Hydroclimate variability in the Caribbean during North Atlantic Heinrich cooling events (H8 and H9)

Abstract

We present a speleothem record from western Cuba, spanning the period 98.7-84.9 ka BP. Our record shows two distinctive periods of high δ¹⁸O corresponding to dry and/or cold periods during 85-87.6 and 90.2-93.1 ka BP, synchronous with Heinrich events 8 and 9 (H8 and H9). Hence, we provide the first proxy evidence of the local Caribbean climate response to H8 and H9. Interestingly, H8 is more pronounced compared to H9, which may be a local response to lower temperatures in the North Atlantic resulting in a weak AMOC and reduced deep water formation, therefore a stronger south shift of the ITCZ. Our data complement existing speleothem records from western Cuba which, collectively, provide a nearly continuous paleoclimate time-series spanning the last 100 ka BP, indicating a consistent response to millennial-scale events as dry and/or cooler conditions. The comparison with regional paleoclimate records reveals an anti-phased relationship with South America, caused by the southern movements of the ITCZ during millennial-scale events which lead to dry conditions in the Caribbean and a stronger South American Monsoon System.

Figure 1

Map indicating the approximate locations of the most relevant sites discussed in this manuscript: (1) western Cuban caves, including the Majaguas Upper cave (this study), Torch Cave (Santo Tomás Cave System), and Dos Anas Cave (Majaguas-Cantera Cave System); (2) Terciopelo Cave, Costa Rica; (3) Sediment core ODP 999A, Caribbean Sea; (4) Sediment core MDO3-2621, Cariaco Basin; (5) Sediment cores ODP 1063 and CDH19, Bermuda Rise^–; (6) Botuvera cave, SE Brazil; (7) NGRIP Ice core, Greenland^–. The Atlantic Ocean circulation patterns and typical positions of winter and summer ITCZ are shown as well. The map was made using Adobe Illustrator.

Figure 2

Stalagmite MCS-01 indicating the location of U–Th dating samples, with ages reported in B.P., and the approximate sample positions for the X-Ray Diffraction (green shading).

Figure 3

Chronological age-depth model developed for the stalagmite MCS-01 during MIS 5b based on StalAge Algorithm. The grey lines represent the age model’s 95% confidence intervals. Age uncertainties of dated samples used in the age model are shown as well. Minor outliers shown in the age model we still considered as their uncertainty is increased using an iterative procedure by the StalAge Algorithm.

Figure 4

The MCS-01 speleothem isotope (δ¹⁸O and δ¹³C) records from Majaguas Upper cave in Western Cuba (a); the early Holocene record of stalagmite MCS-01 is shown as well (b).

Figure 5

Comparison of the MCS-01 speleothem δ¹⁸O record from Majaguas Upper cave in western Cuba (c) with the North Greenland Ice Core Project (a)^– and the reflectance record from the sediment core MDO3-2621 from the Cariaco Basin (b) representing ITCZ movements.

Figure 6

Comparison of the MCS-01 speleothem δ¹⁸O record from Majaguas Upper cave in western Cuba (b) with the Terciopelo speleothem record from Central America (a) and the Botuvera speleothem record from SE Brazil (c).

Figure 7

Comparison of the Western Cuban speleothem records (normalized data) (e) by Fensterer et al. (dark blue), Warken et al. (light blue) and this study (blue), with the North Greenland Ice Core Project (a)^–, the Bermuda Rise ²³¹ Pa/²³⁰Th as a proxy of AMOC strength (b)^,,, the North Atlantic SST from core SU90-03 (c), the Caribbean Sea Surface Temperature (SST) and δ¹⁸O records (d) inferred from planktonic foraminifera from the sediment core ODP 999A, and the reflectance record from the sediment core MDO3-2621 (f) from the Cariaco Basin reflecting ITCZ movements. Gray bars indicate the timing of the North Atlantic abrupt cooling episodes (Heinrich events).