Mid- to Late Holocene Sea-Level Record in French Polynesia, South-Central Pacific

The Mid- to Late Holocene provides the opportunity to study the coastal response to sea-level change that has a similar amplitude (i.e., a few decimetres up to 1 m) to the sea-level rise that is likely to occur before the end of the current century. Furthermore, this time period provides an important baseline of natural climate variability prior to the industrial revolution. This study aims to reconstruct Mid- to Late Holocene relative sea-level change in French Polynesia by examining coral reef records from ten islands, which represent ideal settings for accurate sea-level change studies because: 1) they can be regarded as tectonically stable during the relevant period (slow subsidence), 2) they are located far from former ice sheets ('far-field'), 3) they are characterized by a low tidal amplitude, and 4) they cover a wide range of latitudes which produces significantly improved constraints on GIA (Glacial Isostatic Adjustment) model parameters. The accurate reconstruction of sea-level change relies on absolute U/Th dating of in situ coral colonies and their accurate positioning via GPS RTK (Real Time Kinematic) measurements with a vertical and horizontal precision of ± 2.5 cm and ~1 cm, respectively. We focus mainly on the analysis of coral microatolls, which are sensitive low-tide recorders, as their vertical accretion is limited by the water level. Their growth patterns allow the reconstruction of low-amplitude and high-frequency sea-level changes on centennial to sub-decadal time scales. A sea-level rise of less than ~1 m is documented between 6 and 3-3.5 ka, and is followed by a gradual fall in sea level that started around 2 ka and persisted until the past few centuries. The reconstructed sea-level curve therefore extends the Tahiti sea-level curve [Deschamps et al., 2012, Nature, 483, 559-564], and is in good agreement with a geophysical model tuned to fit far-field deglacial records [Bassett et al., 2005, Science, 309, 925-928].