Keywords

These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

5.1 Introduction and Aims

The case study presented in this chapter forms a component of continuing investigations on coastal geomorphic change experienced on Taveuni Island in northern Fiji, especially the effects of Tropical Cyclone (TC) Tomas in March 2010. The summary here focuses on the characteristics of reef-platform boulder fields and beachrock excavation in four locations. Readers are directed to Etienne and Terry (2012) for a more comprehensive account dealing with a wider range of coastal features than those described here.

TC Tomas in mid-March 2010 was a category-4 intensity cyclone that affected the northern and eastern islands of Fiji in the southwest Pacific. The eastern side of Taveuni Island in particular experienced monstrous waves and powerful storm surge and as the eye passed within 30 km of the coast. Impacts on coastal geomorphology were investigated within a few months of the cyclone on Taveuni. Clear geomorphic and sedimentary signatures were left by TC Tomas in Taveuni’s eastern coastal landscape, distinguishable as both quarrying and scouring features related to the production, removal and/or remobilisation of material, and associated depositional assemblages occurring as various types of sedimentary structures. Amongst the principal aims were to identify the geomorphic imprints resulting from this major storm event, to examine spatial variations in storm effects along the coastline, and to extract quantitative information on the grain-size distributions of coarse deposits, including fresh reef-platform coral boulders (RPCBs) and fragments of beachrock excavated from known sources. The physical characteristics of these deposits were then subject to numerical modelling in order to determine flow velocities and sediment transport capability of the storm-wave conditions prevailing during TC Tomas. This further allows comparison with similar but more substantial reef-top deposits laid down by unknown high-energy marine inundation (HEMI) events at earlier times.

5.2 Background to Study Area

Taveuni Island is an elongated volcanic island in the north of the Fiji archipelago, traversed by the 180° meridian. Approximately 42 km long and 10–14 km wide, the island is oriented in a north east—south west direction. Although Taveuni’s basaltic shield volcanoes are inactive, the topography is dominated by an impressive series volcanic peaks forming a highland spine along the centre of the island; several peaks reach elevations over 1000 m a.s.l (Fig. 5.1). Facing the moist south east trade winds means that the south east coast receives annual rainfall of approximately 5,000 mm, whilst precipitation near the summit of the interior mountains is almost double that figure, approaching 10,000 mm/year. Owing to the year-round very wet climate, the island’s rugged landscape is deeply dissected by steep river valleys and thickly vegetated with tropical rainforest (Ash 1987). Taveuni is therefore fondly called the ‘Garden Island of Fiji’.

Fig. 5.1
figure 1

Taveuni Island in Northern Fiji. Most of the coastline and coral reefs in the boxed area on the map have been established as a Marine Protected Area. The satellite image (courtesy of Google Earth, GeoEye) shows the location of coral boulder fields on fringing reefs and displaced fragments of beachrock that were examined on the central east coast between Bouma and Lavena villages

Full size image

Much of Taveuni’s coastline is pristine and undisturbed. The attractive landscapes and rich biodiversity of the island, both in terms of terrestrial and marine life, have enabled small scale ecotourism ventures to grow and become an important component of the local economy. Although coral reefs do not encircle the island entirely, they are well developed along the northern and eastern central coasts, extending as 100–500 m broad platforms from the shoreline. To safeguard these reefs and adjacent shorelines, a Marine Protected Area (MPA) was established in 1998. The MPA lies within the greater Bouma National Heritage Park (BNHP 2011). At Lavena and Bouma villages (Fig. 5.1), locally-managed visitor centres organise eco-adventure activities that attract a daily flow of international tourists vacationing at resorts elsewhere on the island. Fees payable provide a steady cash income that benefits the rural population, who otherwise pursue traditional Fijian village-based subsistence lifestyles and are the proud custodians of their unspoilt natural heritage. Consequently, tropical cyclones that cause damage to coral reefs and coastlines are perceived as ‘negative’ events by local communities, so the TC Tomas event provided additional impetus for this study within this context.

5.3 Features of Tropical Cyclone Tomas, March 2010

TC Tomas (12–17 March 2010) was the ninth cyclone to develop within the South Pacific basin over the 2009/2010 wet season (November to April). The latter part of this season was marked by unusual climatic activity east of 180°, where four cyclones reached category-3 or 4Footnote 1 intensity in rapid succession over a relatively brief period from February to March 2010 (Terry and Etienne 2010a, b). Of these storms, TC Tomas was the most destructive for Fiji in almost a decade (Terry et al. 2004). El Niño conditions of moderate strength played a part in the heightened levels of disturbance in early 2010 (FMS 2010), allowing warm sea temperatures to stretch far east of the Date Line. This warm pool developed into a broad region of low pressure within which atmospheric instability gave rise to numerous eddies showing potential for spin up into cyclonic vortices.

The nascent depression that later matured into TC Tomas first became organised north east of the Fiji Islands in Samoan waters. Initially following a west south-westwards track, the storm strengthened and attained cyclone status (winds >34 kts) north of Wallis Island at 12 p.m. Fiji Standard TimeFootnote 2 (midnight GMT) on 12 March. Thereafter, the system matured rapidly. It decelerated in speed but intensified to category-4 strength, as illustrated in Fig. 5.2. The eye of TC Tomas passed within 30 km of the east coast of Taveuni Island in northern Fiji on 15 March and then migrated through the Lau group of islands over the next day. The significance of this track for the Taveuni coastline is that sustained forceful winds of 100 knotsFootnote 3 (185 km/hr) and gusts up to 140 knots (259 km/hr) were blowing onshore as the storm approached. The furious winds drove large waves, heavy swells and powerful storm surge ashore, causing severe flooding of low-lying coastal areas and much damage to coastal vegetation, villages and transport networks.

Fig. 5.2
figure 2

Path of Tropical Cyclone Tomas in mid-March 2010 as the system traversed eastern Fiji waters. Positions of the cyclone eye at six-hourly intervals are shown by circular markers, with adjacent timings given in Fiji Standard Time (FST, 12 h ahead of Greenwich Mean Time). The intensity categories written in the eye markers correspond to the following wind speeds (sustained maximums over 10 min intervals): category-1: 34–47 knots; category-2: 48–63 knots; category-3: 64–85 knots; category-4: 86–107 knots. Inset figure at bottom left shows a satellite thermal image of TC Tomas at 12:30 a.m. FST on 16 March. The storm eye is located near the northern tip of Taveuni Island while at category-4 intensity. Original satellite image courtesy of NOAA

Full size image

5.4 Field Procedures and Observations

Four months after TC Tomas, detailed geomorphic field investigations were carried out in July 2010 along the eastern coast of Taveuni Island, with a deliberate concentration of effort between Bouma and Lavena villages. This section of coastline experienced the full brunt of storm-driven waves due to its exposure in relation to the position and orientation of the cyclone track. Table 5.1 provides specific locations and details of bouldery coastal deposits examined at four field sites.

Table 5.1 Details of coastal measurement sites in eastern Taveuni: coral boulders and beachrock fragments
Full size table

Coral boulder fields in two in reef-platform areas were mapped using a GPS Garmin 76 CSX device to 2–3 m precision. The dimensions of individual boulders (a, b, c axis length) were concurrently measured and recorded. During field work, a number of Fijian villagers with an intimate knowledge of the local coastline acted as guides. A principal task was to identify which reef-platform boulders were fresh deposits that had appeared in the aftermath of TC Tomas. This was easily accomplished through recognition by the guides and the noticeable greenish-white colouration of new boulders, which stood out clearly against their older weathered neighbours that pre-dated the cyclone in question (Fig. 5.3). The fresh clasts were dispersed across the reef platform and thus augmented the existing boulder fields. Of interest was that no rubble ramparts of mixed coralline gravels and boulders were observed parallel to reef crests, such as of the type described both in Tuvalu and Samoa after violent cyclones struck those neighbouring countries in 1972 and 1990 respectively (Maragos et al. 1973; Zann 1991). Instead, however, three tongues of loose coral gravel were formed on Taveuni’s reef flats, oriented perpendicular to the reef edge. Several storm ridges of shingle were also built up along the highest part of beaches between Bouma and Lavena. Several villagers independently confirmed these as new constructional features resulting from TC Tomas (see Etienne and Terry 2012, for further description).

Fig. 5.3
figure 3

a Photograph of the fringing reef near Lavena village showing several of the numerous coral boulders strewn across the reef surface. The boulder with the greenish-white colour nearest the observer is a new clast thrown up by TC Tomas. It contrasts markedly with the surrounding group of older boulders behind, brown and weathered in appearance, which pre-date this cyclone. b Angular beachrock fragments excavated from in situ exposures along the Bouma–Lavena shoreline. (Photos by J. Terry, July 2010)

Full size image

A total of 196 coral boulders were observed, of which 38 were fresh boulders produced by TC Tomas, and 158 were old boulders. According to the methods outlined by Etienne and Paris (2010), any scars, gouges, percussion marks or striations on boulder surfaces, or on the adjacent reef platform, were recorded as indicators of the mode of boulder transport. Nevertheless, the majority of pre-existing coral boulders bore no indication of recent movement. The lack of surficial scorings on them or on the reef flat in their immediate vicinity is interpreted as evidence that TC Tomas did not generate sufficient wave energy to set them in motion.

In several places, exposed beachrock at the shoreline provided a source of coarse clastic debris. Angular slab-shaped beachrock fragments had been broken off and subsequently moved. Using similar procedures to those mentioned above, two groups of clasts were inspected approximately 1 km north of Lavena (Fig. 5.3). Fifty clasts in total were measured, all transported by TC Tomas. Careful scrutiny revealed that beachrock-derived material had been carried in two directions, either pushed inland by wave runup or pulled backwards by return flow towards the beach–reef interface and onto the adjacent landward edge of the fringing reef.

5.5 Results and Discussion

5.5.1 Boulder Quarrying and Remobilisation

Expansive coral boulder fields occupy the fringing reefs in two areas near Bouma and Lavena villages. On many individual boulders, however, features of bioerosion (cavities eroded by sea urchins and other marine creatures), solution weathering (karstification) and marine erosion (nascent wave-cut notches) revealed that the majority had stood in place for a considerable period of time prior to TC Tomas. Among the total number of boulders inspected (n = 196), the recent deposits (n = 38) were easily distinguished by their outer greenish-white colouration. All but two of these are relatively small boulders (<0.5 m3), with the most sizeable fresh boulder having a volume of 4.85 m3. Information on boulder volume frequencies for both fresh and older deposits is displayed in Fig. 5.4. Perhaps unexpectedly, in as much as the boulder fields did contain some fresh deposits, there was a general scarcity of recently uprooted or overturned coral colonies on the reef flat itself. Such evidence suggests that coral colonies on the reef edge or external reef slopes contributed primarily to the production of new boulders, although sea conditions during the field expedition were unsuitable for submarine work to confirm this.

Fig. 5.4
figure 4

Frequency distribution of coral boulders by volume on Taveuni reef platforms near Bouma and Lavena, as derived from measurements of boulder axial dimensions (a,b,c axis). Fresh boulders emplaced by TC Tomas are represented by solid black bars, while grey bars include all boulders measured (n = 196)

Full size image

It is noteworthy that pre-existing coral boulders are substantially bigger than those delivered by TC Tomas: the largest exceeds 40 m3 in size, while 16 others are bigger than the largest boulder moved during the storm. No evidence of recent mobilisation could be found on these old boulders: neither scars nor crush marks on the boulder faces (which would indicate rolling), nor grooves on the adjacent reef surfaces (which would indicate sliding). From this, the following conclusions may be drawn. First, the boulder fields were mostly constructed prior to TC Tomas by unknown HEMI events, probably of significant antiquity. Second, waves driven onshore by TC Tomas were unable to mobilise the older boulders. Third, TC Tomas added approximately 20 % to the sum of boulders comprising the existing deposits.

At the shoreline, the quarrying and transport of fresh beachrock fragments is of special interest. Examining beachrock fragments offers one key benefit over RPCBs: identification of their original source is often immediately possible, simply from their neat ‘jigsaw fit’ into nearby exposures of in situ beachrock. Knowing the exact source point enables the displacement of beachrock fragments by cyclonic waves to be determined precisely. Subsequently, measuring fragment dimensions and transport distance allows reconstruction of both runup and backwash flow velocities at the shoreline. Such information is valuable in another way: it reveals the available energy at the shoreline after waves have been attenuated by passing over the fringing reef. Measured clasts quarried from beachrock near Lavena are very heterogeneous in size: the smallest weighs 11 kg whereas the biggest is over 400 kg. Although 90 % of the boulders weigh less than 150 kg, this can be explained by the jointed nature of the beachrock material, with calcarenite slabs being easily broken in smaller pieces during wave transport. Hydrodynamic data deduced from beachrock clasts can be compared with wave energy on the reef platform itself, obtained from similar calculations using values derived from the reef-top boulders mentioned above.

5.5.2 Transport Mechanisms and Flow Velocities

During post-fieldwork data treatment, the minimum flow velocities needed to set in motion coral boulders of particular shapes and in various pre-transport settings were estimated using the hydrodynamic transport equations published by Nott (2003) and revised by Nandasena et al. (2011). Owing to a lack of evidence for significant erosion on the reef surface, the assumption is made that the fresh boulders were quarried from submerged reef-crest or fore-reef locations. Using hydrodynamic equations therefore yields minimum velocities for the flows that emplaced these boulders torn from the reef framework by TC Tomas. Furthermore, when applied to pre-existing boulders showing no signs of movement (assumed partly-subaerial environment on the reef platform), the equations give an indication of flow velocities that were not achieved by the strongest waves driven over the reef by the cyclone. Simultaneous mapping of combined data from fresh and older boulders then becomes valuable, as this provides a first approximation of spatial patterns in flow velocity ranges experienced during TC Tomas over the reef platform (Fig. 5.5).

Fig. 5.5
figure 5

Minimum water flow velocities (m/s) needed to set in motion coral boulders resting on the fringing reef near Bouma (above) and Lavena (below), as deduced from hydrodynamic equations (Nandasena et al. 2011) applied to measured boulder dimensions. See Fig. 5.1 for the location of the two boulder fields. Red values refer to fresh coral boulders produced by TC Tomas wave action. Black values represent pre-existing boulders that were not displaced during the storm. Together, the two sets of values give an appreciation of the minimum (red) and maximum (black) flows experienced from place to place across the reef flat during TC Tomas. The eye symbols show the direction of view for the inset photographs. (Photos by J. Terry, July 2010)

Full size image

Following on from this, a transport histogram plotted for the entire dataset indicates the minimum flow velocity values required to move coral boulders according to the various possible modes of boulder transport, i.e. sliding, rolling or lifting (Fig. 5.6). By adopting the conservative approach that fresh coral boulders were transported on the reef surface only by sliding or rolling, i.e. without saltation, minimum flow velocities lie between 1 and 3.8 m/s. This reveals that in spite of TC Tomas attaining category-4 intensity, waves breaking across the fringing reefs generated insufficient flow velocity to set in motion many pre-existing boulders, the largest of which requires flows of more than 5 m/s for transportation (sliding) as inferred from the Nandasena et al. (2011) equations.

Fig. 5.6
figure 6

Transport histogram for reef-platform coral boulders: combined data from Bouma and Lavena boulder fields. Individual boulders are plotted side by side along the horizontal axis. Fresh boulders deposited by TC Tomas have been grouped on the right-hand side of the graph, while pre-existing boulders are grouped on the left

Full size image

Detached fragments of beachrock on the shoreline were similarly helpful. At least some of these were probably plucked out from original (joint-bounded) positions within outcropping beachrock. Imbrication was also common. This implies that incoming waves must have extricated these slabs by lifting and then transported them by saltation. In contrast, sliding is the most likely mode of transport for freed slabs that were then reworked by the backwash flow. Accordingly, it is possible to reconstruct the minimum flow velocities at the shoreline using the familiar hydrodynamic equations. Runup flow velocities range from 2–4 m/s (average 3.41 m/s) while backwash flow velocities fall below 2 m/s (average 1.85 m/s) (Fig. 5.7). For comparison, analogous data obtained in French Polynesia on beachrock dislodgment by Tropical Cyclone Oli in February 2010 (Etienne 2012) yield mean runup flow velocities of 3.98 m/s on Tetiaroa Atoll and 4.24 m/s on Huahine Island. For the present Taveuni case, if instead the conservative option is accepted that all beachrock clasts were freely detached prior to transport (i.e. in non-joint bounded positions), then 2 m/s was the minimum flow velocity required to mobilise the entire set measured (Table 5.2).

Fig. 5.7
figure 7

Transport histogram for beachrock fragments at two adjacent shoreline sites, 1 km north of Lavena. Refer to Fig. 5.1 for location of field sites. The vertical axis shows the minimum velocity required to set a clast in motion according to mode of transport. It can be seen that 46 of the 50 slabs are stable with a flow velocity below 1 m/s. When the flow velocity reaches 2 m/s, all beachrock fragments except one are mobilised (see Table 5.2)

Full size image
Table 5.2 Progressive instability of beachrock fragments with increasing flow velocity (Lavena study site; n = 50)
Full size table

Flow velocities inferred from fresh coral boulders on reef platforms and beachrock fragments at the shoreline are complementary datasets. Comparison between them provides a valuable indicator of flow velocity reduction over the reef flat due to energy dissipation. Here it can be seen that flow velocities generated by waves breaking at the shoreline during TC Tomas are comparable to, if not greater than, flow velocities over the reef flat. This finding is important as it is somewhat counterintuitive and implies that fringing reefs do not necessarily afford a high degree of protection to the shoreline against cyclone-generated waves.

5.5.3 Caveats to Findings

Calculated flow-velocity values assume that the reef platform is horizontal and smooth. In reality, reef surfaces are rough and uneven owing to surface rugosity associated with the patchwork growth of coral colonies such as microatolls in the back-reef environment. Therefore, higher flow values were probably experienced during TC Tomas.

When attempting to reconstruct boulder transport conditions during a coastal inundation event, knowledge of pre-transport settings is critical for selecting the most appropriate hydrodynamic equation from those available (Nott 2003). One reason is because a tropical cyclone typically generates numerous waves and surges over several hours. Correspondingly, coarse sediments are moved in an incremental fashion and their pre-transport environment might vary over time. To illustrate, a fresh beachrock clast may be sourced from a joint-bounded position within the beachrock mass. Initially, it must be exposed by the removal of overlying beach sediments, then quarried out. The flow velocity required to liberate the clast from the beachrock outcrop is especially high. Post-detachment, however, subsequent transport requires lower velocities in either a wholly submerged or partially subaerial environment. This depends on whether the fragment is transferred into a beach, reef flat or back-beach setting. Here, Fig. 5.7 demonstrates how initial flow velocities must exceed 2 m/s to lift beachrock fragments out of the country rock, but less energy is needed to shift loosened clasts in subsequent waves (between 1 and 2 m/s). This idea applies equally to beachrock slabs and reef-platform coral boulders. In consequence, interpreting mapped flow velocities derived from hydrodynamic equations for a cyclone event with numerous successive waves should therefore be tackled with caution, since freshly-produced deposits are likely to have experienced changing transport environments through the lifespan of the storm.

Finally, although TC Tomas passed very close to the studied coastline on Taveuni Island, it is evident that its primary contribution was to augment existing boulder fields rather than the creation of new fields in alternative locations. Moreover, ignoring a few isolated exceptions, both fresh and earlier coral boulders are relatively-well confined to discrete sites and do not demonstrate any remarkable linear continuity along the reef edge, in spite of the availability of reef platforms as suitable sites for deposition. This implies that local variability in physiographic attributes of the affected reef and coastline strongly influences both the production and preservation of boulder fields in eastern Taveuni. These attributes may include reef geomorphology and configuration, ecology, roughness, sensitivity to wave attack and relaxation time since the last episode of disturbance.

5.6 Conclusions

TC Tomas was a category-4 cyclone which tracked through Fiji waters in March 2010 and was the most intense storm to strike Taveuni Island in recent years. Investigation of reef-platform boulder fields comprising large numbers of both fresh and older coral boulders enabled discrimination between the storm in question and prior HEMI events. Of interest, although TC Tomas was sufficiently energetic to produce fresh debris (mostly sourced from reef crests or fore-reef slopes rather than the reef surface), this debris constitutes only a minor fraction of original boulder fields (approximately 20 %). In addition, measurement of coral boulder sizes reveals that unrecorded events in the past delivered much coarser material than TC Tomas. This evidence does not necessarily imply that past storms were of higher intensity, as other explanations are possible for the presence of huge boulders. Possibilities include ancient tsunamis, Holocene inundation events at times of higher sea-level stands, more direct approaches of palaeo-cyclones, or rogue waves coinciding with spring tides. Nevertheless, since TC Tomas was unable to remobilise the larger original boulders, we can be reasonably certain that higher-energy conditions (waves and surging currents driven across reef flats) were a feature of the past coastal inundations that originally emplaced these pre-existing deposits.

Recommended future work would be to investigate the timing and frequency of HEMI events throughout the late Holocene at the Taveuni study site, as has been achieved for other coastlines in tropical regimes elsewhere (Hayne and Chappell 2001; Radtke et al. 2003; Zhao et al. 2009). This might be accomplished through accurate dating of boulder carbonate material by uranium-series methods (Gale 2009; Yu et al. 2012). In the Taveuni context, because conservation of the Bouma–Lavena coastline and adjacent fringing reefs is a matter of priority as a Marine Protected Area, information on the magnitude-timing-frequency of prehistorical HEMI events would certainly assist in assessing the future long-term risks of coastal hazards in this area.