SINCE COLUMBUS: AN AMO CLIMATE PROXY FROM A PUERTO RICAN SPELEOTHEM

A 420-year reconstruction of summer-weighted rainfall variability over Puerto Rico has been derived by analyzing the d18O of calcite from an actively growing speleothem (PDR-1) collected in Perdida Cave (18.3°N). It is in close agreement with the tree-ring-based AMO (Atlantic Multidecadal Oscillation) through the middle of the 19th century, but diverges significantly prior to 1850.

The stalagmite was collected in June, 2006 and an age of ~10.5 ka was determined for the base by U/Th disequilibrium dating using a multicollector ICPMS. Three other dates were obtained for the top 1.7 cm, where calcite accumulations were nearly constant and without visible discontinuity. Oxygen and carbon isotopes measurements on 200 paired samples had analytical reproducibility of ±0.08‰. Samples were taken at intervals of 0.05mm along the growth axis in the top 6.5mm (2005-1833AD), a per sample resolution of ~1 year, then at 4-year resolution to 1580AD.

A glass plate placed beneath the active drip above PDR-1, and a data logger to record temperature and relative humidity, were retrieved after 2 months. Humidity and temperature were constant at saturation and 22°C, respectively. Calcite precipitated on the glass plate between June and September 2006 had a d18O of -3.9‰, approximately -1.8‰ more negative than the calcite from the top of PDR-1. This offset indicates that speleothem samples incorporate calcite that forms throughout the year, although most of the rainfall occurs during in summer. The d18O of the PDR-1 speleothem is interpreted to reflect the integrated inter-seasonal “amount effect” with a summer-time weighting.

The annual rainfall in Puerto Rico reflects the migratory patterns of the Intertropical Convergence Zone and the speleothem oxygen isotope record appears to record distinctive d18O excursions to -2 ‰, that correspond to modal changes in the AMO. These ended at 1690AD and 1900AD, separated by two shorter returns to -3 ‰ by about 1755 AD and 1956 AD. A third excursion appears to be currently underway. The paired d13C fluctuated synchronously, but with a long-term trend to increasingly lighter values (from -10‰ to -2‰).

Recent hurricane activity has been closely allied to the AMO but there is debate about its persistence: the Puerto Rican speleothem reconstruction will be of importance to ascertain its role prior to using it for interpreting significantly older time periods.