Local calibrations for climate proxies - A case study of the TEX86 paleothermometer

Proxy measures are one of the important tools that historical scientists can use to access the past. Paleoclimatologists utilize a number of proxies for temperature, providing important empirical constraints for our historical understanding of earth's climate. In this paper I look to the development of the TEX86 paleothermometer, a relatively new temperature proxy with a complicated history. While initial observations showed a strong correlation between lipid production (GDGTs) of marine archaea and sea surface temperature, further investigation into TEX86 revealed a more complicated set of influences on these GDGTs. In response to these confounds, many researchers have taken to developing regionally specific calibrations, quantifying the correlation between temperature and GDGT production relative to specific environments. I argue that while these local calibrations may find some limited use, their dependency on such specific conditions makes the proxies unreliable and unusable for their intended historical applications. The ability to vicariously control confounding influences does allow a single proxy to reliably measure multiple target phenomena, in principle, but the empirical demands for correctly applying regional black-box calibrations make such measurement models unsuitable for the investigation of deep time.