Long-Term Analysis of Tropical Cyclones in the Southwest Pacific and Influences on Tuvalu from 2000 to 2021

Abstract

Tropical cyclones frequently occur in the Southwest Pacific Ocean and are considered one of the driving forces for coastal alterations. Therefore, this study investigates the frequency and intensity of tropical cyclonesfrom 2000 to 2021 and their influence on the surface winds and wave conditions around the atoll nation Tuvalu. Cyclone best-track and ERA5 single-level reanalysis data are utilized to analyze the condition of the surface winds, significant wave heights, mean wave direction, and mean wave period. Additionally, the scatterometer-derived wind information was employed to compare wind conditions with the ERA5 data. On average, nine cyclones per year originated here, and the frequency increased to 11 cyclones during the last three years while the intensity decreased by 25 m/s (maximum sustained wind speed). Besides, a total of 14 cyclones were observed around Tuvalu during the period from 2015 to 2021, which showed an increase of 3 cyclones compared to the preceding period of 2001 to 2007. During cyclones, the significant wave height reached the highest 4.8 m near Tuvalu, and the waves propagated in the east-southeast direction during most of the cyclone events (52%). In addition, prolonged swells with a mean wave period of 7 to 11 seconds were generated in the vicinity of Tuvalu, for which coastal alteration can occur. After this preliminary analysis, it was found that the waves generated by cyclones have a crucial impact in altering the coastal area of Tuvalu. In the future, remotely sensed high-resolution satellite data with this wave information will be used to find out the degree of alterations that happened in the coastal area of Tuvalu before and after the cyclone events.

1. Introduction

Several natural disasters, including excessive rainfall, floods, earthquakes, droughts, storms, cyclones, etc., are continuously occurring around the globe and have an impact on the land area and its population. Among these disasters, cyclones are considered the costliest extreme natural event as they have a very devastating impact on life and property (Deo et al., 2022). During a cyclone, multiple extreme situations arise, which include strong winds, storm surges, extreme rainfall, and flooding that concludes with several fatalities and substantial damage to the community, human assets, infrastructure, and economic sectors. This natural disaster usually occurs at a particular time of the year, and that period is widely denoted as the cyclone season (Terry, 2007). The development of cyclones occurs due to the elevated sea surface temperature, which leads to the generation of low-pressure centers that are concentrated on a single area above the surface of the ocean. This kind of natural disaster is confined to particular regions across the planet; however, the recognition of this event depends on the ocean area within which it occurs. These events are known as hurricanes in the western North Atlantic and the Caribbean, typhoons in the western North Pacific Ocean and China Sea, and tropical cyclones in the western South Pacific and Indian Ocean (Terry, 2007). The highest number of tropical cyclones occurs during the periods of summer and autumn; however, cyclones with the most intensity are generated in autumn compared to summer (Wang et al., 2023).

The intensity of tropical cyclones is categorized into seven classes by considering their maximum sustained wind speed in the center based on the Saffir-Simpson hurricane wind scale (Simpson, 1974; Delaporte et al., 2022). Cyclones with less than 17 m/s maximum sustained wind speed are known as tropical depressions, and when the wind speed is increased to 32 m/s, the cyclone belongs to the tropical storm category (National Oceanic and Atmospheric Administration, 2023). Tropical cyclones with maximum sustained wind speeds ranging between 33 and 42 m/s are known as Category 1 cyclones that produce strong winds. The maximum wind speed of Category 2 cyclones remains between 43 and 49 m/s, which generates very strong winds that cause extensive damage to human property (National Oceanic and Atmospheric Administration, 2023). More devastating damage happened to infrastructure and vegetation due to the Category 3 cyclone, and the maximum sustained wind speed remained between 50 and 58 m/s (National Oceanic and Atmospheric Administration, 2023). Cyclones that belong to Category 4 (sustained wind speed between 58 and 70 m/s) and 5 (sustained wind speed >70 m/s) cause catastrophic damage to the population as well as infrastructure (National Oceanic and Atmospheric Administration, 2023). The generated strong winds and extreme height of waves pose a significant danger to the low-lying atoll nations (Avia, 2020). The nations located in the Southwest and Northwest Pacific Oceans are frequently impacted by strong winds, storm surges, extreme precipitation, and flooding during cyclone events. Especially the island nations in the Southwest Pacific Ocean become more vulnerable to tropical cyclone events as they possess an unfavorable and smaller land area due to the combination of low-lying atolls and reef islands (Magee et al., 2016). Due to the occurrence of tropical cyclone Kina (December 1992–January 1993), 25 fatalities took place in Fiji, and it also caused huge damage worth USD 120 million (Salinger and Lefale, 2005). Topical cyclone Heta with an intensity of Category 5 occurred in January 2004 and caused significant damage to Samoa (USD 226 million loss) that included the destruction of infrastructure, crops, and human property (Salinger and Lefale, 2005).

In the previous work (Harun-Al-Rashid and Yang, 2023), the coastal changes of Tuvalu, one of the world’s smallest nations situated in the Southwest Pacific Ocean, from 1897 to 2015 concerning tropical cyclones and sea level rise were depicted. In that work, tropical cyclones were identified as one of the driving forces responsible for the alteration of Tuvalu’s coastal region as different categories of tropical cyclones passed across or near the islets of Tuvalu in previous years. Cyclones with less intensity play a crucial role in the erosion of coastlines by extensively removing sediments, while high-intensity cyclones lead to the formation of storm ridges (Harun-Al-Rashid and Yang, 2023). For example, tropical cyclone Bebe (October 21, 1972), a Category 3 storm that passed across the Funafuti atoll of Tuvalu, transported an extensive volume of sediments to the reef part of Funafuti, and as a result, a storm ridge was formed (Maragos et al., 1973; Kench et al., 2018). Besides, a high storm surge above 4 m from the mean high water level was created during the cyclone Bebe event, which caused flooding to land and destruction of infrastructure as well as loss of human life (Maragos et al., 1973).

Additionally, considerable alterations in the coastline happened to the islets of Tuvalu, excluding the Funafuti atoll, during the occurrence of tropical cyclone Pam due to sturdy storm surges and sea swells (Hisabayashi et al., 2018). Thus, the surrounding environmental conditions (wind, the height of sea waves, etc.) of Tuvalu are influenced during the passing period of cyclones and lead to significant changes in the coastal morphology. Additionally, remotely sensed satellite images are generally used extensively to detect and analyze the coastal alterations, and previously, different resolution satellite data were used to quantify the coastal alteration in Funafuti (Hisabayashi et al., 2018; Kench et al., 2015). These studies only analyzed the spatiotemporal changes that happened in Tuvalu especially in Funafuti atoll during the strong intensity cyclone event such as cyclone Pam in 2015 (Hisabayashi et al., 2018) where the influence of cyclone on winds and wave condition were not described. However, the cyclone frequency is changing with time and therefore, it is required to comprehend the long-term circumstances of cyclone events and their influence on the surface wind and wave condition around the atoll island.

Therefore, in this study, the occurrence of tropical cyclones was investigated in the Southwest Pacific region during the period 2000–2021. Then, emphasis was given to the cyclones that passed in the vicinity of Tuvalu during the investigation period to analyze their frequency and intensity. Additionally, the conditions of surface winds, significant wave heights, mean wave direction, and mean wave period were also evaluated around Tuvalu during the cyclone events, and wind vectors from scatterometers were used to compare the wind condition. The outcome of this study will serve as the basis for analyzing the extent of coastal changes that occurred in Tuvalu due to the cyclone events.

2. Materials and Methods

Tropical cyclone best-track data for the Pacific region from 2000 to 2021 was obtained from the Naval Oceanography Portal (https://www.metoc.navy.mil/jtwc/jtwc.html?cyclone). An archive of tropical cyclone track data, referred to as best-tracks, is maintained by the Joint Typhoon Warning Center (JTWC), with each best-track file containing information about the location of the tropical cyclone’s center and its intensities (maximum sustained 10-m wind speed) at 6-hour intervals (Chu and Sampson, 2002). To create the JTWC archive, data from three sources are used, namely the National Climate Data Center database, a Fleet Numerical Meteorology and Oceanography Center database, and the Automated Tropical Cyclone Forecasting System database (Sampson and Schrader, 2000). The positions of tropical cyclones are established every 6 hours after the cyclones have dissipated, and these positions can deviate from the “working best-track” positions in the tropical cyclone warnings by up to 120 nautical miles (Chu and Sampson, 2002). The JTWC archive data are generally rectified by implementing several steps (Chu and Sampson, 2002). Firstly, a visual comparison is conducted between the JTWC archive and the automated tropical cyclone forecasting system database. Subsequently, discrepancies are checked by cross-referencing them against tracks sourced from other databases. The culmination of these efforts is resulting in the formulation of recommendations for refining the JTWC archive. Afterward, the final correction is conducted in the archive data (Chu and Sampson, 2002). Fig. 1 depicts the track of tropical cyclones along with their intensity that occurred in the Pacific area in the previous two decades. In this study, two Groups were created to divide the different intensities of cyclones, while cyclones belong to the tropical depression and tropical storm categories as expressed in their original definition. According to the Saffir-Simpson wind scale, cyclones that fell into Categories 1, 2, and 3 were denoted as Group 1, whereas Group 2 consisted of the strongest intensity (Categories 4 and 5) cyclones.

Fig. 1. Trajectories of cyclones originated in the Pacific region from 2000 to 2021. Trajectory colors represent the maximum lifetime intensity of cyclones. TD: tropical depression, TS: tropical storm, Group 1: cyclone Categories 1, 2, and 3, Group 2: cyclone Categories 4 and 5.

Tuvalu, the 4th-smallest nation worldwide and located in the Southwest part of the Pacific Ocean, was considered the study area to evaluate the cyclone’s influence. It consists of six true atolls and three reef islands: Nanumea, Nui, Vaitupu, Nukufetau, Funafuti, Nukulaelae, Nanumaga, Niutao, and Niulakita (Milan et al., 2016). The total land area and the exclusive economic zone of Tuvalu are 26 km² and 900,000 km², respectively (Ceccarelli, 2019). The coastlines of Tuvalu’s islets are long and directly exposed to the sea. The highest elevation of all islets is barely 4 m above sea level, which makes Tuvalu highly vulnerable to extreme environmental events like storm surges, sea level rise (Ceccarelli, 2019), precipitation, and flooding. The islets of Tuvalu are situated near the equator, and hence, they have a lower threat from tropical cyclone events than other Pacific Island countries such as the Cook Islands, Fiji, Kiribati, Samoa, Papua New Guinea, etc., which are directly hit by cyclones. Although the threat is lower, during the passing period of a tropical cyclone, strong winds and waves with extreme heights are generated, which can contribute significantly to the alterations of the coastal region (Avia, 2020). A region extending from 5°S to 15°S latitude and 170°E to 175°W longitude was specified, which covers Tuvalu (Fig. 2), and cyclones were listed that passed within this specified region. The cyclone’s intensity (maximum sustained wind speed) was assessed during its existence within the selected area, and the surface winds and wave conditions within this region were also evaluated. From the Funafuti atoll (8.5°S, 179.2°E), the distance of the specified region’s boundary was 1,000 km in the west, 635 km in the east, 389 km in the north, and 720 km in the south, respectively.

Fig. 2. Location of Tuvalu along with the bathymetry map and boundary of the specified region (yellow dashed rectangle). Black polygons indicate the islets of Tuvalu.

The ERA5 dataset was acquired from the European Union’s Copernicus Climate Change Service website (https://cds.climate.copernicus.eu/cdsapp#!/dataset/reanalysis-era5-single-levels?tab=form). The European Centre for Medium-Range Weather Forecasts produced the ERA5 data, the fifth generation of atmospheric reanalysis, that spans the years 1950 through the present (Bell et al., 2021). A major improvement over its predecessor, ERA-Interim (78 km horizontal resolution), ERA5 offers hourly estimates of the world’s atmosphere, ocean waves, and land surface at a horizontal resolution of 31 km. Reanalyses have a wide range of applications, spanning from intergovernmental assessments of global climate change at one extreme to specific and unique use cases that require accurate representations of local weather at the other (Bell et al., 2021). In this study, the ERA5 hourly single-level reanalysis dataset was used to analyze the surface winds, significant wave heights, mean wave direction, and mean wave period around Tuvalu during cyclone events. The spatial resolution of the used ERA5 data for the surface winds and waves were 0.25° × 0.25° and 0.5° × 0.5°, respectively. Two distinct locations in the vicinity of Tuvalu were chosen as representative sites for analyzing the wave conditions during the cyclone event due to the similarity in the wave situation around Tuvalu. Furthermore, since a majority of prior investigations on coastal changes have centered on the Funafuti atoll, these two locations were specifically selected from the eastern and western sides of Funafuti, respectively, to facilitate their applicability in future research. The coordinates of the first location were 8.5°S, 179°E, and the second location was 8.5°S, 179.5°E.

Additionally, three Moderate-resolution Imaging Spectroradiometer (MODIS) images, which cover a wide area were acquired during the cyclone occurrence period to spatially visualize the specified cyclone. Besides, sea surface wind vectors over the global ocean observed by the Advanced SCATterometer (ASCAT) satellite were used in this study to compare the wind condition with the ERA5 reanalysis data. ASCAT is a sensor that was integrated into the first Meteorological Operational Polar Satellites launched by the European Space Agency on October 19, 2006 and is functioned by the European organization (Kako et al., 2011). ASCAT employs real-aperture radar technology using vertically polarized antennas. Operating within the C-band at 5.225 GHz, the antennas generate beams that cover around 550 km swaths, spaced around 700 km apart as the satellite orbits. Within these swaths, radar backscatter from the sea surface is measured in 25-km grids, with additional measurements at 12.5-km spacing included in the datasets (Kako et al., 2011). In this study, a scatterometer wind product with 12.5 km horizontal resolution on March 11, 2015, was downloaded from the EUMETSAT website (https://data.eumetsat.int/data/map/EO:EUM:DAT:METOP:OSI-104).

3. Results

3.1. Occurrence of Tropical Cyclones in the Southwest Pacific

From 2000 to 2021, there were 561 cyclone events in the Pacific region, including 211 tropical cyclones that originated in the Southwest Pacific and 350 cyclones that developed in the Northwest Pacific Ocean. The frequency of tropical cyclones over the Pacific area each year is shown in Fig. 3. Compared to the Southwest Pacific region which had an average of 9 cyclones per year, the Northwest Pacific experienced more cyclone events every year (15 on average). In the Southwest Pacific, the highest number of 12 cyclones was developed in 2003, while in the Northwest Pacific, a maximum of 25 cyclones originated in one year (2004). The cyclone generation frequency in the Southwest Pacific increased from 8 cyclones per year on average between 2000 and 2006 to 10 cyclones per year between 2007 and 2011 (Fig. 3). The frequency decreased over the next two years; however, after 2014, it subsequently approached an average of 10 cyclones annually. In contrast, between 2000 and 2006, an average of 18 cyclones per year developed in the Northwest Pacific, and over the following seven years, that frequency decreased to 13 cyclones per year (Fig. 3). After that, from 2014 to 2021, there were 16 cyclones on average every year.

Fig. 3. Annual frequency of tropical cyclones in the Southwest and Northwest Pacific Ocean from 2000 to 2021.

The source positions of cyclones across the Pacific Ocean from 2000 to 2021 are depicted in Fig. 4. It was found that the cyclone origin point in the Southwest Pacific was scattered over a vast area that expanded between 8°S–20°S and 130°E–150°W, whereas the majority of the cyclones in the Northwest Pacific originated between 3°N–18°N and 130°E–175°E (Fig. 4). As a consequence, countries that are between 10 and 30 degrees from the equator are particularly vulnerable to the threat posed by cyclones (Arthur and Woolf, 2013). In the case of the Southwest Pacific, Fiji, Kiribati, the Solomon Islands, Vanuatu, Tonga, Samoa, the Cook Islands, Tuvalu, Niue, and the Mashall Islands are a few of these nations. Additionally, in the Southwest Pacific, a majority of over 75% of tropical cyclone generation points were located between 150°E and 175°W, while in the Northwest Pacific, more than 85% of cyclone origin points were dispersed across the range of 138°E to 165°E (Fig. 4). Moreover, it was found that, tropical storms and Group 1 cyclones were more prevalent in the Southwest Pacific while Group 2 cyclones were more frequently developed in the Northwest Pacific region (Fig. 4). Besides, throughout this 22-year period, a substantial number of Group 2 cyclone origin locations were limited to the area spanning from 165°E to 175°W in the Southwest Pacific Ocean (Fig. 4). Conversely, in the Northwest Pacific, a notable majority of Group 2 cyclones originated within the range of 138°E and 165°E (Fig. 4). From 2000 to 2021, 49% of tropical strom events occurred in the Southwest Pacific, compared to 22% of storms in the Northwest Pacific (Fig. 4). As a result, the Northwest Pacific region experienced a larger percentage (41%) of strong intensity cyclone (Group 2) than the Southwest Pacific (17%) (Fig. 4). Moreover, the proportion of Group 1 cyclone development in the Southwest Pacific region was relatively higher at 30% compared to the Northwest Pacific, which was at 27% (Fig. 4).

Fig. 4. Visualization of the source location of different intensity tropical cyclones in the Pacific Ocean from 2000 to 2021. The shape and color of the source points indicate the maximum lifetime intensity of cyclones. TD: tropical depression, TS: tropical storm, Group 1: cyclone Categories 1, 2, and 3, and Group 2: cyclone Categories 4 and 5.

Fig. 5(a) depicts the general trends of different intensity cyclones (tropical storm, Group 1, and Group 2) in the Southwest and Northwest Pacific from 2000 to 2021. Tropical storms and Group 2 cyclones, in particular, were more prone to develop in the Southwest and Northwest Pacific areas, respectively (Fig. 5a). The line graphs of tropical storms in the Southwest Pacific and Group 2 cyclones in the Northwest Pacific, consequently, exhibit a distinct pattern of occurrence. In the Southwest Pacific, four storms formed in 2000, which increased to seven storms in 2021 (Fig. 5b), and there was a consistent change in the frequency of storm generation during this period (Fig. 5b). A similar pattern was observed in the Northwest Pacific, where only three storms originated in 2000 and then seven storms were recorded in 2021 (Fig. 5c). The line highlights that Group 2 cyclones in the Northwest Pacific displayed an irregular variation during the overall period (Fig. 5c), even though the generation frequency of Group 2 cyclones in the Northwest Pacific was higher than in the Southwest Pacific. Both in the Southwest and Northwest Pacific regions, Group 1 cyclone development was less regular and inconsistent resulting in an ambiguous trend of Group 1 cyclone occurrence (Fig. 5a). Overall, during the 22 years, tropical storm development increased more frequently in both the Southwest and Northwest Pacific (Figs. 5b, c). Additionally, Group 2 cyclone generation in the Northwest Pacific region continued during those two decades, despite a minor drop in frequency (Fig. 5c).

Fig. 5. Annual variation (a) of different intensity tropical cyclones in the Southwest and Northwest Pacific from 2000 to 2021. The overall trend (dashed black line) of TS and Group 2 cyclones in the Southwest and Northwest Pacific is illustrated by (b) and (c), respectively. TS: tropical storm, Group 1: cyclone Categories 1, 2, and 3, Group 2: cyclone Categories 4 and 5, WSP: Western South Pacific, and WNP: Western North Pacific.

3.2. Frequency and Intensity of Tropical Cyclones around Tuvalu

A total of 31 tropical storms passed through the designated regions (Fig. 2), which covers Tuvalu, out of 211 tropical cyclones that developed in the Southwest Pacific between 2000 and 2021. Table 1 demonstrates that the intensities of the tropical cyclones throughout their passage around Tuvalu were not equal to their maximum lifetime intensities. A tropical cyclone moves across numerous places between its generation and extinction period, and its intensity changes over time (Wang et al., 2023). It was found that the intensity of 64% of cyclones was altered during their passing period within the specified zone. The majority of cyclones passed around Tuvalu during the initial phases of their development when their maximum sustained wind speed was not as strong as their highest sustained wind speed throughout their entire lifetime.

Table 1. The description of the cyclones passed around Tuvalu from 2000 to 2021

As cyclones move, the sustained wind speed progressively increases with them, and as they intensify from a tropical depression to a storm to a severe tropical cyclone, the weather is significantly altered (Terry, 2007). Moreover, it was found that during their existence around Tuvalu, 13 cyclones attained intensity levels consistent with tropical depression (Table 1). The second-highest number of cyclones (11) fell into the category of tropical storm intensity (Table 2).

The tracks of tropical cyclones that passed near Tuvalu from 2000 to 2021 are depicted in Fig. 6. The overall frequency of cyclones passing near Tuvalu increased throughout this time, indicating a rise in cyclone activity around Tuvalu. However, there were noticeable fluctuations in the strength of these cyclones. Between 2000 to 2021, there was an obvious decrease in the development or passage of cyclones with strong intensity near Tuvalu (Fig. 6).

Fig. 6. Trajectories and source points of tropical cyclones passed around Tuvalu from 2000 to 2021. The yellow dashed box indicates the specified area boundary (Fig. 1). Trajectory color represents the cyclone intensity and different shapes indicate the source point of cyclones. Black polygons indicate the islets of Tuvalu. TD: tropical depression, TS: tropical storm, Group 1: cyclone Categories 1, 2, and 3, and Group 2: cyclone Categories 4 and 5.

From January 2001 to April 2007, a total of 11 cyclones passed (Fig. 6a) within the specified region whereas only 6 cyclones passed from November 2007 to April 2013 (Fig. 6b). The highest number of cyclones (14) were passed in the last 7 years (Fig. 6c).

3.3. Wind and Ocean Wave Conditions around Tuvalu during Cyclone

Tropical cyclones have an impact on the wind and wave conditions, and as a result, during their passage, the wind speed and wave states near Tuvalu were influenced. The wind and wave conditions surrounding Tuvalu during the cyclone events are described in this section. In this study, the majority of the cyclones (68%) were found to move Southward (Fig. 6), and their influence on the wind and waves increased as they approached the open ocean. Table 2 depicts the information on the wave condition in the two representative locations during the existence of cyclones within the specified region. Tropical cyclones that produced ≥2 m significant wave height were considered in this case.

Table 2. Description of the wave conditions during tropical cyclone events in the vicinity of Tuvalu

The average maximum sustained wind speed of cyclones was found to be 29 m/s (Table 2) while they existed within the specified region, it gradually increased along with their movement. Therefore, the wind speed in the vicinity of Tuvalu was around or greater than 10 m/s during the cyclone events. The highest significant wave height was found at 4.8 m around Tuvalu during the cyclone events (Table 2). The highest elevation of Tuvalu is less than 5 m, which makes the islets vulnerable to flooding (Lin et al., 2014), and during cyclones, the extreme height of waves causes coastal alterations (Connell, 2003). After analyzing the mean wave direction, it was found that the waves propagated in the east-southeast direction in most of the cases (52%) (Table 2). On the other hand, in the case of 30% of cyclone events, the waves propagated in the south-southwest direction (Table 2). Because it has been seen that most of the cyclones passed on the east side of the islets of Tuvalu (Fig. 6), the storm surge generated by cyclones can hit the outer area of the islets situated on the east side (Taupo and Noy, 2017). As a result, an assumption can be made that the eastern part of Tuvalu’s islets is significantly influenced by the waves. After examining wind speed and wave parameters, which include significant wave height, mean wave period, and wave direction, variations were noted based on cyclone intensity and their distance of occurrence. A majority of cyclones passed through the vicinity of Tuvalu during their developmental stage when their intensity was relatively weaker. As they advanced, they gained strength and moved farther from Tuvalu, resulting in varying effects. Table 2 highlights the contrast in conditions: During cyclone Ami with a maximum sustained wind speed of 38 m/s, the significant wave height reached 3 m, and the mean wave period ranged between 7.9 and 8.2 seconds. In contrast, during cyclone Percy, with a maximum sustained wind speed of 28 m/s (approximately 26% lower than cyclone Ami), the significant wave height around Tuvalu ranged from 3.6 to 4 m, and the mean wave period was 9 seconds. This variance can be attributed to the differing locations of their occurrences. Despite cyclone Ami having a higher intensity than cyclone Percy, the proximity of cyclone Percy to Tuvalu at that time contributed to its impact (Fig. 6a). Furthermore considering cyclone Pam, originated at a greater distance (>1,000 km) from Tuvalu compared to cyclone Ami and cyclone Percy (Fig. 6), there was a significant elevation in wave height to 4.7 m, coupled with an extended wave period exceeding 11.5 seconds (Table 2). This elevation resulted from cyclone Pam’s higher intensity, standing around 21 m/s and 31 m/s above the intensities of cyclone Ami and cyclone Percy, respectively. Additionally, wave direction correlated with cyclone trajectory: during cyclone Ami, originating on Tuvalu’s western side and progressing southward (Fig. 6a), waves propagated east-southeast. Conversely, during cyclone Kerry’s development on the eastern side of Tuvalu and subsequent southeastward movement, wave propagation shifted to a southwestward direction. Moreover, long waves were generated around Tuvalu during the cyclone events, and it was found that the mean wave period varied from 7 to 11.5 seconds (Table 2). Due to these prolonged swells, coastal alteration in Tuvalu’s islets can occur, and the alterations include both erosion and accretion (Hisabayashi et al., 2018; Kench et al., 2015; Sierra and Casas-Prat, 2014).

This study presents data from three distinct cyclone events to provide a comprehensive description of wind and wave conditions in the vicinity of Tuvalu, while also describing wind and wave conditions during non-cyclone periods (Fig. 7).

Fig. 7. Wind and wave conditions around Tuvalu without cyclone events. (a), (b), and (c) indicate the condition before the cyclones Ami, Pam, and Bina, respectively. The black and cyan arrow indicates the wind and wave direction, respectively. Blue polygons indicate the islets of Tuvalu.

Tropical cyclone Ami, classified as a high-intensity (Group 1) cyclone, originated on January 9, 2003, from a location 372 km away from the Funafuti atoll. Fig. 8 illustrates the cyclone’s track (Fig. 8a) along with the wind and wave conditions observed on January 12, 2003, at 15:00 UTC (Fig. 8b), when the sustained wind speed reached 38 m/s (Fig. 8a; denoted by the maroon dot). During this period, the wind speed around Tuvalu exceeded 10 m/s and it flowed in the east direction. From Table 2. it can be seen that the significant wave height was 3 m at both locations (Fig. 8; green and black dots) and the waves propagated in the east-southeast direction. Moreover, the mean wave period ranged between 7.5 and 8.2 seconds. On the contrary, three days before this cyclone event (Fig. 7a), the surface wind speed was around 5 m/s (eastward direction) and the significant wave height was 1.7 m at both locations while the direction of wave propagation was also different (southwest direction). Besides, the mean wave period was found 9 seconds on January 6, 2003.

Fig. 8. Wind and wave distribution map. (a) True color MODIS image (RGB channels: bands 1, 4, and 3) of cyclone Ami on January 11, 2003. (b) Wind and wave conditions within the specified region (yellow dashed box) during the occurrence period (maroon dot) of cyclone Ami. Green and black dots indicate the first (8.5°S, 179°E) and second (8.5°S, 179.5°E) locations, respectively. Blue polygons indicate the islets of Tuvalu.

Tropical cyclone Pam, displaying the characteristics of a Group 2 cyclone, originated on March 6, 2015, at a distance of 1,139 km from the Funafuti atoll. The track of cyclone Pam and the wind and wave conditions recorded on March 11, 2015, at 21:00 UTC, with a sustained wind speed of 59 m/s (indicated by the maroon dot), are illustrated in Fig. 9. At that period, the wind speed in the vicinity of Tuvalu surpassed 10 m/s and the direction was east-southeast. A consistent significant wave height of >4 m was observed at both locations (Fig. 9; denoted by green and black dots), with the wave propagated towards the southeast direction (Table 2). Moreover, the mean wave period at that time was <11 seconds (Table 2). In contrast, on March 2, 2015, in the absence of cyclone Pam (Fig. 7b), the significant wave height was <1.5 m at both specifications and the waves also directed in different ways (southwest direction). Moreover, during this usual situation, the mean wave period was found 9 seconds the surface wind speed was around 5 m/s and the direction was eastward. Due to the limitation in the satellite coverage in the vicinity of Tuvalu, the visualization of wind conditions was confined to the area spanning from 5°S to 10°S and 175°E to 180° (Fig. 9c). A comparison between the scatterometer wind vectors (March 11, 2015, at 21:00 UTC) and ERA5 wind data is presented in Fig. 9(c). The scatterometer-derived wind speed recorded during the cyclone event exceeded 10 m/s, and the wind direction indicated an east-southeast flow (Fig. 9c), which concurred with the ERA5 wind information. As a result, the comparison highlights a higher similarity between the scatterometer-derived wind conditions and the ERA5 dataset.

Fig. 9. Wind and wave distribution map. (a) True color MODIS image (RGB channels: bands 1, 4, and 3) of cyclone Pam on March 11, 2015. (b) Wind and wave conditions within the specified region (yellow dashed box) during the occurrence period (maroon dot) of cyclone Pam. The wind distribution from the scatterometer and ERA5 data is shown in (c). Green and black dots indicate the first (8.5°S, 179°E) and second (8.5°S, 179.5°E) locations, respectively. Blue polygons indicate the islets of Tuvalu.

Originating on January 29, 2021, at a distance of 1,139 km from the Funafuti atoll, cyclone Bina exhibited the characteristics of a tropical storm. Fig. 10 visually maps the trajectory of cyclone Bina along with the recorded wind and wave conditions on January 31 at 03:00 UTC. During this time, the sustained wind speed was reached at 23 m/s (marked by the maroon dot), and the wind speed in the vicinity of Tuvalu exceeded 10 m/s. For both designated locations, the significant wave height was found 2.4 m (Fig. 10; indicated by green and black dots), and the waves were directed toward the south-southwest (Table 2). The mean wave period during the cyclone event was found >7 seconds (Table 2). However, both the wind and wave conditions were different on January 21, 2021 (Fig. 7c) when there was no existence of a cyclone. At that time, the wind speed and direction were identical to the previous two conditions (Figs. 7a, b), and the significant wave height was found 1.4 m while the waves propagated to the southwest direction. Also, the mean wave period (10 seconds) was different compared with the cyclone occurrence time.

Fig. 10. Wind and wave distribution map. (a) True color MODIS (RGB channels: bands 1, 4, and 3) image of cyclone Bina on January 31, 2021. (b) Wind and wave conditions within the specified region (yellow dashed box) during the occurrence period (maroon dot) of cyclone Bina. Green and black dot indicates the first (8.5°S, 179°E) and second (8.5°S, 179.5°E) locations, respectively. Blue polygons indicate the islets of Tuvalu.

4. Discussion

From the investigation, it was found that the frequency of cyclones increased throughout the 22 years in the Southwest Pacific, whereas it decreased in the Northwest Pacific. Furthermore, it was depicted that although tropical cyclones occurred more frequently in the Southwest Pacific, their maximum sustained wind speeds were reduced by 25 m/s, indicating that the cyclone intensity was lowered (Fig. 5b). During this period, a total of 31 cyclones passed in the vicinity of Tuvalu, and 32% of these cyclone’s origin locations were 500 km or less from the Funafuti atoll, indicating a substantial percentage of cyclones generated in proximity to the capital (Fig. 6). In contrast, 23% of cyclones generated 1,000 km away from the capital of Tuvalu which indicates that cyclones developed in distant regions can pass near Tuvalu (Fig. 6). A significant number of cyclones (45%) were found to originate between 501 and 1,000 km from the Funafuti atoll, highlighting a frequent occurrence of cyclones in the middle distance range (Fig. 6). Due to the cyclone event, the wind and wave conditions around Tuvalu are significantly altered and the changes in wind and wave conditions during the cyclone events are depicted in Figs. 8, 9, and 10. These three representative cyclone events consistently resulted in a wave height increase of over 1.2 m in each case. Before the cyclone events, waves near Tuvalu propagated towards the southwest direction. However, as a consequence of these cyclone events, the wave direction shifted to the Southeast. However, the alteration in the mean wave period was marginal, with changes of around 2 seconds.

Several studies have investigated the spatio-temporal coastal changes in Tuvalu using remotely sensed satellite data. However, only a limited number of these research works have depicted the coastal changes induced by intense cyclones, and the wind and wave conditions in Tuvalu’s vicinity during cyclone events have been less explored. With the increasing frequency of cyclones around Tuvalu, understanding the wind and wave conditions during cyclones has gained importance. Reanalysis data, which provide information on wind and wave conditions, can serve as valuable resources in this context. Furthermore, the comparison result between the scatterometer-derived wind and reanalysis wind also showed a strong relation as they provided identical information on the surface wind. By subsequently integrating satellite data, the impact of different intensities of cyclones on the coastline can be analyzed, facilitating the establishment of a link between wave parameters during cyclone events and their impact on the extent of coastal alterations. Combining satellite and reanalysis data can provide clearer insights into coastal changes such as accretion, erosion, or stability. Therefore, it becomes crucial to collect data on cyclone occurrences and their influence on wind and wave conditions around Tuvalu. These findings will subsequently complement the application of satellite data in future studies.

5. Conclusions

According to our previous work on the coastal changes at Funafuti, Tuvalu, most of the alterations in the coastal area happened due to tropical cyclones. As a result, cyclones were identified as one of the driving forces for coastal changes. In this work, we investigate the cyclone occurrences in the Southwest Pacific Ocean from 2000 to 2021 and their effects on Tuvalu. Despite a decrease in intensity, the frequency of cyclones in the Southwest Pacific increased over the 22 years. After the analysis, it was found that the cyclone events had a significant impact on the wind and wave conditions around Tuvalu. During the cyclone, the wind speed and waves in the vicinity of Tuvalu were elevated, and prolonged swells were generated. The waves advanced in an east-southeast direction and these prolonged waves have the potential to significantly change the coastline. In the future, satellite data will be used to comprehend the spatio-temporal changes resulting from the cyclone events in Tuvalu’s coastal zone. Additionally, this wave information will be used to determine the extent of wave influence in the case of coastal changes.