Climate- and human-driven shoreline changes in a remote Pacific Island: Tubuai, French Polynesia Camille SALMON a,*, Virginie K.E.

DUVAT a, Victoire Laurent ba UMR LIENSs 7266, Université de la Rochelle-CNRS, 2 rue Olympe de Gouges, 17000 La Rochelle, Franceb Direction Inter Régionale de Polynésie Française, Météo France, 98702 Tahiti, French Polynesia* Corresponding author.E-mail address: [email protected] studies have focused on shoreline change in high mountainous islands, whereas their coastal low-lying areas host most population centres and human assets. This paper contributes filling this gap by assessing shoreline change in a remote Pacific island, Tubuai, French Polynesia.

Different shoreline proxies and timescales are considered, based on the 32-year available imagery (1982-2014). Over the study period, erosion was the dominant pattern, especially when considering the base of the beach as a shoreline proxy (57% of transects experienced retreat). The other investigated proxy, the stability line -which consists of the vegetation line or the base of coastal defences-, exhibited logically more stability (61% of transects). At shorter timescales, results show high spatial-temporal variability. Investigating the respective contributions of tropical cyclones and human activities to shoreline change, we found that the former had a moderate influence, despite intense cyclones struck the island over the study period, while the latter had a major influence, despite low population density. In fact, human disturbances have extensively destabilized sediment cells, which is likely to exacerbate the negative impacts of future cyclones on both these systems and the human society.

This case study is particularly interesting to address risk reduction strategies in remote island contexts (i.e. with low population densities and available space).Keywords: shoreline change, tropical cyclones, anthropogenic disturbances, Tubuai Island, French Polynesia, Pacific Ocean?1. IntroductionSmall tropical islands are considered particularly vulnerable to coastal hazards due to their morphological and human characteristics, especially their small size, remoteness, high population pressure on ecosystems and natural resources, and concentration of most human assets in low-lying coastal areas (Pelling and Uitto, 2001; Nurse et al., 2014; Scandurra et al.

, 2018). Coastal hazards, which comprise shoreline retreat and marine inundation, are caused by both extreme episodic events -as distant-source swells and tropical cyclones (TCs) – and gradual changes, such as sea-level rise (Nurse et al., 2014; Ranasinghe, 2016; Duvat et al., 2017a;). Moreover, these risks are expected to increase, as sea-level rise is accelerating (Ranasinghe, 2016).

In order to better estimate the past-to-present and future trajectories of change of islands, and design relevant adaptation strategies, baseline scientific information, such as event-induced and multi-decadal shoreline change, is needed (Hapke et al., 2013; Duvat et al., 2017a). Appreciating shoreline positional change (including rhythm and spatial variability) and its drivers thus constitutes highly valuable information for coastal planning and risk management (Romine and Fletcher, 2012; Duvat et al., 2017b). This need is notably illustrated at the global scale by the recent beach erosion assessment (Luijendijk et al., 2018), which follows the previous attempt of “the International Geographical Union working group on the Dynamics of Coastal Erosion” (1972–1976) and the “Commission on the Coastal Environment” (1976-1984) (Bird, 1985).

Beyond these global assessments, numerous regional or local studies have been realized during the last decades, but concerning small islands, we observe significant gaps according to their type and their geographical extent. Indeed, since the study published by Webb and Kench (2010), the persistence of atoll islands over the 21st century has raised major attention, as indicated by the increasing number of publications on this concern (Duvat, submitted). On the opposite, high islands from volcanic or continental origin were not subject to such scientific efforts, whereas their coastal low-lying areas host most population centres and human assets, including critical infrastructure (De Scally, 2014; Nurse et al.

, 2014). However, these islands are particularly affected by coastal erosion, as showed by Romine and Fletcher (2012a) for Hawaii, by Bheero et al. (2016) and Duvat et al. (2016) respectively for Mauritius and Reunion Island, and by Jackson et al. (2012) for Porto Rico.

The latter islands are the only ones for which data on multi-decadal shoreline change are available. Despite the abovementioned studies did not use exactly the same methods, they all reported shoreline retreat as being a dominant pattern, e.g. reaching up to 85% of Maui’s beaches (Hawaii). In addition, these studies also investigated the contribution of climate-related (e.g.

sea-level rise in Romine et al., 2013, the wave climate in Bheero et al., 2016) and human (e.

g. shoreline armouring in Romine and Fletcher, 2012b and in Jackson et al., 2012) drivers of change.

Although these studies brought significant insights for shoreline change understanding, their number and spatial extent remain limited. Importantly, documenting a large number of islands, according to their types (from low-lying reef islands to mountainous islands, from urban to rural islands, from different areas and political contexts), is crucial to reveal the diversity of existing storylines (including both morphological change and human development), to be able to design context-specific solutions to coastal risks. In particular, it seems important, beyond considering densely-populated capital islands that constitute key hotspots of risks, to also address the situation of rural islands, because they also face important challenges regarding coastal hazards and climate change (Duvat et al.

, 2017a).This paper addresses this gap by studying shoreline change on the northern coast of Tubuai Island, a remote island located in the Austral Archipelago in French Polynesia (Pacific Ocean). Interestingly, Tubuai provides a good example of a territory showing a high-level of exposure of people and human assets to coastal hazards despite a low population density, due to the concentration of the population and human assets within a 200 m-wide coastal strip around the island’s perimeter. This case study thus brings key insights to think risk reduction and adaptation to climate change strategies in remote and rural tropical islands.Based on image analysis and fieldwork, this study aims to: (i) assess temporal multi-scale shoreline change, from multi-decadal (1982-2014) to event-related (i.e. cyclonic), using the stability line and base of the beach as shoreline proxies (Duvat and Pillet, 2017; Duvat et al.

, 2017b); (ii) appreciate the role of climate- and human-related drivers of change. Here, climate-related drivers refer to TCs or tropical depressions (TDs). The fact that three significant events struck the island during the study period (with TC Oli in 2010 being the most intense cyclone ever recorder in French Polynesia), made it possible to analyse their contribution to shoreline positional change. In addition, in line with previous studies (e.g.

Cooper and Pilkey, 2012), we investigated the contribution of human activities to shoreline change and coastal systems’ destabilization. We more particularly assess the role of land reclamation, shoreline armouring, and interception of the longshore sediment drift by transversal structures (wharves, harbours, groynes…) in shoreline change and coastal systems’ destabilization. After presenting the general context of the study, we describe the methodological framework used to measure shoreline change and to estimate the contribution of these two types of drivers. Results are then presented for each time period, followed by a discussion focusing on the attribution of change, the limitations of the study, and its implications for risk reduction strategies are analysed.

?2. Context of the study2.1. Presentation of Tubuai IslandThe Austral Archipelago, which is one of the five archipelagos composing French Polynesia, consists of high volcanic or limestone islands, and of one atoll. Tubuai Island, which is located 640 km south to Tahiti Island, is both the largest and the capital island of the Archipelago (Fig. 1). It is a high volcanic island of 45 km² encircled by a barrier reef and a large lagoon covering 94 km² (Fig.

2A). Mount Taita’a is the highest point of the island, reaching 422 m in elevation. A large coastal plain exhibiting small rivers and swampy areas has formed around the inner mountainous relief. Tubuai’s coast consists of narrow (generally