Return to search

Assessing the Reproducibility of Coral-based Climate Records: A Multi-proxy Replication Test using Three <em>Porites lutea</em> Coral Heads from New Caledonia

Coral-based climate studies commonly use elemental ratios and stable isotopes of coral skeletons to address seawater temperature and hydrologic balance issues in the tropical surface oceans. Replication, or cross-checking, a standard technique used to assess the fidelity of proxy records in paleoclimatology has not been widely applied in coral-based climate studies, primarily because of the time and cost associated with generating multiple records from a single reef site. Modern and paleoclimate reconstructions based on a single proxy-coral record from a site may contain errors if individual corals from the same reef record different geochemical signals. In this study we perform a replication test using elemental ratios and stable isotopes in three Porites corals from New Caledonia.
The reef complex offshore Amédée Island, New Caledonia is an ideal site to perform a coral replication test because instrumental sea surface temperatures (SST) and sea surface salinity measurements (SSS) have been made there for over 25 years. In this study, we compare sub-monthly resolved, geochemical variations (Sr/Ca, δ18O and δ13C) in three Porites lutea coral heads, located ~500 m apart, with the instrumental SST and SSS records over the interval 1992-1967. The monthly coral Sr/Ca and δ18O time series are well correlated to each other (r=0.86, p<.0001) and to the monthly instrumental SST record (r= -0.86, p<.0001, coral Sr/Ca to SST; r= -0.77, p<.0001, coral δ18O to SST). The three, sub-monthly resolved, 30-year coral Sr/Ca-SST time series have mean SST values that agree within 0.2o C with the instrumental mean SST value. A similar comparison for the coral δ18O-SST records indicate a maximum difference between predicted and observed mean SST of 0.5˚C. Analysis of the monthly climatological means also indicates that Sr/Ca-SST records closely match the instrumental SST record ±0.4˚C; a similar comparison using the δ18O-SST record yields an average offset of ±0.6˚C between observed and predicted monthly SST. Stacking the three records to form composite Sr/Ca-SST and δ18O-SST records does not appreciably improve the goodness of fit between the proxy and instrumental SST records; hence a coral-based proxy climate record from a single coral accurately reflects the observed record of climate variability at this locality. These results support the concept that high fidelity climate records can be generated using a single coral core.

Identiferoai:union.ndltd.org:USF/oai:scholarcommons.usf.edu:etd-2485
Date05 November 2003
CreatorsStephans, Christie L
PublisherScholar Commons
Source SetsUniversity of South Flordia
Detected LanguageEnglish
Typetext
Formatapplication/pdf
SourceGraduate Theses and Dissertations
Rightsdefault

Page generated in 0.0019 seconds