system (RTK-GPS). Vertical and horizontal accuracies of both RTK GPS and digital theodolites are generally less than ±5 cm

and ±2 cm respectively (Table 1). Elevations were determined along a total of 100 shore-normal beach transects, the beach berm crest, vegetation line, cliff top, beach toe, and across the island interior and all reduced to a common LAT datum. Using ESRI ArcInfo 9.3, each point elevation dataset was georeferenced using N5 ground control points (root mean square error = 0.12 m), and used to create Triangulated Irregular Network (TIN) models (Longley and Batty, 2003). TINs were selected over interpolation methods to constrain elevations at each data point and to avoid difficulties with addressing uncertainties that are common with interpolation models (Zhang and Goodchild, 2002). TINS also handle the irregular distribution of sampling points more effectively and have a comparatively low computational cost (Leon and Woodroffe, 2007). The accuracy of a TIN model is governed by the quality of the source data (Leon and Woodroffe, 2007) which for Raine Island is high (point accuracy (b±0.05 m; Table 1), density (N15,000 points/km2), relatively even distribution of data points). TIN model accuracies were validated against the original datasets resulting in an average root mean square (RMS) error of 0.27 m. The locations of profiles 1–8 were identified in the GIS by digitising and georeferencing a planimetric map of historic profile positions from the report of Gourlay and Hacker (1991). Berm elevations and swale widths for the years 1998, 2006, and 2007 were then measured at profiles 1–8 using each underlying TIN model within the GIS (Fig. 4). 3.2. Historic shoreline change (1967–2007) Annual rates of change in swale width were calculated for the intervals between data collection using least-squares linear regression (Norcross et al., 2002). This method was chosen over more simplified methods because: 1) all the data are used, regardless of changes in trend or accuracy; 2) it is a method based on accepted statistical concepts; and 3) basic statistical software can be used. Some error is inevitable in measurements derived from historic surveys; however, it is assumed that the uncertainties associated with each dataset are independent, uncorrelated and random. Therefore errors associated with annual rates of change in swale width are included in the uncertainty term (standard error estimate) of the linear regression models (Norcross et al., 2002). Rates of change in swale width at Raine Island have an average uncertainty of ±0.23 m.

Table 1 Summary of historic topographic survey datasets — Raine Island 1967–2007. Number of usable profiles

Reference

8

Gourlay and Hacker (1991)

– b±0.05 m

8 4

Stoddart et al. (1981) Gourlay and Hacker (1991)

b±5 m

b±0.05 m

7

Gourlay and Hacker (1991)

Optical theodolite

b±5 m

b±0.05 m

3

Gourlay and Hacker (1991)

±10 m

Gourlay and Hacker (1991)

b±5 m

? (Probably b ±0.5 m) b±0.05 m

7

December 1984

Orthophomap (survey data+ aerial photograph) Optical theodolite

8

Gourlay and Hacker (1991)

August 1987

Optical theodolite

b±5 m

b±0.05 m

5

Gourlay and Hacker (1991)

December 1998

Digital theodolite

b±0.05 m

b± 0.05 m

8

Raw data processed in this study

December 2006

Real-time kinematic GPS

b± 0.05 m

b± 0.05 m

8

Raw data processed in this study

December 2007

Real-time kinematic GPS and total station survey system (TSSS)

±0.02 m

±0.01 m

8

Raw data processed in this study

Original source

Month/year surveyed

Survey method

Horizontal uncertainty

Pritchard, 1967 (planimetric survey map)

July 1967

Plan and cross section profiles

Stoddart et al., 1981 (planimetric survey map) QNPWS (Environmental Protection Agency) (8 beach profiles) QNPWS (Environmental Protection Agency) (18 beach profiles) QNPWS (Environmental Protection Agency) (9 beach profiles) Australian Survey Office (AUSLIG) (contour survey map) QNPWS (now Environmental Protection Agency) (23 beach profiles) J. Hacker (University of Queensland) (5 beach profiles) Environmental Protection Agency (EPA) (2400 point elevations) Environmental Protection Agency (EPA) (5000 point elevations) This study (2600 point elevations)

November 1973 June 1981

Pacing and compass survey Optical theodolite

? (Probably N ±10 m) ±10 m b±5 m

December 1981

Optical theodolite

July 1982 September 1983

Vertical uncertainty –

J.L. Dawson, S.G. Smithers / Global and Planetary Change 72 (2010) 141–154

145

Fig. 4. (A) A Triangulated Irregular Network (TIN) elevation model of Raine Island derived from point elevation data and showing the locations of beach profiles used in this study. The location of each profile on the TIN was spatially matched with those from historic data by digitising, importing and georeferencing a planimetric map from Gourlay and Hacker (1991) into a GIS. Examples are given of (A) a cross-section profile derived from 1998, 2006 and 2007 TIN elevation models and (B) the corresponding cross-section profile from the historic surveys of Gourlay and Hacker (1991). Dates and arrows indicate the relative positions of the berm crest.

3.3. Historic volume change (1967–2007) At each profile, the change in swale width was used to calculate volume change using the equation, modified from Hanson (1989), shown below. ΔV = ΔSðBi  ΔBÞ where total volume change (ΔV) between any two consecutive beach profiles (i and ii) is calculated as a function of the change in swale width, ΔS (m) and the height of the active beach, (Bi + ΔB) during positive ΔS and (Bi − ΔB) during negative ΔS. The parameter Bi (m) is equivalent to the initial berm elevation of profile i and (ΔB) (m) is the change in elevation between Bi and Bii (Fig. 5A). Reef flat elevation varies between 0.3 and 0.6 m above LAT. Assuming independent errors in measurements for each of the above parameters the uncertainties associated with estimates of volume change at each profile are calculated from the standard error of each (Warrick et al., 2009). The average uncertainties of each parameter are given in Table 2. To spatially assess volume change over the 40-year period, the beach at Raine Island was divided into 8 compartments that extend from each profile location half way to the next profile. Volume change at each profile is extrapolated along shore to estimate the volume change within each compartment. To account for variations in shoreline translation either side of a given profile within a given compartment, the planform areas of lateral progradation and retreat were each measured in the GIS and multiplied by the height of the at-a-profile volume change, H ± ΔB (Fig. 5B). The sum of these two volume changes gives the best-estimate of net volume change for a given compartment over a given time period. Change in beach volume, over a given time period, can also be calculated by determining the

cross-sectional area between sequential beach profiles. This is the preferred technique given it does not require the assumptions of an equilibrium beach profile (Ravens and Sitanggang, 2007). However, it was not possible to use this method for all historic survey datasets. Nevertheless, the cross-sectional area of all beach profiles from 1998 to 2007 were determined from the TIN models within the GIS and were found to be consistent with volume changes based on methods employed in this study (Fig. 6), suggesting the approach used here to calculate beach volume and changes is sound.

3.4. Cut/fill volumetric analysis (1998–2007) TIN models of Raine Island (1998, 2006, and 2007) were interpolated to digital terrain models (DTMs) using ordinary kriging and a grid comparison (cut/fill) analysis was applied to investigate high-resolution volume change between 1998 and 2007. The advantage of a ‘cut and fill’ analysis is that it directly compares the volume of two surface models and is able to display areas and volumes of surface material that have been modified by removal or addition over time. In each DTM, the beach and swale were isolated from the island interior and anomalous negative values were reduced to zero in the subtractions matrix in order to compensate for error commonly introduced during the interpolation method (Alexander et al., 2008). A limitation of the cut and fill analysis is that few elevation data below 1.8 m above LAT are available due to logistical difficulties during data collection, mostly due to the tide. Nonetheless, given that the beachface slopes are relatively uniform both through time at a transect and also around the island, the analysis is useful to investigate volume change of the beach at Raine Island, particularly within the swale, landward of the berm crest where most of the nesting occurs.

146

J.L. Dawson, S.G. Smithers / Global and Planetary Change 72 (2010) 141–154

Fig. 5. A schematic diagram showing (A) horizontal change in swale width (ΔS) and vertical change in berm elevation (ΔB) between two consecutive beach profiles (i and ii). The area between profiles i and ii (volume change per unit length of beach) is estimated as the product of the change in swale width, ΔS (m) and the height of the active beach, (Bi + ΔB) during positive ΔS and (Bi − ΔB) during negative ΔS, where H (m) is the berm elevation of profile i above the reef-flat and (ΔB) (m) is the change in berm elevation over time. Volume change within a given beach compartment (n) is calculated by multiplying the at-a profile rate (A) by the planform areas of horizontal erosion and accretion (B) to give gross accretion/erosion in m3. The sum of gross changes gives the net volume change within compartment n. Note the reef-flat elevation varies from ∼0.3 to 0.6 m LAT.

4. Results 4.1. Historic shoreline change (1967–2007) Between 1967 and 2007, 34% of Raine Island's shoreline experienced net retreat and 44% net progradation. The greatest variations in

Table 2 Parameters used to calculate volume change (m3/m) between two consecutive beach profiles. Parameter

Unit of measure

Notation

Profile i (oldest) Profile ii (youngest) Time period (profile ii–profile i) Swale width Berm elevation Berm elevation of profile i Change in berm elevation (B) between profiles i and ii Height of the active beach Change in swale width (W) between profiles i and ii Volume change between profiles i and ii

yra yra yrsa m m m m

i ii ii–i W B Bi ΔB

– – – 3.0 0.1 0.1 0.2

m m

Bi ± ΔB ΔS

0.2 4.6

a b

3

m /m

Average uncertainty (1SE)

b

ΔS(Bi ± ΔB)

17.2

Expressed as number of years after July 1967. Bi + ΔB is used for +ve shoreline change and Bi − ΔB is used for −ve shoreline change.

Fig. 6. A comparison of Raine Island volume change (m3/m of shoreline) calculated from shoreline change (swale width) using ΔS(Bi ± ΔB), and the cross-sectional area between profiles derived from digital TIN models. Each square represents a comparison at a particular profile location and a particular time period (i.e. 8 locations for both 1998–2006 and 2006–2007). The solid line represents equivalence between the two volume calculation techniques. The slope and R-square value of the correlation are given.

J.L. Dawson, S.G. Smithers / Global and Planetary Change 72 (2010) 141–154

swale width occurred at the western and southeastern ends of the island; the width of the swales at profiles 6 (west) and 3 (southeast) were as much as 43 and 47 m different at some time within the

147

40 year study period (see Fig. 7A; Table 3). Swale widths on the north–northeast to east–northeast shores (profiles 1 and 2) varied least, with maximum ranges of 11.9 and 20 m respectively (Table 3;

Fig. 7. Historic changes in swale width and volume at Raine Island (1967–2007). (A) superimposed maps showing the location of berm crests relative to the phosphate cliff. Black lines indicate surveys conducted during the northwesterly monsoon winds and grey lines indicate surveys conducted during the southeasterly tradewinds. Arrows indicate the direction and relative magnitudes of the movement of the berm crest over the 40 years; large arrows= N0.3 m/yr; medium-sized arrows=0.1–0.3 m/yr; and small arrows=b 0.1 m/yr. Data are compiled from GPS survey results and maps of Pritchard (1967); Stoddart et al. (1981) and Gourlay and Hacker (1991). Beach profile locations are given. (B) a map of the spatial variability of volume change at Raine Island between 1967 and 2007. Dashed lines indicate beach profile locations and the boxes represent corresponding beach compartments. Compartments extend alongshore either side of the profiles and half way to the adjacent profiles. Shades of grey indicate levels of volume change. Dominant wind directions during SE-Trades and NW-monsoons are also given.

148

J.L. Dawson, S.G. Smithers / Global and Planetary Change 72 (2010) 141–154

Table 3 Summary of change in swale width (m) 1967–2007 (width relative to 2007). Profiles Date

1

2

3

4

5

6

7

8

Min.

Max.

July 1967 November 1973 June 1981 December 1981 July 1982 September 1983 December 1984 August 1987 December 1998 December 2006 December 2007 Range Jul-67–Dec-07 (m/y)a

− 10.6 − 10.9 – 1.0 – – − 6.8 − 10.7 − 4.5 − 0.9 0.0 11.9 0.28

− 0.6 − 4.6 – – – – 13.0 – 1.1 − 7.0 0.0 20.0 − 0.05

2.9 − 5.4 7.3 3.8 32.6 15.6 2.6 – − 1.6 − 13.9 0.0 46.5 − 0.33

− 3.3 − 10.3 – − 1.3 – 4.0 − 6.3 1.1 12.6 − 2.1 0.0 22.9 0.24

3.6 − 10.2 – − 8.2 – − 6.9 − 5.9 – 20.5 − 1.3 0.0 30.7 0.23

− 29.1 14.2 − 27.2 − 20.7 − 23.1 − 26.0 −5.0 − 17.8 − 13.8 − 21.7 0.0 43.3 0.22

− 13.7 − 2.3 − 27.4 − 16.3 − 29.2 − 14.5 − 9.7 − 3.4 − 17.3 − 11.3 0.0 29.2 0.24

− 21.4 − 17.7 − 26.3 − 34.3 − 27.8 – − 14.9 − 23.6 − 19.8 − 13.9 0.0 34.3 0.40

− 29.1 − 17.7 − 27.4 − 34.3 − 29.2 − 26.0 − 14.9 − 23.6 − 19.8 − 21.7 0.0 11.9 − 0.33

3.6 14.2 7.3 3.8 32.6 15.6 13.0 1.1 20.5 − 0.9 0.0 46.5 0.4

a

Rate of change (metres per year) is derived from least squares linear regression.

Fig. 7A). Elsewhere, the change in swale widths fluctuated by a maximum of around 20 to 35 m between 1967 and 2007 (Table 3). The overall pattern expressed in the data indicates a trend of retreat of the east–southeast shore (between profiles 2 and 3) and progradation of the north–northwest swale between profiles 1 and 7 (Fig. 7A). For example, the swale on the north–northwest shore (profile 8) prograded 34 m between December 1981 and 2007 whereas on the opposite side of the island at the east–southeast shore (profile 3) swale width had reduced by 33 m between July 1982 and December 2007. This trend is also evident in the long-term average rates of swale width change derived from linear regression models (Table 3). The swale on the east–southeast shore tends to be narrower at the end of the southeast tradewinds/beginning of the northwest monsoon (November–December; see Table 3). In contrast, the swale at this location is widest at the end of the northwesterly monsoons, suggesting intra-annual oscillation of shoreline accretion and erosion driven by seasonal variations in wind direction and intensity.

(west–southwest) and between 1981 and 1982 at profile 3 (southeast). The largest volumes of sediment were eroded from the profiles during the period between 1973 and 1981, where as much as 150 m3 was lost from profile 6 (west–southwest) and a total of 214 m3 was lost from all measured profiles, representing about 7% of the total island volume in 2007 (Table 4). However, neither erosion nor accretion typically occurs at all profiles during any measurement interval. Usually some profiles are eroding and some accreting, again suggesting alongshore sediment redistribution as an important process. An exception to these general trends occurred during the last measurement period, when accretion occurred at all profiles — as discussed later we infer that this reflects the rapid deposition of sediment onshore from the inner reef flat during a cyclone that occurred two weeks prior to the survey. The 8 compartments delineated to spatially assess volume change over the 40-year period at Raine Island varied in area from ∼11,000– 26,000 m2 (Fig. 7B). Between 1967 and 2007, the west–southwest shore (compartment 6) was the most dynamic, showing both the largest volume of accretion (∼52,800 m3) and erosion (∼40,900 m3) for any of the measurement periods, but a net accretion of 11,900 m3 over the 40-year period. In contrast, the northeast to southeast sector (compartments 2 and 3) changed relatively little (net accretion of 2500 m3 and 3800 m3 respectively; Table 5). Net accretion was highest for compartments 1, 6, and 8, representing about 40% of the total beach area (Fig. 7B) and approximately 60% of the total net volume change (Table 5). For the entire 40-year period, the island as a whole showed a net accretion of 68,400 ± 6700 m3, equivalent to an average annual rate of approximately 1700 m3/yr (Table 5).

4.2. Historic volume change (1967–2007) At Raine Island, volume increases of up to 160.8 m3/m (volume per linear metre of shoreline) and decreases of −150.1 m3/m have occurred along the eight historic profiles during the past 40 years (Table 4). Net accretion occurred along all profiles between 1967 and 2007, ranging from just 6.8 m3/m at profile 5 to 138.9 m3/m at profile 6. The largest accretion volumes for any measurement period and profile occurred between 1967 and 1973 at profile 6 Table 4 Summary of historic volume change (m3/m) 1967–2007. Profiles Time period

1

2

3

4

5

6

7

8

Min.

Max.

Jul67–Nov73 Nov73–Jun81 Nov73–Dec81 Nov73–Dec84 Jun81–Dec81 Dec81–Jul82 Dec81–Sep83 Dec81–Dec84 Jul82–Sep83 Sep83–Dec84 Dec84–Aug87 Dec84–Dec98 Aug87–Dec98 Dec98–Dec06 Dec06–Dec07 Jul67–Dec07

− 1.3 – 52.4 – – – – − 35.1 – – − 15.6 – 20.5 15.5 18.5 54.8

− 15.3 – – 66.9 – – – – – – – − 41.7 – − 24.3 31.9 17.5

− 30.6 49.5 – – − 13.7 118.1 – – − 68.0 − 50.7 – − 13.0 – − 27.1 54.3 18.9

− 28.1 – 41.4 – – – 23.9 – – − 42.2 28.1 – 43.7 − 51.5 25.6 40.9

− 57.6 – 9.0 – – – 5.7 – – 4.4 − 100.3 – − 67.6 12.9 6.8

160.8 − 150.1 – – 28.6 − 10.3 – – − 12.5 73.5 − 41.0 – 16.8 − 26.9 99.9 138.9

41.0 − 90.1 – – 38.9 − 40.0 – – 58.8 15.8 19.5 – − 30.6 20.4 51.4 85.2

12.5 − 23.6 – – − 24.0 28.0 – – – – − 33.9 – 15.2 24.8 47.0 45.9

− 57.6 − 150.1 9.0 66.9 − 24.0 − 40.0 5.7 − 35.1 −68.0 − 50.7 − 41.0 − 41.7 − 30.6 − 67.6 12.9 6.8

160.8 49.5 52.4 66.9 38.9 118.1 23.9 − 35.1 58.8 73.5 28.1 100.3 43.7 24.8 99.9 138.9

J.L. Dawson, S.G. Smithers / Global and Planetary Change 72 (2010) 141–154

149

Table 5 Summary of changes in island beach volume (m3) 1967–2007 based on profile data. Volume change

Beach compartment

Jul67–Nov73

Nov73–Dec84

Dec84–Dec98

Dec98–Dec06

Dec06–Dec07

Jul67–Dec07

Gross accretion

1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8

4100 200 0 0 300 21,100 7300 1100 − 1500 − 4000 − 5300 − 6100 − 10,000 − 1400 0 − 2200 2600 − 3800 − 5300 − 6100 − 9700 19,700 7300 − 1100 3600

8200 15,200 11,600 18,000 5100 12,400 5900 10,500 − 8900 − 4000 − 4800 − 14,800 − 2900 − 28,000 − 11,100 −13,900 − 700 11,200 6800 3200 2200 − 15,600 − 5200 − 3400 − 1500

9300 100 2700 18,400 25,500 4300 0 5800 0 − 4800 − 4200 0 0 − 4800 − 2900 0 9300 − 4700 − 1500 18,400 25,500 − 500 − 2900 5800 49,400

0 300 400 0 0 800 1900 2500 − 6100 − 2800 − 4100 − 14,800 − 17,500 − 6700 0 − 300 − 6100 − 2500 − 3700 − 14,800 − 17,500 − 5900 1900 2200 − 46,400

8300 3400 7700 7400 5200 14,200 7000 11,400 0 − 1100 − 200 0 0 0 0 0 8300 2300 7500 7400 5200 14,200 7000 11,400 63,300

29,900 19,200 22,400 43,800 36,100 52,800 22,100 31,300 − 16,500 − 16,700 − 18,600 − 35,700 − 30,400 − 40,900 − 14,000 − 16,400 13,400 2500 3800 8100 5700 11,900 8100 14,900 68,400

Gross erosion

Net volume change

Total

There has been a general alternation of erosional and accretionary phases for the island as a whole over time; total island volume changes were small in the first 17 years of study (+3600 m3 July 1967–November 1973; −1500 m3 November 1973–December 1984) but have dramatically increased since 1984 (Table 5). In the 14 years between December 1984 and 1998 net accretion of 49,400 m3 occurred (annual average of ∼ 3500 m3/yr), and in the 8 years between December 1998 and 2006 a net loss of 46,400 m3 occurred (annual average erosion of 5800 m3/yr). Measurements over the last survey period (December 2006–2007) indicate very large accretion of 63,300 m3, possibly reflecting the fact that the last survey occurred just two weeks following a cyclone (Guba), and after a prolonged period of erosion when some of the sediments removed from the island may have been stored on the nearshore reef flat. This remarkable annual volume change represents approximately 4% of the current island volume and 10% of the current beach volume, but it is unclear how much of this sediment will be stored onshore in the longer term.

historic profile data indicate the greatest erosion occurred along the southern and south–southwestern shore (compartments 4 and 5; see Table 5). The analyses capture different components of sediment dynamics on the cay, with reduced sediment volume at compartment 6 quantified by the cut and fill analysis attributed to reduced swale elevation over the broader compartment area. In contrast, the shoreline at profile 6 shows a net progradation of 14 m from 1998 to 2007, however, the shoreline within the compartment overall shows a variable pattern, retreating to the southeast and prograding to the north of profile 6, for relatively little net change at the compartment scale (Fig. 7). On the southern and south–southwestern shore (compartments 4 and 5) the swale elevations have remained constant or modestly increased, but marked net shoreline retreat post 1998 (10–20 m) has occurred. We note that the erosion trend for 1998–2007 indicated for compartment 6 differs from the net accretion established for the entire 40-year period, demonstrating the temporal variability in shoreline and beach morphology around the island. 5. Discussion

4.3. Volume change of the swale between 1998 and 2007 The cut/fill analysis indicates that between 1998 and 2007 net accretion (sediment fill) occurred over 72% of the swale area, 20% experienced net erosion (sediment cut), and 8% showed no net change for the 9-year period. Erosion from the swale was mostly concentrated at the western end of the cay (Fig. 8), especially around profile and compartment 6 where the shoreline receded ∼ 8 m between 1998 and 2006 (Table 3) and the volume of the entire compartment was reduced by about 6000 m3 (Table 5). The elevation of the swale around the entire island changed little between 2006 and 2007, but over the same period accretion and progradation dominated most of the shoreline, especially around profiles/compartments 6 and 8, where the coast built out by up to 20–25 m (Fig. 7; Table 5). Comparison of the accretion and erosion patterns revealed by cut and fill analysis for 1998–2007 with the historical profile record for the same period demonstrates both the temporal and spatial variability in sediment deposition and erosion on Raine Island. For example, the cut/fill analysis shows that erosion was focussed on the western end of the island between 1998 and 2007 (Fig. 8) whereas

Detailed quantitative surveys and analyses demonstrate that Raine Island increased in area (∼ 6%) and volume (∼ 4%) between 1967 and 2007. The analyses also revealed that the island undergoes substantial morphological change from season to season and year to year. Changes are most pronounced at the southeast and western ends of the cay, where large shoreline changes probably reflect seasonal shifts in dominant wind direction, as described for other cays on the GBR (Flood, 1984, 1986, 1988) and elsewhere (Kench and Brander, 2006a). 5.1. Historic shoreline behaviour 5.1.1. Lateral shoreline change At Raine Island, the phosphate cap and vegetation of the island interior impart a degree of stability. However, the beach and swale seaward of the phosphate cliff are dynamic and changed morphologically between 1967 and 2007 (Figs. 7A and 8). The shoreline on the southeastern flanks of the island showed a recessionary trend between 1967 and 2007, with an average annual rate of retreat of −0.33 ± 0.32 m/yr whereas the north–northwestern shore prograded

150

J.L. Dawson, S.G. Smithers / Global and Planetary Change 72 (2010) 141–154

Fig. 8. Cut and fill analysis of the subaerial beach of Raine Island (1998–2007) using georeferenced digital elevation models (DEM's). The subaerial beach is defined as the portion of beach between the berm crest (dashed line) and the cliff. The position of the beach toe (dotted line) is also given. Calculated cut (erosion), fill (accretion), and total (net) island volume change are in m3. Planform areas (m2) and average change in beach elevation are also given for areas of cut and fill. It is assumed that the island interior is stable and has not changed over the 9-year period, therefore the Island interior has been omitted from the analysis to remove the potential for volumetric change arising from inconsistencies in topography.

over the same period at an average annual rate of 0.40 ± 0.19 m/yr. These annual average rates of change are, however, indicative only, as large seasonal and interannual variability is usual rather than gradual change. Examination of shoreline changes around the entire island for the survey period suggests a clockwise rotational trend for sediment movement and shoreline accretion, as has also been reported for other GBR cays such as Green Island (Beach Protection Authority, 1989).

from the surrounding reef flat back onto the island. Importantly, the net loss of −46 400 m3 between 1998 and 2006 may represent a temporary loss of sediment to the adjacent reef flat rather than a permanent loss, and the large accretion in 2007 may reflect the rapid return of this sediment back onto the beach during cyclone Guba. It is unknown how much of this sediment will remain onshore in the longer term.

5.1.2. Volumetric change The 40-year record for Raine Island reveals that erosion (volumetric decreases) and accretion (volumetric increases) of the beach do not occur uniformly around the island for any given time period or at any given location through time (Table 5). Between 1967 and 2007, Raine Island accreted a net volume of 68,400 ± 6700 m3 of sand, of which approximately 40% was deposited along the north–northwestern shore as the shoreline has laterally built out. The average annual rate of accretion of approximately 1700 m3/yr is low compared to the estimate of 5000 m3/yr previously calculated for the period of 1984– 1990 (Gourlay and Hacker, 1991; Gourlay, 1997), however large interannual variations clearly occur and are evident in the record since 1967 (Tables 4 and 5). For example, the volume change calculated from our dataset for between 1984 and 1998 is 3500 m3/yr. The large accretion volume measured between 2006 and 2007 (63,000 ± 1000 m3) followed a net volumetric increase of just 5100 m3 between 1967 and 2006 (Table 5). This anomalously large accretion may reflect rapid transport of sediment stored on the adjacent reef flat onshore during a cyclone that occurred just 2 weeks prior to the 2007 survey, some of which may be later removed off the beach and back onto the reef flat. This cycle may occur through time as high-energy events are responsible for quick delivery of sediment

5.2. Response to environmental driving forces 5.2.1. Seasonal wind patterns The distinct wind and wave seasons experienced at Raine Island generate alongshore sediment transport with monsoonally-driven accretion to the southeast matched by erosion to the west-southwest, with a reversing pattern occurring during the southeasterly tradewinds (Fig. 7A). Consequently, the opposing ends of the island are very dynamic and exhibit some of the greatest variability in lateral shoreline movement. Seasonally ‘full’ beach compartments at the southeast of the island at the end of the monsoon are gradually depleted during the southeasterly tradewinds, when beach width on this shore narrows. The reverse is true for the west–southwest to northern shore during the monsoons (Table 3). For example, the width of the swale at profile 3 was on average 17 m narrower at the end of the summer northwest monsoons than in June–August when the island is visited, noting that the southeasterly trade season usually has several months left to run at this time during which sediment movement to the west will continue. The reverse pattern is true for the opposing end of the island, particularly at or close to profiles 6 where the width of the swale was on average 18 m wider at the end of the southeast trade season (Fig. 7A). Seasonal oscillations similar to

J.L. Dawson, S.G. Smithers / Global and Planetary Change 72 (2010) 141–154

those observed at Raine have been reported from mid-ocean atoll reef islands in the Maldives (Kench and Brander, 2006a). Reef islands are inherently sensitive to changes in the intensity and/or the direction of wave energy on a reef platform (Gourlay, 1988). Seasonal wind reversals can result in significant changes in island size, shape and position on a reef (Hopley, 1981; Kench and Brander, 2006a). On the GBR, seasonal oscillations in the orientation of spits at opposing ends of a reef island have also been well documented (Gourlay, 1983; Flood, 1984, 1986, 1988). Changes in wind direction influence the pathways of propagating waves across the reef-flat and can reposition the wave convergence zone, where sediments are deposited as a result of attenuated wave energy (Flood, 1986; Hopley et al., 2007). On the central GBR, Flood (1986) recognised a long-term oscillation shift in mean annual wind energy vector of about 45° from south–southeast to east–southeast (1964– 1978) was the driving force for the observed shifts in orientation of a windward spit to the southeast. At Raine Island, large increases in

151

swale width on the east–southeast shore can result from the seasonal extension of the island during the northwest monsoon season of certain years. On the other hand, large decreases represent the return of this material back towards the western end of the island during the southeast tradewinds (Fig. 9). Consequently, changes in beach width on opposing ends of Raine Island can be in the order of 40–80 m over a single season (Fig. 9), while seasonal volumetric changes up to 120 m3 per linear metre of shoreline can occur (Table 4). 5.2.2. Wave modification across the reef-flat The major process mechanism controlling the stability of reef islands is wave action and interactions with the reef platform topography (Gourlay, 1988). Reef elevation and width are especially important in governing levels of geomorphic activity during a tidal cycle as they determine: 1) the proportion of time that wave energy can propagate across a reef-flat; and 2) the relative degree of bed friction imposed (Kench and Brander, 2006b). Such differences can, in

Fig. 9. Aerial photographs of Raine Island taken in (A) November 1990; and (B) June 1991 (NorthAir Surveys) illustrating the substantial shoreline changes that occurred between summer 1990 (dashed line) and winter 1991 (solid line). Note the wider beach to the west and exposure of beachrock (filled area) to the northeast in November (A) and the extension of the southeast shoreline in July (B). A rapid seasonal change in vegetation in the swale region can also be seen between the two photographs.

152

J.L. Dawson, S.G. Smithers / Global and Planetary Change 72 (2010) 141–154

part, account for variations in shoreline morphology on opposing sides of a reef island (Kench and Brander, 2006b). By measuring the reef width at each profile within the GIS and using the reef top elevations of Gourlay's (1995) model, the reef energy window index (ψ) of Kench and Brander (2006b) was calculated at each profile. Highest and lowest values for ψ occur on the west–northwest (0.05) and east–southeast (0.001) reef flat reflecting gross differences in wave energy that reach the shore. The significantly wider reef-flat to the east–southeast (2010 m versus 46 m for the reef flat to the west– southwest) will dissipate most wave energy, thus reducing the potential for waves to geomorphically influence the shore. This may, in part, explain the comparatively small volumetric increases that have occurred on the northeast to southeast sector of the island (beach compartments 2 and 3) between 1967 and 2007 (Fig. 7B; Table 5). It could also explain why the maximum variation in swale widths was observed at profiles 6 and 8 on the western side of the island, where ψ is relatively large suggesting a high potential for geomorphic work. In contrast, Gourlay (1995, 1997) contends that relative tidal levels do not affect the patterns of reef-top wavegenerated currents at Raine Island and that westward flowing wavegenerated currents transport most sediment to the island. Gourlay's model provides an explanation for the ESE–WNW shoreline migration trend illustrated over the 40-year study period (Table 3; Fig. 7A). However, it is likely that both processes are important drivers of longterm shoreline changes at Raine Island. 5.2.3. Tropical cyclones A major environmental driving force on the morphological change of reef islands on the Great Barrier Reef (GBR) is elevated storm surge/ waves associated with tropical cyclones (Blackman et al., 1986). During extreme events such as tropical storms or cyclones, storm surges may overtop the beach face causing a recession of the upper and back beach area (Nott, 2006; Masetti et al., 2008). Cyclones can severely erode reef islands in a short period of time; island sediment may be transported out onto the reef-flat awaiting subsequent replenishment back to the island or may be permanently lost off the reef edge (Bayliss-Smith, 1988; Solomon and Forbes, 1999; Smithers et al., 2007). Cyclones as far north as Raine Island are rare, and typically are fast moving and low intensity (Puotinen, 2007). Nevertheless, 14 tropical cyclones have passed within 100 km of Raine Island between 1967 and 2007, of which 5 have passed within 40 km, the average diameter of a cyclone eye and area of maximum wind strength (Australian Bureau of Meteorology, 2009; Fig. 10). The eyes of two cyclones (1970 and 2005) passed within 20 km of Raine Island between 1967 and 2007 (Fig. 10), presumably increasing wave energy across the reef and reaching the shore. However, it is not possible to confidently determine the impact of these events on the shoreline change due to the time windows captured by the historical survey data, and the likelihood of overprinting by large seasonal shoreline changes. One possible exception is during the last year of study, when tropical cyclone Guba (reaching category 3) slowly (∼5 km/h) passed ∼ 70 km to the east of Raine Island a fortnight prior to survey. The marked accretion at Raine Island between December 2006 (no cyclone) and December 2007 (cyclone) is consistent with the observations of Gourlay and Hacker (1991) who suggested that a northeasterly swell generated by either of two cyclones during 1970 and 1972 caused significant accretion of parts of the island between 1967 and 1973. 5.3. Applicability of the shoreline model for volumetric change In this study we used shoreline change as a proxy for volume change over the past 40 years. The theoretical relationship between volume change (per metre of shoreline) and shoreline change at Raine Island was given by ΔV = ΔS(Bi ± ΔB) where ΔV is volume change under a given beach profile, ΔS is the change in width of the swale

Fig. 10. Tropical cyclone tracks passing within 100 km of Raine Island between 1967 and 2007. Bold lines indicate cyclone tracks passing within 40 km (average diameter of a cyclone eye) and the months/years are given. Arrows indicate the direction of tracks. The dashed line represents the track of cyclone Guba, passing Raine Island just two weeks prior to the December 2007 survey. Data is compiled from the Australian Bureau of Meteorology, 2009 (http://australiasevereweather.com/cyclones/index.html).

region, Bi is the initial berm elevation of profile i and ΔB is the change in elevation between Bi and Bii (Fig. 5A). The application of this model is justifiable for Raine Island (Fig. 6), where beach slope varies little around the island, but careful assessment is necessary before it is applied to other reef islands. However, at Raine Island we are confident that it reasonably accommodates and represents changes in shoreline morphology and volume, as indicated by the accord between volume changes calculated from the product of lateral shoreline change, and volume changes measured between beach profiles. 6. Conclusions This study examined trends of shoreline and volumetric change at Raine Island between 1967 and 2007. Contrary to perceptions, Raine Island did not erode but instead modestly accreted during the 40-year study period (6% in area, 4% in volume). Significant lateral shifts in shoreline position were recorded around most of the island (∼ 78% of the total perimeter). A longer-term trend in which the beach rotates in a clockwise direction was identified, with east–southeast shoreline retreating by an average of 0.3 m per year whereas the north– northwest shore prograded at an average rate of 0.4 m per year over the 40-year period. A pronounced seasonal oscillation in shoreline position was evident with an amplitude of change that often far exceeds the longer-term trends (e.g. an average seasonal shift in

J.L. Dawson, S.G. Smithers / Global and Planetary Change 72 (2010) 141–154

shoreline position of 17 m is calculated, but the net change from year to year is generally 0.3–0.4 m), driven by seasonal changes in wind direction and intensity. In the 40 years between 1967 and 2007 Raine Island underwent a net accretion of 68,400 ± 6700 m3. This accretion did not occur at a steady rate, but was characterised by significant variability from year to year. Large changes in total island volume have occurred since about 1984 while volume changes as large as 4% of the total island volume can occur in a single year (e.g. December 2006–December 2007). Sand sheets on the reef flat immediately adjacent to the cay are the probable sources/sinks for episodes of rapid accretion/erosion, with exchange between the cay and reef flat indicated by the morphology and sedimentology of these deposits. The two analyses used to investigate volume change (historic beach profile data (1967–2007) and a GIS-based cut/fill analysis (1998– 2007)) capture different components of sediment dynamics on Raine Island. The techniques applied in this study establish a record of historic shoreline fluctuations and provide an understanding of how beach behaviour evolves and responds to environmental driving forces. We have been able to utilise sources of survey data other than beach profile data including 2D planimetric maps, aerial photographs and satellite imagery, thus, greatly extending the length of the historic record. The observed interactions between volume change and morphology at Raine Island are critical for improved resolution of island response to future climate change and for managers and decision makers to adopt the most suitable strategies for the protection of turtle nesting as well as historically important infrastructure (e.g. navigation tower). It is clear that future management strategies of Raine Island and other islands of the Great Barrier Reef should recognise that perceptions of reef island erosion can arise from large short-term seasonal and stormderived sediment redistribution from one part of the island to another or to a temporary storage on the adjacent reef flat, and not necessarily a net permanent loss from the island sediment budget. Acknowledgements The authors wish to thank the EPA (Environmental Protection Agency) for the supply of the 1998 and 2006 elevation datasets and for their assistance in the field. Additional gratitude goes to the captain and crew of the QPWS research vessel MV Kerra Lyn and two anonymous reviewers for providing objective and thoughtful comments and suggestions that helped improve early manuscript drafts. We also acknowledge Professor David Hopley for providing valuable information and insightful comments on early drafts. Funding for this project was provided in part by EPA Queensland and the Queensland Government through the Smart State PhD Scholarships Program. References Alexander, R.W., Calvo-Cases, A., Arnau-Rosalén, E., Mather, A.E., Lázaro-Suau, R., 2008. Erosion and stabilisation sequences in relation to base level changes in the El Cautivo badlands, SE Spain. Geomorphology 100 (1–2), 83–90. Aston, J.P., 1995. The relative mobilities of coral cays on the GBR can be modelled, M.S. thesis, James Cook University, Townsville, Australia. 267 pp. Australian Bureau of Meteorology, 2009. Tropical cyclone database for Australia. . Published online at http://australiasevereweather.com/cyclones/index.html. Date accessed: 02/09/2009. Baker, J.C., Jell, J.S., Hacker, J.L.F., Baublys, K.A., 1998. Origin of insular phosphate rock on a coral cay — Raine Island, Northern Great Barrier Reef. Australia Journal of Sedimentary Research 68 (5), 1001–1008. Baker, A.C., Glynn, P.W., Riegl, B., 2008. Climate change and coral reef bleaching: an ecological assessment of long-term impacts, recovery trends and future outlook. Estuarine, Coastal and Shelf Science 80 (4), 435–471. Bayliss-Smith, T.P., 1988. The role of hurricanes in the development of Reef Islands, Ontong Java Atoll, Solomon Islands. Geographical Journal 154 (3), 377–391. Beach Protection Authority Queensland, 1989. Green Island Data Report. Brisbane. 39 pp. Blackman, J.G., Winter, J.W., King, B.R., 1986. Effects of cyclone Winifred on coastal and island fauna. In: Dutton, I. (Ed.), Workshops on the Offshore Effects of Cyclone Winifred. Townsville. Great Barrier Reef Marine Park Authority, Townsville, pp. 60–70. Calhoun, R.S., Field, M.E., 2008. Sand composition and transport history on a fringing Coral Reef, Molokai, Hawaii. Journal of Coastal Research 24 (5), 1151–1160.

153

Chivas, A., Chappell, J., Wallensky, E., 1986. Radiocarbon evidence for the timing and rate of island development, beach rock formation and phosphatization at Lady Elliot Island, Queensland, Australia. Marine Geology 69, 273–287. Church, J.A., White, N.J., Hunter, J.R., Lambeck, K., 2008. Briefing: a post-IPCC AR4 update on sea-level rise. Antarctic Climate & Ecosystems Cooperative Research Centre, Hobart, Tasmania, pp. 1–12. Flood, P.G., 1984. Variability in shoreline position on five uninhabited islands of the Capricorn Section, GBR Marine Park. In: Ward, W.T., Saenger, P. (Eds.), The Capricornia Section of the GBR: Past. Brisbane Royal Society of Queensland and Australian Coral Reef Society, pp. 17–24. Flood, P.G., 1986. Sensitivity of coral cays to climatic variations, Southern Great Barrier Reef, Australia. Coral Reefs 5, 13–18. Flood, P.G., 1988. Shoreline changes on coral cays, Capricornia section, Great Barrier Reef Marine Park, Australia. Proceedings of the 6th International Coral Reef Symposium, Townsville, vol. 2, pp. 219–224. Frank, T.D., Jell, J.S., 2006. Recent developments on a nearshore, terrigenous-influenced reef: Low Isles Reef, Australia. Journal of Coastal Research 22 (3), 474–486. Gourlay, 1983. Accretion and erosion of coral cays and some consequent implications for the management of Marine Parks. In: Baker, J.T., Carter, R.M., Sammarco, P.W., Stark, K.P. (Eds.), Proceedings of the inaugural Great Barrier Reef Conference, August 29–September 2, 1983, Townsville. James Cook University Press, North Queensland, Australia. Gourlay, M.R., 1988. Coral cays: products of wave action and geological processes in a biogenic environment. Proceedings of the 6th International Coral Reef Symposium, Townsville, Australia, vol. 2, pp. 491–496. Gourlay, M.R., 1995. Wave generated currents at Raine Island: laboratory model experiments. University of Queensland Civil Engineering Report CH44/95. 89 pp. Gourlay, M.R., 1997. Raine Island coastal processes and geology: a review of investigations 1991–1997. Report to the Raine Island Corporation. 19 pp. Gourlay, M.R., Hacker, J.L.F., 1991. Raine Island: coastal processes and sedimentology. CH40/91, Department of Civil Engineering, University of Queensland, Brisbane. 151 pp. Hanson, H., 1989. Genesis-a generalized shoreline change numerical model. Journal of Coastal Research 5 (1), 1–27. Hayne, M., Chappell, J., 2005. Cyclone frequency at Princess Charlotte Bay, Queensland over the last 2, 500 years. Palaeogeography, Palaeoclimatology, Palaeoecology 168, 207–219. Hopley, D., 1981. Sediment movement around a coral cay, Great Barrier Reef, Australia. Pacific Geology 15, 17–36. Hopley, D., 1993. Coral reefs in a period of global sea level rise. In: Saxena, N. (Ed.), PACON international: Recent advances in marine science and technology. PACON, Honolulu, pp. 453–462. Hopley, D., Smithers, S.G., Parnell, K.E., 2007. The geomorphology of the Great Barrier Reef: development, diversity and change. Cambridge University Press, Cambridge. 532 pp. IPCC (Intergovernmental Panel on Climate Change), 2007. Climate Change 2007: The Physical Science Basis. Contribution of Working Group 1 to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change. In: Solomon, S., Qin, D., Manning, M., Chen, Z., Marquis, M.C., Averyt, K., Tignor, M., Miller, H.L. (Eds.), Intergovernmental Panel on Climate Change, Cambridge and New York. Jukes, J.B., 1847. Narrative of the Surveying Voyage of H.M.S. Fly, vol. 1. T. & W. Boone, London. 423 pp. Kench, P.S., Brander, R.W., 2006a. Response of reef island shorelines to seasonal climate oscillations: South Maalhosmadulu atoll, Maldives. Journal of Geophysical Research 111, F01001. doi:10.1029/2005JF000323. Kench, P.S., Brander, R.W., 2006b. Wave processes on coral reef flats: implications for reef geomorphology using Australian case studies. Journal of Coastal Research 22 (1), 209–223. Leon, J., Woodroffe, C.D., 2007. The use of digital terrain models in coral reef geomorphology. Proceedings of CoastGIS 07 — 8th International Symposium on GIS and Computer Mapping to Coastal Zone Management, Santander, Spain, vol. 2, pp. 454–463. Limpus, C.J., Miller, J., Parmenter, C.J., Limpus, D., 2003. The green turtle, Chelonia mydas, population of Raine island and the northern Great Barrier Reef 1843–2001. Memoirs of the Queensland Museum 49, 349–440. Longley, P.A., Batty, M., 2003. Advanced spatial analysis: the CASA book of GIS 8. ESRI, Redlands, USA. 463 pp. Maragos, J., Baines, G., Beveridge, P., 1973. Tropical cyclone Bebe creates a new land formation on Funafuti Atoll. Science 181, 1161–1164. Masetti, R., Fagherazzi, S., Montanari, A., 2008. Application of a barrier island translation model to the millennial-scale evolution of Sand Key, Florida. Continental Shelf Research 28 (9), 1116–1126. McLean, R.F., Tsyban, A., 2001. Coastal zones and marine ecosystems. In: McCarthy, J.J., Canziani, O.F., Leary, N.A., Dokken, D.J., White, K.S. (Eds.), Climate Change 2001: Impacts, Adaptation, and Vulnerability. Cambridge University Press, Cambridge, UK, pp. 343–379. Mimura, N., Nurse, L., McLean, R.F., Agard, J., Briguglio, L., Lefale, P., Payet, R., Sem, G., 2007. Small islands. Climate Change 2007: Impacts, Adaptation and Vulnerability. In: Parry, M.L., Canziani, O.F., Palutikof, J.P., van der Linden, P.J., Hanson, C.E. (Eds.), Contribution of Working Group II to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge University Press, Cambridge, UK, pp. 687–716. Neil, D.T., Jell, J.S., Brownlaw, R.L.S., Frankel, E., Phinn, S.R., 2000. Geology and geomorphology of Raine Island and its Reef, northern GBR, Australia, and implications for green turtle nesting and conservation management. Report to Raine Island Corporation. 73 pp. Norcross, Z.M., Fletcher, C.H., Merrifield, M., 2002. Annual and interannual changes on a reef-fringed pocket beach: Kailua Bay, Hawaii. Marine Geology 190 (3–4), 553–580.

154

J.L. Dawson, S.G. Smithers / Global and Planetary Change 72 (2010) 141–154

Nott, J., 2006. Tropical cyclones and the evolution of the sedimentary coast of Northern Australia. Journal of Coastal Research 22 (1), 49–62. Pritchard, P.W., 1967. Raine Island phosphate deposit — examination 12th−20th July 1967. Geological Record 1967/19. Australian Fertilizers Ltd, Sydney. Puotinen, M.L., 2007. Modelling the risk of cyclone wave damage to coral reefs using GIS: a case study of the Great Barrier Reef, 1969−2003. International Journal of Geographical Information Science 21 (1), 97–120. Ravens, T.M., Sitanggang, K.I., 2007. Numerical modeling and analysis of shoreline change on Galveston Island. Journal of Coastal Research 23 (3), 699–710. Scoffin, T., 1993. The geological effects of hurricanes on coral reefs and the interpretation of storm deposits. Coral Reefs 12, 203–221. Sheppard, C., Dixon, D.J., Gourlay, M.R., Sheppard, A., Payet, R., 2005. Coral mortality increases wave energy reaching shores protected by reef flats: examples from the Seychelles. Estuarine, Coastal and Shelf Science 64, 223–234. Smithers, S.G., Harvey, N., Hopley, D., Woodroffe, C.D., 2007. Vulnerability of geomorphological features in the Great Barrier Reef to climate change. In: Johnson, J.E., Marshall, P.A.

(Eds.), Climate Change and the Great Barrier Reef. Great Barrier Reef Marine Park Authority and Australian Greenhouse Office, Australia, pp. 667–716. Solomon, S.M., Forbes, D.L., 1999. Coastal hazards and associated management issues on South Pacific Islands. Ocean and Coastal Management 42, 523–554. Stoddart, D.R., Gibbs, P.E., Hopley, D., 1981. Natural history of Raine Island, Great Barrier Reef. Atoll Research Bulletin 254, 1–137. Veron, J.E.N., Hoegh-Guldberg, O., Lenton, T.M., Lough, J.M., Obura, D.O., Pearce-Kelly, P., Sheppard, C.R.C., Spalding, M., Stafford-Smith, M.G., Rogers, A.D., 2009. The coral reef crisis: the critical importance of b350 ppm CO2. Marine Pollution Bulletin 58 (10), 1428–1436. Warrick, J.A., George, D.A., Gelfenbaum, G., Ruggiero, P., Kaminsky, G.M., Beirne, M., 2009. Beach morphology and change along the mixed grain-size delta of the dammed Elwha River, Washington. Geomorphology 111 (3–4), 136–148. Woodroffe, C.D., 2007. Reef-island topography and the vulnerability of atolls to sealevel rise. Global and Planetary Change 62 (1–2), 77–96. Zhang, J., Goodchild, M., 2002. Uncertainty in geographical information. Taylor & Francis Inc., New York, USA. 266 pp.