Glacial/interglacial temperature variations in Soreq cave speleothems as recorded by ‘clumped isotope’ thermometry

‘Clumped isotope’ thermometry is based on analyzing mass 47 in CO 2 extracted from carbonates and uses the tracer mass 47 anomaly (Δ 47). Δ 47 is defined as the deviation of R 47 from that expected for a random distribution of isotopologues and reflects a temperature dependent preference of 13C and 18O to create a bond with each other in CO 2 or in the carbonate lattice. Being an internal characteristic of the carbonate mineral, it is independent of the isotopic composition of the water in which equilibrium precipitation of the carbonate occurs and can therefore be used to independently determine carbonate growth temperatures. This work provides a first examination of the applicability of ‘clumped isotopes’ thermometry to reconstructing the growth temperatures of speleothems, by examining the glacial/interglacial variations of the Δ 47 values of speleothem carbonates from Soreq cave, Israel. The results indicate that the last glacial maximum temperatures were 6–7 °C colder than modern day temperature and a sample at 56 Ky BP was 3 °C colder than the modern. Early Holocene temperatures were slightly above modern day, and late Holocene temperatures were slightly below modern day. These temperature variations are similar to those previously estimated for Eastern Mediterranean sea surface water. Cave water was 18O depleted in the Holocene compared to modern day (by 0.6–1‰) and 1.1‰ more enriched in the last glacial maximum. Comparison of these cave water δ 18O values with fluid inclusion δ D values indicated a late Holocene d-excess value within the range of modern rainfall, implying ∼45% relative humidity. Last glacial maximum and early Holocene d-excess values were significantly lower, suggesting relative humidity of ∼60% and ∼70%, respectively. The temperatures reported in this study were empirically corrected for a non-equilibrium artifact observed in a modern speleothem. The similarity of the temperature variations obtained here to other, independent, records in the region suggests that the Δ 47–temperature calibration slope observed in inorganic synthetic calcite and marine organisms may also be applied in speleothems. But the offset observed in modern temperature suggests that the intercept is different so that a separate calibration is needed for accurate absolute temperature reconstruction using speleothem ‘clumped isotopes’. Similar examination of additional caves would be necessary to determine whether such empirical correction can be generally appl