An Econometric approach to build a new calibration model for Continental TEMPerature proxy? (ECCOTEMP)
The recent discovery of biomarkers such as the Glycerol Dialkyl Glycerol Tetraethers (GDGTs) for reconstructing continental temperatures and hydrological changes has raised a considerable interest in the paleoclimatic community. The applications of these new tools to marine and lacustrine archives are however still at their beginning and require an important calibration effort. One of the to-date limitations relies on the number of calibration models that have been proposed in the last years to predict the temperatures from sedimentary values.
All calibrations (i) use linear ordinary least squares (OLS) regression, most often with aggregate indices (Methylation index of branched tetraethers -MBT- and Cyclisation ratio of branched tetraethers -CBT-) based on GDGTs as the independent variable and temperature the explanatory variable, (ii) rely on a spatially constant relationship between modern soils or lacustrine sediment core tops and observed overlying temperature, (iii) do not consider each GDGT as a separate variable, (iv) do not use the two types of samples (soil and lacustrine sediments) simultaneously to calibrate the models. The present project proposes to extend existing approaches along these four dimensions.
With this project we propose to revisit the existing calibration data-set of modern soils and sediments by means of statistical technics commonly used in economics. We will propose to extend the four dimensions of existing calibration detailed above. The proposed approach is interdisciplinary and relies on a strong collaboration between researchers from CEREGE, who will bring the data sets and their expertise in paleoclimatology, and from GREQAM, who will bring their knowledge about the econometric methods that can be adapted to paleoclimatology issues.