Iterative land proxy based reconstruction of SST for the simulation of terrestrial Holocene climate

Abstract. Sea surface temperature (SST) is the main driver of simulated climate in coupled atmosphere-ocean general circulation models. A reliable reconstruction of past SST is necessary to simulate past climate realistically. We here present a novel method for reconstructing SST on the basis of terrestrial Holocene palaeothermometer data such that a climate model is able to represent the climate mean state in the land temperature time series.

For our study, we use the Earth system model of intermediate complexity Planet Simulator (PlaSim). The land climate is represented by the high-resolution and long-term palaeothermometer time series from Lake Ammersee (Southern Germany), where the temperature is derived from the stable δ¹⁸O isotope in ostracod valves. To provide a climate simulation which reflects the proxy-derived climate during the Holocene, we (i) determine the sensitivities of the terrestrial PlaSim climate with respect to SST anomalies for present day conditions; (ii) define the inverse of these sensitivities to find the SST conditions necessary for representing past land proxy climate; and (iii) reapply the climate model to this newly reconstructed SST. We iterate over steps (ii) and (iii) until the mean model and proxy climate converge.

We demonstrate the applicability of this new method to reconstruct past climate by comparing the simulated land temperatures to an independent (pollen derived) proxy data set of land temperatures for Europe. The implementation of a wider range of terrestrial palaeotemperature information from proxy archives analogous to our method will foreseeably yield better reconstructions of past SST. These can, for example, be used to overcome many models' difficulties with simulations extending to the time before 8500 yr before present, when the North American (Laurentide) ice sheet caused a no-analogue climate.