The strontium to calcium ratio (Sr/Ca) in aragonitic skeletons of massive corals provides a proxy for sea surface temperature (SST) that can be used to reconstruct paleoclimates across decades, centuries, and, potentially, millennia. Determining the reproducibility of Sr/Ca records among contemporaneous coral colonies from the same region is critical to quantifying uncertainties associated with the Sr/Ca‐SST proxy. We evaluated both intracolony and intercolony variability in Sr/Ca using five modern colonies of Orbicella faveolata collected live from the Dry Tortugas National Park, FL. We regressed all available Sr/Ca‐SST data pairs from the five O. faveolata colonies against the Advanced Very High Resolution Radiometer gridded SST data set to produce a new Sr/Ca‐SST calibration equation (Sr/Ca = −0.049 × SST + 10.460), which we suggest can be applied to O. faveolatacolonies collected throughout the Gulf of Mexico/Caribbean region. We estimated total uncertainty by calculating the root‐mean‐square of the intracolony, intercolony, and analytical error terms. Our (1σ) uncertainty estimates of 0.082 mmol/mol (1.66 °C) for subannual Sr/Ca‐SST and 0.070 mmol/mol (1.43 °C) for mean annual Sr/Ca‐SST represent conservative error terms that can be applied to individual data points in single‐colony Sr/Ca‐SST reconstructions. We illustrate how these uncertainties can be significantly reduced by generating multicolony reconstructions and/or through replication of sampling within individual coral colonies. Although the uncertainties on absolute Sr/Ca‐based SST are likely too large to allow researchers to evaluate subdecadal temperature variability, we show that the O. faveolata paleothermometer can reliably detect changes of ~2 °C across decadal timescales and ~1 °C over multidecadal timescales.
Additional publication details
|Publication Subtype||Journal Article|
|Title||Quantifying uncertainty in Sr/Ca-based estimates of SST from the coral Orbicella faveolata|
|Series title||Paleoceanography and Paleoclimatology|
|Contributing office(s)||St. Petersburg Coastal and Marine Science Center|
|Google Analytic Metrics||Metrics page|