Improving the fidelity of coral-based climate records: A roadmap for bypassing intercolony variability and diagenesis
Sayani, Hussein R.
MetadataShow full item record
The tropical oceans are a critical part of climate system, modulating global temperature and precipitation patterns on a variety of timescales. Yet, decadal-to-centennial scale climate variability in the tropics remains poorly characterized due to the dearth of instrumental observations prior to 1970. Massive scleractinian corals are excellent records of past oceanographic variability as they are widely distributed across the tropical oceans, form seasonally-banded skeletons, and can be absolutely dated via U/Th. However, accuracy of coral-based climate reconstructions has been questioned as corals growing on the same reef can produce very different geochemical signals. Diagenetic alteration, which is prevalent among both modern and fossil corals, is also known to significantly compromise the fidelity of coral-based reconstructions. In this thesis, I assess the reproducibility of coral-derived temperature and salinity records using a collection of neighboring corals that grew at Palmyra Atoll between 1980-2010. I also explore the potential of using secondary ion-probe mass spectrometry (SIMS) to extract reliable climate information an altered fossil coral. Chapter 2 assesses intercolony reproducibility of coral δ18O and Sr/Ca records using 5 cores from Palmyra Atoll (6˚N, 162˚W). In general, Palmyra coral δ18O and Sr/Ca records exhibit monthly to interannual variability that is consistent with temperature observations, and reproducible among different colonies. However, the absolute values of individual coral δ18O and Sr/Ca records are consistently offset from each other. The temperature sensitivity of coral Sr/Ca also varies among different colonies, producing 5 distinct Sr/Ca-SST calibrations. The presence of intercolony variability among Palmyra corals implies that reconstructions based on individual fossil corals from this site may be unable to resolve mean climate changes smaller than ~1.4˚C. However, multiple overlapping corals can be used to constrain climate-driven differences in mean δ18O and Sr/Ca between two time periods, providing more reliable estimates of mean climate change. Chapter 3 presents a comprehensive assessment of the reproducibility of SIMS Sr/Ca measurements, in order to extract a reliable Sr/Ca-based SST record from an altered fossil coral. At micrometer-scales, the coral skeleton is comprised of two basic features: (i) centers of calcification (COCs), which account for <5-8% of the coral skeleton and contain consistently higher Sr/Ca, and (ii) aragonite fibers, which account for >90% of the coral skeleton and are consistently lower in Sr/Ca. The geochemical composition of both these features is highly variable at micrometer-scales, however reliable estimates of bulk Sr/Ca can be readily obtained using SIMS by making 3-4 measurements, per month of skeletal growth, on only the aragonite fibers. With this approach, credible Sr/Ca-based SST records are obtained in all pristine modern coral samples where COCs were successfully avoided. Application of this technique to two lightly-altered sections of a young fossil coral reveals that low levels of diagenesis do not impact the fidelity of bulk Sr/Ca measurements. Across a heavily-altered patch on the same fossil coral, where bulk measurements erroneously suggest the ocean was up to ~6˚C cooler, targeted SIMS analyses of only pristine material provides both bulk Sr/Ca estimates and a ~4yr time series that are consistent with SST. Overall, these results demonstrate that SIMS is a powerful tool that can be used to verify the accuracy of reconstructions from fossil corals.