HR: 17:45h
AN: PP44A-08 [Abstracts]
TI: Growth-Dependent Calibration of Coral Sr/Ca-SST From Multiple Colonies Provides Potential for Long SST
Records from Fossil Corals
AU: * Goodkin, N F
EM: goodkin@whoi.edu
AF: Woods Hole Oceanographic Institution, 360 Woods Hole Road, M.S. 25, Woods Hole, MA 02543
AB:
Extended reconstructions of sea surface temperature (SST) are critical to examining long-term climate variability not
captured in instrumental records. Coral skeleton, which continuously accretes in annual density bands, preserves unique,
multi-century archives of sub-annual resolution SST. Despite the promise of coral proxies, however, SSTs derived from corals
are often several degrees cooler than those derived from other archives.
Here we present strontium to calcium ratios (Sr/Ca) for four brain corals (Diploria labyrinthiformis) collected from the
south shore of Bermuda that are strongly correlated with both instrumental SST (Hydrostation S, 30km southeast) and annual
skeletal extension rate. High Sr/Ca ratios correspond with cold SSTs and slow skeletal growth rate, and vice versa. Over a
~25 year calibration period, the four corals have distinct average growth rates (2.57, 2.68, 3.55 and 4.03 mm/yr). For each
colony, we provide a quantitative calibration of annual Sr/Ca to annual extension rate and annual SST along the axis of
maximum growth and derive an individual growth dependent Sr/Ca-SST calibration equation:
Sr/Ca = m*(SST) + n*(annual growth rate)*(SST) + b
The slopes and intercepts of the four equations are found to be linearly related to the average growth-rate during the
calibration periods of each colony, and a final multi-variant regression is performed to establish one final Sr/Ca-Growth
Rate-SST calibration, in the form:
Sr/Ca = m*(SST) + n*(annual growth rate)*(SST) + o*(average colony growth rate)*(SST) + b
This growth-dependent calibration is shown to be applicable to a fossil coral of the same species in order to reconstruct
SSTs at Bermuda for 223 years. A reconstruction excluding the influence of growth yields SSTs that exaggerate both cool and
warm periods. SST anomalies near the end of the Little Ice Age (~1850) that are derived using a non-growth dependent
calibration are exaggerated by a factor of two relative to those from a growth-dependent model which yields SSTs ~1.5 §C
cooler than today. Our results indicate that incorporation of growth rate effects and multiple colonies into coral Sr/Ca
calibrations improves the accuracy of SSTs derived from living and fossil corals, and increases the utility of the
calibrations for application to other individual colonies.
DE: 1050 Marine geochemistry (4835, 4845, 4850)
DE: 4916 Corals (4220)
DE: 4954 Sea surface temperature
SC: Paleoceanography and Paleoclimatology [PP]
MN: Fall Meeting 2005