Publication Date:
2023-10-28
Description:
Reconstructions of last interglacial (LIG, MIS 5e, ~127-117 ka) climate offer insights into the natural response and variability of the climate system during a period partially analogous to future climate change scenarios. We present well preserved fossil corals (Diploria strigosa) recovered from the southern Caribbean island of Bonaire (Caribbean Netherlands). These have been precisely dated by the 230Th/U-method to between 130 and 120 ka ago. Annual banding of the coral skeleton enabled construction of time windows of monthly resolved strontium/calcium (Sr/Ca) temperature proxy records. In conjunction with a previously published 118 ka coral record, our eight records of up to 37 years in length, cover a total of 105 years within the LIG period. From these, sea surface temperature (SST) seasonality and variability in the tropical North Atlantic Ocean is reconstructed. We detect similar to modern SST seasonality of ~2.9 °C during the early (130 ka) and the late LIG (120-118 ka). However, within the mid-LIG, a significantly higher than modern SST seasonality of 4.9 °C (at 126 ka) and 4.1 °C (at 124 ka) is observed. These findings are supported by climate model simulations and are consistent with the evolving amplitude of orbitally induced changes in seasonality of insolation throughout the LIG, irrespective of wider climatic instabilities that characterised this period. The climate model simulations suggest that the SST seasonality changes documented in our LIG coral Sr/Ca records are representative of larger regions within the tropical North Atlantic. These simulations also suggest that the reconstructed SST seasonality increase during the mid-LIG is caused primarily by summer warming. A 124 ka old coral documents, for the first time, evidence of decadal SST variability in the tropical North Atlantic during the LIG, akin to that observed in modern instrumental records.
Keywords:
Center for Marine Environmental Sciences; Integrierte Analyse zwischeneiszeitlicher Klimadynamik; INTERDYNAMIK; MARUM
Type:
Dataset
Format:
application/zip, 7 datasets
Permalink
