Exploring errors in paleoclimate proxy reconstructions using Monte Carlo simulations: paleotemperature from mollusk and coral geochemistry
1Institut des Sciences de l'Evolution, CNRS-UM2-IRD, UMR5554, Université Montpellier 2, Pl. Eugène Bataillon, 34095 Montpellier, France
2University of Washington, School of Oceanography, Box 355351, Seattle, WA 98195, USA
3University of Washington, Department of Atmospheric Sciences, Box 351640, Seattle, WA 98195, USA
Abstract. Reconstructions of the past climate from proxy records involve a wide range of uncertainties at every step of the process. These uncertainties and the subsequent error bar in the reconstruction of a paleoclimatic variable need to be understood and quantified in order to properly interpret the reconstructed variability and to perform meaningful comparisons with climate model outputs. Classic proxy calibration-validation techniques are not well-suited for identifying the causes of reconstruction errors, estimating their relative contribution, or understanding how errors accumulate from a multitude of sources. In this study, we focus on high resolution proxy records based on calcium carbonate geochemistry of sessile organisms such as mollusks, corals, or sclerosponges, and propose an approach based on Monte Carlo simulations with simple numerical surrogate proxies. A freely available algorithm (MoCo, http://www.isem.cnrs.fr/spip.php?rubrique472) is provided for estimating systematic and standard errors of mean temperature, seasonality and variance reconstructed from marine accretionary archive geochemistry. This algorithm is then used for sensitivity experiments in a case study to characterize and quantitatively evaluate the sensitivity of systematic and standard errors to sampling randomness, stochastic uncertainty sources and systematic proxy limitations. The results of the experiments yield an illustrative example of the range of variations that climate reconstruction errors may undergo, and bring to light their complexity. One of the main improvements of this method is the identification and estimation of systematic bias that would not otherwise be detected. It thus offers the possibility of correcting the proxy-based climate from these biases for a more accurate reconstruction. Beyond the findings of error sources for coral and mollusk-based reconstructions, our study demonstrates that numerical simulations based on Monte Carlo analyses are a simple and powerful approach to improve the proxy calibration process. A thourough understanding of the proxy record errors is essential for the interpretation of paleoclimate records from proxies derived from accretionary skeleton geochemistry. The error estimates provided by MoCo are much more comprehensive and therefore closer to reality than error estimates provided by typical calibration studies.