The Medieval Climate Anomaly in Greenland ice core data

Abstract

Temperature signals in ice core data Greenland ice core data can be used to derive information on past climatic conditions in the Greenland area. Recently, Vinther et al. (2010) demonstrated a high correlation between δ18O data and Greenland coastal temperature observations even on seasonal timescales, reaffirming the ability of ice core δ18O to capture local temperatures, first proposed by Dansgaard (1954). The Vinther et al. (2010) analysis of the seasonal ice core δ18O data showed that the winter δ18O contains the strongest temperature signal, and that even annual average Greenland temperatures are more accurately captured in the winter δ18O rather than annual average δ18O data. This surprising fact is partly explained by the observation that Greenland winter temperatures are much more variable than summer temperatures and thus dominate the annual average variability (Vinther et al., 2010). Past Greenland climatic conditions can also be derived directly from Greenland ice core borehole temperature observations, by solving the equation of heat conduction in moving ice and firn (Dahl-Jensen et al., 1998). Using a Monte Carlo inversion technique, Dahl-Jensen et al. (1998) derived temperature histories from both southern and central Greenland from high precision borehole temperature measurements obtained at the DYE-3 and GRIP ice core drill sites. A third way of estimating past Greenland temperature conditions is to measure isotopes on nitrogen and argon trapped in air-bubbles within the ice (Severinghaus et al., 1998). These gases undergo a temperature dependent mass fractionation in the snow column before the air-bubbles are formed in the ice, thus retaining a retrievable temperature signal.