A Review on Multidecadal Coastal Changes at Funafuti, Tuvalu from 1897 to 2015

Abstract

Tuvalu is a small reef islands country in the Pacific Ocean. Its coastal regions are very much dynamic due to the profound effects of tropical cyclones and sea level rise (SLR). However, research works on coastline dynamics of Tuvalu mainly cover its capital, Funafuti. Therefore, this review summarizes the extent of long-term coastal changes in different islets of Funafuti and on overall Tuvalu. In Funafuti, highly accreting areas are Te Afualiku, Fuafatu, Motugie, and Amatuku, and highly eroding areas are Fuagea and Tefala with the fully disappeared islet of Vasafua after 2005. However, in spite of different causes and supposition of scientists on disappearing these lands the accretion is more dominant than erosion which resulted in 7.3% net increase of land areas of Tuvalu over 117 years till 2015. Severe tropical cyclones mainly caused accretion of land areas by forming coral rubble rampart formation and further reworks and erosion to small sandy islands whereas frequent low-energy cyclones mainly caused erosion. Though, till now severe erosion of coastal areas are not evident by global SLR, islets of Funafuti experienced remarkable shoreline increase as formation of 30-40 m wide rubble rampart formation along 19 km in 1971 by tropical cyclone Bebe and net increase of area of 3.45 ha by tropical cyclone Pam in 2015. In spite of such overall accretion of coastal areas several scientists suspect drowning of its areas in future because of high SLR (~5.1±0.7 mm/year) at Funafuti which supposedly will not work as a breakwater anymore. Thus, protection measures should be taken to prevent coastline erosion as well as land reclamation activities should be done following the global examples.

1. Introduction

Tuvalu is a microstate (4th smallest country of the world) consisting of only around 26 km² land area (McLean and Hosking, 1991; Siaosi et al., 2012). Formerly it was named as Ellice Islands till its independence in 1979. It consists of a widely scattered group of nine coral islands and atolls, namely Nanumea, Niutao, Nanumaga, Nui, Vaitupu (the largest), Nukufetau, Funafuti, Nukulaelae and Niulakita from north to south in the Southwest Pacific Ocean (Rodgers and Cantrell, 1988; Xue, 2005).In spite of its very small total land area this coral reef island chain extends for around 579 km and bounded by Kiribati to the north, Fiji to the south, and the Solomon Islands to the west. The word Tuvalu means the cluster of eight, so named because the tiny southernmost Niulakita was not considered as part of the group in traditional times (Munro, 1982). The small islands and atolls of Tuvalu have very long coastlines exposed to the open sea. All these islands are of very low height with elevations not greater than 4.6 m above the sea level. Due to having long coastlines with smaller land areas most of the islets areas remains very close to the coastlines, and the coastal areas are very much dynamic due to this low elevation, unconsolidated coralline calcareous sand and gravel dominated soils, small land areas, etc. (McLean and Hosking, 1991; Munro, 1982; Siaosi et al., 2012; Siose et al., 2018).

Tuvalu is a densely populated country which population increased approximately by 12.8% (increased by 1,221 people) from 2002 to 2012 (Apinelu, 2022). As the country is composed of very small isletsin the atolls or small islets, the whole population of the nation has to live very close proximity to coastlines, and thus exert enormous pressure on their physical and biological processes and resources. Hence, any change, even a minor change in coastline will have great impacts on the whole population of the country. Among various types of landforms coastline is very much prone to sudden or very short period temporal change (Darwish and Smith,2021).Coastal erosions are mostly contributed by climate changes and supposed to be intensified globally because of global warming (Yang et al., 2021). McLean and Hosking (1991)suspected minor changes in the reef island outlines of Nanumaga, Niutao, Niulakita, and Vaitupu, whereas larger atolls’ reef platform space is still available for island accretion, and storm waves could play a substantial role in future islet development. In order to develop sustainable coastline zone management policies of a country, especially very small island country like Tuvalu, proper understanding of coastline changes could serve as a strong basis. Therefore, it is necessary to know the coastline changes of Tuvalu. Literatures on coastline changes in Tuvalu are reviewed here to find out the nature of coastline changes and vulnerable areas.

Coastline alteration is usually caused by effects of natural factors and human activities, and thus has important impacts on regional ecology, economy, and society (Fan et al., 2020). In general, the major factors affecting coastline changes are extreme weather events like hurricane (Badru et al., 2018),sea level rise (SLR) (Fan et al., 2020; Ranasinghe et al., 2011; Yang et al., 2013), and human disturbances (Fan et al., 2020; Mujabar and Chandrasekar, 2013; Syvitski et al., 2009). Coastline change could include diverse ranges of time scales covering long-term geological evolution to sudden extreme weather events (Fan et al., 2020). Currently most of the world’s beaches are under the threat of long-term changes (Fan et al., 2020). The change in Tuvalu islands’ coastlines were mainly a result of natural environmental changes, such as tropical cyclones, SLR (Ceccarelli, 2019), raining and flooding, etc. of which tropical cyclones and SLR are greatly responsible. McCubbin et al. (2015) identified that vulnerability to climate change in Funafuti is greatly influenced by several climatic forces including strong winds and extreme SLR. Considering these, in this study we will review long-term coastline changes of Tuvalu from 1897 to 2015 in relation to tropical cyclones and SLR. However, before reviewing the coastal changes, it will be reasonable to define coastline, and then describe the basic geographical and geological characteristics of Tuvalu. Coastline can be simply defined as the tangent line that intersects sea and land surfaces. For practical purpose of coastline detection Boak and Turner (2005) defined coastline as visibly recognizable features in coastal images, or the intersection of a tidal datum from coastal profiles, or detectable proxy shoreline features in digital images regardless of human eye discrimination capability.

Through the next sections we will describe the methods of data selection followed in this study, then will briefly describe the geology of the islands and atolls, followed by brief description of the major changes occurred to the islets including identification of hotspots of changes. Afterwards, we will present the major devastating tropical cyclones that caused coastline changes in Tuvalu, and finally analyze the coastal changes in relation to SLR followed by conclusion of the review.

2. Materials and Methods

At first the Google Scholar, then Researchgate, and finally other Google databases were utilized from 1950 to 2022 for document searches by applying combinations of keywords as (Funafuti / Tuvalu) and (shoreline change/shoreline dynamics/shoreline erosion/coastline change/coastline dynamics/coastline erosion) and/or (sea level rise/tropical cyclone/hurricane/storm). We only included articles, reports, etc. published in English-language and excluded other types of documents, for instance articles which couldn’t access as fully except only the abstract. In this review protocols of the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement (Page et al., 2021) is followed to select articles for analysis(Fig. 1). Moreover, during studying the selected documents some other important documents were also identified from cross-references which were also included for reviewing after through studying. Variations in the spellings of some islands and islets of atolls are found in different sources; however, in this review we have used only single spelling for each case. Preliminary results of the search were 1,784, but after final screening only 80 documents were reviewed. It could be noted here that most of the in depth coastal change detection works were done on Funafuti rather than all islands and atolls of Tuvalu; thus in this study we also emphasize on Funafuti.

Fig. 1. Flow diagram for choosing review documents using Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidance.

Generally, for coastline dynamics research ancient maps, traditional actual trace measurements, remote-sensing images, etc. are chiefly used as sources of data. Most of the coastline change literatures in Tuvalu reviewed here mainly focused on analysis of space-time changes with some articles correlating it to coastline driving forces. In this century various geographic information technologies are widely applied to study coastline changes of Tuvalu. Most of the Tuvaluan islands and islets have mangroves that occur as fringes around the enclosed lagoons. Edge of vegetation (EoV), i.e. seaward edge of island mangrove was thus used by the most of the literatures reviewed here to detect coastline change (Ceccarelli, 2019; Woodroffe, 1987).

3. Discussion

Acquisition of coastline data is prerequisite for coastline change detection. Long ago traditional paper based ancient maps and historical coastline data from geological survey were used to obtain coastline information which are usually acquired by survey based actual measurement. Later on with the development of photography historical aerial photographs were interpreted for the same purpose. Geology based coastline change research works originated in 1990s’ and multi-temporal remote-sensing data are nowadays become the most commonly used method for coastline change detection research; however it is subjected to the question of data accuracy that could affected by resolution and interpretation method (Li et al., 2002). Thus, from the beginning of this century case studies on detection of short-term spatiotemporal changes have sharply increased with further advances in geographic information technology. The documents reviewed here used these various types of data to detect coastline.

The islands and atolls of Tuvalu were originated from a volcanic seamount chain that raised from the Central Pacific Basin (Ceccarelli, 2019). Seamounts’ tops formed the islands and the atolls were formed biogenically on top of submerged volcanoes (Ceccarelli, 2019;McLean and Hosking, 1991). Each atoll possesses many small islets at the outer boundary of the atoll and thus forms at least one lagoon that is usually connected to the sea with one to multiple natural channels (Ceccarelli, 2019). The islets’ soils are dominated by loosely attached sediment deposited on consolidated coral rubble formed underneath the emergent reef flats (Ceccarelli, 2019; Dickinson, 1999). Phosphate limestone type soils occur to various extents(Ceccarelli, 2019; Rogers, 1992). Regardless of island size and shape the islands have common characteristic natural landform units like high ridges composed of coarse coral grave, cobbles or platy rubble on the windward or oceanward side, low-height ridges on the leeward or lagoonward side which are generally sandier, broader and have more gentle slopes, and depression or flat among these (McLean and Hosking, 1991).

Only five of the Tuvalu reefs, namely Nukufetau, Funafauti, Nanumea, Nui, and Nukulaelae, can be classed as true atolls due to having continuous rim of reef at floor near the surface of the sea surrounding deeper lagoon (McLean and Hosking, 1991). The remaining three true atolls have a broad continuous reef rim which is exposed at extreme low water, and thus isolating the lagoon from the open sea. Out of these true atolls only Nukufetau and Funafuti have natural passages that permit ship access to the lagoon (McLean and Hosking, 1991). Nanumea has an artificial channel on its south-western side. Nanumaga, Nuitao and Niulakita possesses continuous land margin that occupy the flat tops of table reefs without surface draining lagoons (McLean and Hosking, 1991). Vaitupu possesses composite characteristics of the atoll and table reef forms as its small lagoons are almost blocked by land and reef flat, and the lagoons only fill through shallow channels at spring tides. However, for the convenience of description henceforth in this review Vaitupu will be mentioned as table reef island and the term reef platform (RP) will be used to mention these table reefs including Vaitupu (Kench et al., 2018a; McLean and Hosking, 1991; Munro, 1982). In Tuvalu the reef tops vary greatly in alignment, size and shape (Fig. 2), and an indication of this variability is given in Table 1.

Fig. 2. Tuvalu maps. (a) Tuvalu (red polygon) in the world map. (b) Distribution of atolls and table reefs of Tuvalu Archipelago. (c) Outer boundaries of reefs and landmasses (blue and red lines, respectively) in same scale regardless of exact geographic positions.

Table 1. Some basic characteristics of islands of Tuvalu

3.1. Coastal Change Result Analysis

3.1.1. Overall Changes in Tuvalu Coastline

Xue (2005) conducted surveys to identify coastal changes for several months of 1995 to 1996 on Fongafale and Amatuku Islets of Funafuti Atoll, and to the atolls of Nukufetau, and most of the remaining atolls and islets of Tuvalu. Based on his own surveys and other research works he identified intense erosions immediately after several human activities. In 1942 a 2.3 km long and around 25–40 m wide area was reclaimed by the US army along the beach of central Fongafale in order to build a runway. For that purpose, several pits were dug on the lagoon reef flats which caused erosion due to change of sediment transportation patterns along that lagoon coast. In Amatuku Islet, lagoon excavation of artificial channel and quarry pit during the World War-II caused most of the loose sediment to move away causing significant erosion of the northwest edge of this islet by 45 m from 1943 to 1996. The mostly sandy and gravelly nature soil of islets of Nukufetau Atoll facilitated the earliest recorded coastal erosion after the World War-II at the northwest coast of Savave Islet from mining on beach and reef flat. In Vaitupu, cyclones Kina and Nina caused erosion of sandy southwest coast which resulted in receding the shoreline by around 5–6 m. Moreover, construction of fisheries harbour in 1992 and mining activities at the northeast, northwest, and southeast coasts of Vaitupu for the purpose of roads and houses construction also caused erosion at sandy areas, whereas in some other places shorelines were accreted by 12–13 m till 1996. However, on its northeast coast erosional and prograding shores appeared alternately which are supposed to be from unusual northerly storm waves. In Nanumea, a boat channel of 450 m length, 18 m width and 2 m depth was excavated through the reef flat during 1943 which caused coastal erosion by transporting sediment out of the channel. In the Nui Atoll, excavation of boat channel and taking of stones out of the reef flats for seawall construction caused erosion (Xue, 2005).

By summarizing the findings of Kench et al. (2018a) it is found that from 1971 to 2014 only some smaller-sized islands eroded, whereas majority of the islands (73.5%) expanded. Only one islet (0.08 ha) of Nukufetau Atoll was entirely eroded and this atoll decreased in land sizes by 0.01% whereas the remaining eight atolls increased in size. Total 73 islands thus increased in size by 80.7 ha (2.18 ha mean increase) showing <1 to 113% growth which resulted in net total increase of 73.5 ha (2.9%) land area in Tuvalu. Largest absolute increase occurred to Vaitupu (11.4 ha, 2.2%), Nui (10.4 ha, 7.1%), Nukufetau (5.3 ha, 4.2%), Nanumaga (4.7 ha, 1.6%), and Fongafale Islet (4.6 ha, 3%) of Funafuti Atoll. The other 28 islands (27.7% of total) of Funafuti Atoll were net decreased by 7.24 ha only, and erosion was prevalent to the smaller islets. Four islands (initial size <0.5 ha) were decreased by more than 50% (Fig. 2). Great absolute decreases occurred in Nanumea RP (2.88 ha, 1.32%), Tepuka (0.89 ha, 8.35%), Fuagea (0.74 ha, 45.5%), and Fualifeke (0.51 ha, 6.16%) islets of Funafuti Atoll. Out of the 19,403 shoreline transects taken by Kench et al. (2018a) over different areas of all 9 islands, 44% were found to be accreted, 23% were found to be eroded, and the rest 33% remained stable. The balance of land morphology through erosion and accretion of islands over a time span resulted in net change of island area and effective island migration. Greatest variation in shorelines occurred to the islets on the larger atoll rims.

3.1.2. Coastal Changes in Funafuti

Funafuti Atoll (8.5°S, 179°E; Fig. 2) is the capital of Tuvalu. It extends around 25 km from north to south and about 17 km along east to west (Hisabayashi et al., 2018). Hisabayashi et al.(2018) and Kench et al.(2015) listed 32 islands, but McLean and Hosking (1991) mentioned 34 islands of Tuvalu. These difference in islands numbers occurred due to fuse of closely located islets over time, disappearing of small islets, or not considering very small islets (<0.5 km²). Considering the land area, Funafuti is the 6th largest among the islands and atolls of Tuvalu, but it is the most widespread with largest lagoon area and longest coastline among all atolls and islands. According to Kench et al. (2015) only Amatuku and Fogafale Islets of Funafuti Atoll are inhabited, but houses and buildings (probable evidence of human settlements) of various sizes and numbers can visually be also identified in Funafala, Funagongo, Mulitefala, Fualefeke from Google Earth Pro.

Table 2. Overall changes to the coasts of Funafuti Atoll

Hisabayashi et al.(2018)showed substantial decreases in land area to Fuagea, Vasafua, and Tefala from 2005 to 2015. Vasafua as the smallest islet in 2005, and as a consequence of erosion it was found to be disappeared in 2015. Except this no other islet is found to be disappeared. Another small islet, Tepukasavilivili, is mentioned by only Webb (2006) among the listed references and found to be vegetated only before 1990 and later found to be as only an elevated rubble shoal. For example, two islands disappeared or fused with neighbouring islet. Fongafale is its capital islet which is the largest islet of Funafuti Atoll. However, this islet possesses a narrow and very low elevation geography which made it vulnerable to flooding during extreme natural events like tropical cyclones (Government of Tuvalu, 2012). Though Tuvalu has experienced very high rates of SLR (~5.1±0.7 mm/year) during the last few decades, a great degree of uncertainty still exists in case of coastline dynamics with respect to accelerating SLR (Ballu et al., 2011; Becker et al., 2012).

The study period of coastline changes works on Funafuti are presented as timeline in Fig. 3. It can be seen that in depth changes study over long periods were mainly done by Yamano et al. (2007), Kench et al. (2015), Webb (2006), and Webb and Kench (2010). Other researchers conducted studies for short time period equal or less than 10 years timespans. Among these Webb (2006) found the greatest changes of net 2.75 ha increase of total land area of Funafuti from 1941 to 2003. The major cyclones’ names according to year are also placed in the timeline which indicate that the major change was chiefly caused by tropical cyclone Bebe in 1972 which will be discussed later in related subsection. All these articles identified increase or no major change in net area except Hisabayashi et al. (2018) showing decrease of 0.35 ha area which was mainly caused by the devastating cyclone Pam in 2015. However, this net decrease was estimated over around 10-years period from 2005 to 2015. If the changes caused by cyclone Pam in 2015 is considered comparison with those of the previous available year data, that depicts different pictures of changes; 3.45 ha net increase in 2015 (Fig. 4).

Fig. 3. Key changes in Funafuti coastlines from reviewed sources (top row), and timeline of major tropical cyclones in Tuvalu (bottom row).

Fig. 4. Overall changes in Funafuti Islet’s area from 1897 to 2015: (a) increased by >15%, (b) decreased by >15%, and (c) changed by 15 to -15. For (a) and (b) the left side Y-axis indicate the land sizes for the red dashed lines (Funafala and Fongafale, respectively).

It can be seen that islets of Funafuti showed different ranges of changes over time and in most cases major changes in percent size occurred to comparatively small islets (Figs. 4 and 5). Major increase in size is evident to Te Afualiku, Fuafatu, Motugie, Amatuku, Aalau-Teafuafou, Falaoigo, Mateiko, Funafala, Luamotu, Telele-Motusanapa, Motuloa. Here, the highest increase is seen for the sandy islets Te Afualiku and Fuafatu, and gravel dominated Motugie Islet. Major decrease is evident to gravel dominated Tefala and sandy Faugea, and only Vasafua fully eroded. To the capital, Fongafale, very minor increase occurred which is partly supposed to be due to some man-made constructions (Kench et al., 2018a). For Motungie Islet, Kench et al. (2015) showed 0.07 to 0.19 ha from 1897 to 2013, and Hisabayashi et al. (2018) similarly showed 0.18 and 0.19 ha from 2005 to 2015, whereas Webb and Kench (2010) showed 4.97 to 5.03 ha from 1984 to 2003 which seems to be erroneous.

Fig. 5. Soil types and percent changes in islets’ size of Funafuti Atoll from 1897 to 2015. Exceptions: Fualefeke-Paava (1897–2013 by circle placed at between the enlarged islets), Fualifeke (2003–2015), Paava (2003–2015), and Vasafua (2010–2015).

From the total percent area changed calculated for the whole period, the highest area increase can be accounted for Teafualiku, followed by Fuafatu, Moutulo, Amatuku, and Avalu-Teafuafou (Fig. 6). However, the highest decrease in percent area for the whole period is accounted for Vasafua, Fuagea, and Tefala. From the observation of yearly change of islets’ sizes of Funafuti in graphs(only for Kench et al., 2015) in the year 2013 most of the areas decreased and only some areas increased. However, in the previous year (2012) and next few years (2014, 2015) from other sources the areas were found to be almost unchanged. This might be the difference in methods for identifying land areas. According to the Hisabayashi et al. (2018) the highest unit percent decrease in land area occurred to Fuagea for 10-years period. On the contrary, the highest unit percent increase in land area occurred to Teafualiku according to Kench et al. (2015) for the period of 117 years (1897–2013), whereas the second highest decrease was accounted for the same islet according to Hisabayashi et al. (2018). On the contrary, very little changes were observed for the land from Webb (2006) for 1941 to 2003, and Webb and Kench (2010)for 1984 to 2003.

Fig. 6. Comparison of Funafuti Islet’s land area changes (percent)from different sources: land area changes at whole period (top), and per year (bottom).

Coral islands are found to be changed morphologically like migrating, prograding, eroding, rotating and merging with adjacent islands over time (Fellowes et al., 2022). Kench et al. (2015) classified island adjustment in Funafuti into five major types: (1) accretion or erosion of some islets around their entire perimeter, (2)shift of islands through differential erosion of ocean and lagoon shorelines: lagoonward migration in most cases (e.g., Paava, Fualefeke, Tepuka), and migration of many islets towards the ocean reef edge on the leeward (western)side of Funafuti Atoll with no cases migration exceeding 50 m, (3) islets at the eastern margin of Funafuti were laterally extended as gravel spits from longshore transportation of sediments supplied during several storm events (e.g., maximum 60 m extension at Fatato),(4) merging of neighboring islands by filling up the shallow passages in between through alongshore sediment fluxes and then embayment infilling by lagoonward-directed trapped sediments (e.g., Avalau, Teafuafou), and (5)rotation of sandy islets (e.g., Paava, Fualefeke, Tepuka). Kench et al.(2014)showed the net area reduction of Tupuka by 23% over the past century through shoreline erosion at lagoon from 1897 to 1984, and later on elongation towards the terminal end sfrom 1984 to 2011 causing clockwise rotation its land area on the reef platform.

Some islands become fused as seen in the images in some years like Avalau-Teafuafou (1971), Fualefeke-Paava (1971), and Telele-Motusanapa (Kench et al., 2015). Changes in size among the studied years are shown as line-graphs in Fig. 4. Based on the intensity of overall percent changes in size among the studied first and last year the islands are separated into three groups for the purpose of discussion, namely greatly increased (>15%), greatly decreased (>15%), and moderate change (15% to -15%). However, to fit the lines in the graph the changes for Funafala and Fongafale are indicated by right side Y-axes in the graphs. Due to long gap between 1897 and 1941, no intermediary changes could be shown in the graphs. It can be seen that for the island of sizes larger than 5 ha comparatively steep changes can be seen in their sizes. Here, in Fig. 7 an example of review changes to the Fualafake Islet of Funafuti Atoll from 1987 to 2017 is shown depicting large changes in most of the parts which can be described as first accretion and then erosion to the northern part, erosion, and then accretion to the southern and south-eastern parts, and fusing with Paava in 1971.

Fig. 7. Changes in boundary for Fualafake Islet of Funafuti Atoll from different sources.

3.2. Driving Forces of Tuvalu Coastline Changes

Tuvalu experiences disadvantages of very small size, extreme isolation and fragmentation as well as very low height (most areas much below than 5 m above sea level), and thus become extremely vulnerable to different types of environmental shocks. Tuvalu is already experiencing increased frequency of extreme climatic events like storm surges, cyclones, hurricanes, tide surges, and flooding (Connell, 2003).The problems of episodic cyclones in Tuvalu are supposed to be intensified by SLR and thus will be reviewed here in relation to coastline dynamics (Connell, 2003).

3.2.1. Tropical Cyclones and Coastline Changes to Tuvalu

Cyclone can be agents causing sharp changes in an island’s coastline and its morphology (Connell, 2015). Webster et al. (2005) noticed increase in the frequency of categories 4 and 5 cyclones in the Southwest Pacific region from 1975 to 2004 (Rasmussen et al., 2009). The earliest research works on examining the effects of extreme storms on atoll and resulted sedimentation were done on Jaluit Atoll and Marshall Islands by McKee (1959) and Blumenstock (1961). Intense tropical cyclonic storms in low-latitude coral reef environments sometimes have the capability to construct large ridges by transporting reef-derived coarse sediment that eventually results in coastland accretion (Maragos et al., 1973). However, there are several prerequisites to develop a storm ridge, such as highly raised water levels on the storm passage, severe waves, ample supply of sediments, availability of suitable substrates for deposition, abundant time for clast excavation and transport, etc. For the formation of steep ridge, deposition of ridge materials with concurrent flooding is also essential. However, reef debris could not be redeposited as ridges on islets which frequently faced hurricanes. These dynamics results from complex relationship among storm actions, place of cyclone passage and effective fetch, sediment availability, storm characteristics including intensity and duration, etc. (Richmond and Morton, 2007).

David and Sweet (1904) identified storms as an important factor in islet development processes (McLean, 2014). Less-intense storms have the capacity for eroding shorelines and removing extensive sand accretions whereas extreme storms cause storm ridge formation. Except storms and cyclones no other natural process can suddenly transport significant amount of coral debris form its generation location in coral reefs fronts to its distant accumulation location creating more or less stable landforms (Connell, 2015). However, not every storm is found to be resulted in formation of coral-gravel storm ridges. Accretion or erosion trajectories of islets vary greatly with storm magnitude and soil composition of the islands (Kench et al., 2018a)

Alike other archipelagos in the Pacific, Tuvalu is periodically influenced by several strengths of cyclones that generate high waves(3 to 4+ m heights)in extreme cases (Rasmussen et al., 2009). The climate of Tuvalu is governed by the El Niño. Cyclone or storm usually occurs during the rainy season (November–April) (Ceccarelli, 2019). From 1940s to 1970s Tuvalu was affected by mean three cyclones per decade whereas later on it greatly increased to eight cyclones during next four decades (Nunn, 1990; Vavae and Epu, 2011). Thus, as influenced by tropical cyclones, from 1945 to 1990 several changes are evident in Tuvalu, particularly on the very small sandy islets (Connell, 2003; McLean and Hosking, 1991). However, Connell (2015) suspected fewer cyclones to Tuvalu in future, but with greater intensity. The major tropical cyclones during the last several decades are shown as timeline in Fig. 3 from which some with their devastating impacts are briefly summarized here. The most dangerous cyclone was Bebe hitting Funafuti in 1972. Such events could account for the predominant expansion of other islands made of mixed sand–gravel or gravel. Kench et al. (2018a) found that majority of sandy islets of Funafuti and Nukufetau located at the leeward northwest and northern rims of those atolls reduced in size mainly by periodic storms. In 1990 Vaitupu was stuck by a devastating cyclone. In a single year of 1997 hit by three tropical cyclones, Gavin, Hina and Keli, especially affected Niulakita and Nukulaelae. Scientists predict less frequent but stronger storms and cyclones in coming future (Chambers and Chambers, 2001; Connell, 2003). Funafuti is also subject to the actions of seasonal high waves and storms, however fortunately remained outside the central pathways of many strong cyclonic storms (Scoffin, 1993).

Tropical cyclone Bebe stroke Funafuti on 21 October 1972 (Maragos et al., 1973). It was a slow-moving but very strong storm (Category-3 storm at Saffir-Simpson Hurricane Wind Scale) that’s eye passed through Funafuti Atoll from east to west (Richmond and Morton, 2007). For that region it was the most studied tropical cyclone as that transported a large volume of coarse sediments to the Funafuti reef flat resulting formation of storm ridges. After formation, the rampart has retreated isletward or lagoonward along most of its length. The ramparts later expanded to the shorelines along the eastern rims of Funafuti (Maragos et al., 1973; Kench et al., 2018a). The leeward margins of the formed ridges were much elevated and steeper (Richmond and Morton, 2007). The bases of the ridges were in intertidal zone whereas their crests remain above the standard tide level. Thus, nearly continuous permanent 19 km long coral-gravel ridge of around 30-40 m (average 37 m) width and maximum 4 m height was formed along the southeastern reef platform around 30 m inward from the seaward edge of the reef (Baines and McLean, 1976a; Maragos et al., 1973; Richmond and Morton, 2007). The mean size of surface fragments was around 10 cm with largest storm block of 7 m diameter (Maragos et al., 1973). The deposits were much larger than those reported earlier. During the succeeding years waves shifted the coral ramparts inshore across the ocean reef flats that formed the eastern shores for most of the eastern islets(Baines and McLean, 1976a; 1976b). The deposits formed by Bebe were comprised of loosely composed cobble to boulder size coral rubble (Maragos et al., 1973; Baines and McLean, 1976a). The rampart materials were derived from deeper water offshore (Baines and McLean, 1976a). However, older and higher storm ridges were also found at the landward side of the Bebe ridge (David and Sweet, 1904; Dickinson, 1999). Successive surveys indicate lagoonward movement of a massive coral rubble rampart (Baines and McLean, 1976a; 1976b).

During March 1997 tropical cyclone Gavin produced extremely high waves in Funafuti Atoll which resulted in large morphological changes of lands. Sato et al. (2010) showed doubled lagoonal wave height than the mean nominal wave height through simulating winds and waves for the Gavin. Connell (2003) mentioned a drama, “Paradise Drowned: Tuvalu, The Disappearing Nation” which focused on several problems of Tuvalu like disappearing of one islet of Funafuti, greater incidence of cyclones and higher tides, increased flooding, and scarcity of potable water in Funafuti (Chambers and Chambers, 2007). Recently Category-5 tropical cyclone Pam hit Funafuti in March 2015. Though Funafuti escaped major damage from Pam, that caused considerable coastline changes to Nanumega, Nanumea, Niutao, Nukulaelae, Nui, Nukufetau, and Vaitupu due to strong storm surges and prolonged sea swells (Hisabayashi et al., 2018; North, 2015).

In the tropical regions storm climates and wave are intensifying with the passage of time (Chand et al., 2020; Meucci et al., 2020; Fellowes et al., 2022). In future, increase in storm intensity and wave height might supersede short-term constructive effects on coral reefs in case of limited sediment supply (Beetham and Kench, 2018; Bramante et al., 2020; Fellowes et al., 2022; Tebaldi et al., 2021).

3.2.2. Sea Level Rise and Coastal Change

The SLR is now a global phenomenon mainly caused by thermal expansion of sea water and addition of water to the oceans from melting ice sheets and glaciers as a result of global warming (Boretti, 2021; Eslamian et al., 2018). However, in general SLR usually results from combinations of multiple natural phenomena which individually may not be much extreme (Leonard et al., 2014). Extreme SLR greatly influences the coastal regions with inundation, cliff instability, beach erosion, etc.(DCC, 2009).By having very low height land areas Tuvalu is one of the small Pacific island countries most vulnerable to SLR where the mean SLR is almost four times to that of global average (Becker et al., 2012; Lazrus, 2012). Kench et al. (2018a) identified total rise of around 0.15+ m SLR from Funafuti tide gauge from 1971 to 2014 (average 3.9±0.4 mm/year). The SLR near Tuvalu is around 5 mm/year since 1993 (Fig. 8) greatly above than that of the global average (2.8–3.6 mm/year). It has already frequently experienced many king tide events causing elevated sea level, and thus results in inundation at the low elevation areas. The risk of such tidal flooding is anticipated to increase in response to SLR. These high tide effects are compounded by storms. Patel (2006) predicted that in the next 100 years Tuvalu could become uninhabitable due to SLR of 20–40 cm. Such SLRs are causing erosion in some areas of Tuvalu. Seawalls were being constructed and mangroves were planted to some extents in order to mitigate coastal erosion (Ceccarelli, 2019).

Fig. 8. Relative sea level (red dashed trendline) from March 1993 to June 2022 from tide gauge daily data at Funafuti (source: https://uhslc.soest.hawaii.edu/).

In spite of suspicion of very high SLR in Tuvalu no great changes are noticed by the researchers (Barnett and Adger, 2003; Connell, 1993; IPCC, 2001). Connell (2003) showed 0.9 mm mean rise in sea level per year from 1993 till 2002. The historical tide gauge record on Funafuti also doesn’t show much acceleration in SLR from 1978 to present (Fig. 8). Thus, scientists argued a discourse of drowning for Tuvalu is created by the media (Connell, 2003; Farbotko, 2005; Farbotko, 2010; Kelman, 2014). The people of Tuvalu are positioned by foreign actors to raise their voice regarding the discourse that the entire planet is under the threat of climate change crisis which is also backed by the government in order to draw attention to the need for international responses get international attention (Farbotko and Lazrus, 2012). Mortreux and Barnett (2009) and Campbell and Warrick (2014) postulated that the impacts of climate change could not be identified as the major driver for the current or future migration of the residents of Funafuti. However, it could not be overlooked that SLR around Tuvalu are the highest in the Pacific region which is still less than those of world oceans during the last century as a result of tectonic plates shifting, changes in sea currents, volcanic events, alteration in long term climate patterns, global warming, etc. (Australian Bureau of Meteorology and CSIRO, 2011; Connell, 2003). In spite of variations in short-term tidal levels since 1978, that may be increased by high spring tides in response to El Niño events as observed at Nanumea in 1993 (Chambers and Chambers, 2001). Similarly, in 1997–1998 the El Niño event have resulted in SLR of more than 35 cm which was much higher than the normal annual fluctuations are 10 cm at Tuvalu (Connell, 2003).

According to Kench et al. (2018a) no uniform and widespread erosion in Tuvalu indicates that observed island dynamics are caused by a set of higher-frequency processes rather than sole contribution by SLR. Analysis of the 30-year wave hindcast data showed no substantial change since 1979 indicating its unlikeliness to be the main causal agent for the observed island adjustments. However, SLR is supposed to enhance remobilization of island shorelines and sediment transfer through greater transfer of wave energy across reef surfaces. Evidence indicates that SLR has influenced the coastal dynamics of atoll rim islets throughout Tuvalu which resulted in net lagoonward migration of beaches. From observations Kench et al. (2018a) depicted that over the past four decades (1970s–2000s) islands could successfully persist under SLR of 3.9±0.4 mm/year.

In spite of high SLR most of the low-lying coral reef islands of the world are still not diminished, rather their surface areas are found either stable or expanding throughout the twenty-first century, and none of islands larger than 10 ha decreased in size since 1980s (Boretti, 2021; Duvat, 2019; Holdaway et al., 2021; Sengupta et al., 2021). From 1999 to 2017 Tuvalu land area increased by around 7% (Holdaway et al., 2021). However, it is still uncertain whether the low lying Tuvalu can adapt to changing climates, and on what timelines (Fellowes et al., 2022). With the rise of sea level, the future stability of Tuvalu land areas will greatly depend upon reef integrity, carbonate productivity, ample sediment supply, wave dissipation, storm events, flooding, island overtopping, human modification, etc. (Baines and McLean, 1976; Beetham and Kench, 2018; Fellowes et al., 2022; Maragos et al., 1973; McLean and Hosking, 1991; Winter et al., 2020). Shallow reef atolls are most likely to be subjected to greater wave-induced run-up and flooding due to SLR (Storlazzi et al., 2015). Many atoll islands are supposed to be flooded annually causing salinization of freshwater resources which in turn will cause its inhabitants to leave their home in decades long before their suspected drowning after centuries (Storlazzi et al., 2015).

Coral reef islands are generally constructed from unconsolidated sediments after reworking by waves and currents (Perry et al., 2011). Being a living object coral grows that supply new sediments to the surrounding environments (Boretti, 2021; Perry et al., 2011). Ample sediment supply can result in formation of sand aprons, coral islands, and shallow lagoons through infilling of the backreef environments (Fellowes et al., 2022; Mandlier and Kench, 2012; Ortiz and Ashton, 2019; Rankey, 2021). Sediments of low lying coral islands can be formed of skeletons and shell of calcifying organisms, chiefly corals, large benthic foraminifera, calcareous algae (e.g., Halimeda), molluscs, echinoderms, etc. (Dawson and Smithers, 2014; Dee et al., 2020; Fellowes et al., 2022). These coral reefs usually generate from the secretion of calcium carbonate skeleton from coral polyps as a result of symbiosis between those polyps and some specific algae (Fellowes et al., 2022; Hoegh-Guldberg et al., 2007). In the shallow lagoon of Fongafale foraminifera is the major component(41%) of sediments, followed by Halimeda (24%), coral and calcareous red algal debris (23%), molluscs (9%), and other organisms (3%), whereas in deeper water Halimeda replaces the foraminifera (Collen and Garton, 2004; Fellowes et al., 2022). Notably, foraminifera are single cell protists which take part in sediment formation.

Coral sand dominated islands are prone to more morphological changes under most near-future ecological change scenarios compared to those dominated by the tropical benthic genus of green algae, Halimeda (Perry et al., 2011). However, islands composed mostly of benthic foraminifera exhibited wide ranges of resilience depending on the specific combination of ecological disturbances encountered (Perry et al., 2011). On most islets of Fongafale large positional adjustments occurred as a result of reworking of island margins, especially from fresh sediment addition to the island reservoir (Kench et al., 2015). Funafala and Luamotu Islets on the eastern atoll rim of Funafuti have accreted through generation of new gravel deposits by cyclones and their subsequent alongshore transportation (Baines and McLean, 1976; Kench et al., 2015; Maragos et al., 1973). Reversal of the erosional trend on other eastern rim islands of Funafuti Atoll is also caused by large volume of rubble as generated during several cyclonic events (Kench et al., 2015). Between 1897 and 1971, Fatato, Funagongo, and Funamanu eroded, later on which accreted with the net minor change (Kench et al., 2015). The islets which mainly showed shrinking trends were typically smaller than 10 ha and were sandy in soil composition located on the west and north rims of Funafuti Atoll (Fellowes et al., 2022; Kench et al., 2015). In case of net decreased islands, their land migration or rotation (e.g.,Tepuka) occurred in net deficit due to no new net sediment input, and adjustments relied on greater reworking of the original core (Kench et al., 2015). These findings indicate the importance of sediment formation for the purpose of stability, growth, and persistence of reef islands as well as reflect the difference in the responses of islands to dynamic change (Kench et al., 2015; Perry et al., 2011). In particular, it can be said that the gravel islands have more stable resistant core around which further accumulation of sediments may occur, whereas sand islands are comparatively more prone to dynamic changes and migration including reworking of island core sediments which rely on regular supply of sand-size materials to maintain their size (Kench et al., 2015). The findings of Masselink et al. (2020) show that the reef islets which are mainly composed of gravel materials, for instance Fongafale is a morphodynamically resilient landforms that have capability to evolve in spite of rise in sea level by accreting to sustain their positive freeboard whereas retreat lagoonward as an effect of wave overtopping processes that transfers sediment from the beachface to the island surface.

By a physical modelling Masselink et al. (2020) showed vertical accretion of the crest and lagoonward retreat by 25 m during 0.5 m increase in the sea level. They stated wave overtopping as the physical mechanism of island adjustment as well as the primary mechanism for vertical island accretion which effectively transport sediment from the nearshore and beachface to the island crest and surface. The net result is a morphodynamic rollover response alike that of gravel barrier systems as identified by Orford et al.(1995) and other analytical modeling studies on atoll islands (Cowell and Kench, 2001; Masselink et al., 2020).

Various literatures show land elevation transects or ground elevation profiles in several areas of Funafuti Atoll, especially Fongafale, depicting large areas under less than 3 m and 1.5 m nearby the central areas of the islets, especially in case of the largest Fongafale Islet (Lewis, 2021; Lin et al., 2015; Henning et al., 2017; Kench et al., 2014; Kench et al., 2018b). Kayanne et al. (2021) showed that most of the areas in the central areas of Fongafale Islet will go under water at 10 cm SLR, but outer boundaries might remain above that increased sea level. However, due to storm effect erosion may occur, and thus parts of boundary areas above 10 cm like storm ridge, seawall, etc. may erode and thus couldn’t act as defense line by making free connection of the drowned areas with the open sea (JICA, 2011a; 2011b). In spite of such considerable suspected vulnerabilities, in Tuvalu till now direct physical interventions to modify coastal processes for the purpose of protecting islands are still scanty (Kench et al., 2018a). Some small-scale land reclamation activities were taken during the last few decades which include minimal direct shoreline modification in Fongafale till 2014, coastal protection work along a short length of Savave shoreline in Nukufetau Atoll, dredging of channels across some reef flats for boat access, building boat-landing facilities, etc. which had no significant direct impact on coastal changes at and around those construction sites (Kench et al., 2018a). Therefore, land reclamation process following the worldwide examples are necessary for Tuvalu (Chee et al., 2017; Holdaway et al., 2021; Sengupta et al., 2018; Tian et al., 2016). Otherwise, due to ever increasing SLR, storm, prolonged flooding, high wave situation, very low elevation areas might go under water or at least will be inhabitable due to lack of potable water caused by trapping of salt water in the inland waterbodies and salinization of groundwater lenses through intrusion of excessive saline water (Chui and Terry, 2012; Nakada et al., 2012; Connell, 2003).

4. Conclusions

There are several research works on both long- and short-term coastal changes of Tuvalu, especially on Funafuti Atoll. Changesin the coastal areas throughout Tuvalu are chiefly driven by environmental processes, mostly by tropical cyclones. Island area changes are not major concern for comparatively larger reef flats. Major documented changes include increased mobility of atoll reef rim islands with erosion of smaller sand islands (10 ha), whereas RP islands were found to be comparatively much stable. However, for small islets changes in land area varies a lot as well as variably changing locations of erosion and accretion which affect those islets a lot due to their small sizes. Though, from long ago it was suspected that many islands might lose much of its area or fully disappear due to ever increasing SLR for Tuvalu, afterward studies evident no such outcomes except disappearance of Vasafua Islet. On the contrary, many islets’ total land area were increased due to strong tropical cyclones except decrease in land area for some small islets. However, those cyclones caused disruption of mangrove area and no effective land reclamation or erosion prevention measures were evident. Moreover, as the increase in land are by strong tropical cyclones have no specific trend of land accretion, scientists couldn’t confirm such accretion in near future which still arise the uncertainty of not drowning the landmasses of Tuvalu by increasing trend of SLR in coming future. Therefore, in order to overcome such suspected effect of losing grounds activities for lagoonward accretion of land could be possible to increase the land areas following similar global examples after thorough research works on its effects to the water quality and biota of lagoon.