Consistent Observational and Numerical Modeling Support for Ice Sheet Forcing of DOI event 8

The Binge/Purge (B/P) hypothesis of abrupt climate change (MacAyeal, 1993a,b, Paleoceanography) suggests that an instability and periodic collapse of the Laurentide ice sheet is responsible for the observed abrupt warmings of Greenland (Dansgaard-Oeschger Interstadials or DOIs) through the influence of ice sheet meltwater on ocean circulation and its poleward heat transport. If the B/P hypothesis is correct, then one should be able to use the Greenland paleotemperature data and a climate model to infer changes in global sea level associated with the DOIs. The B/P hypothesis could then be tested by comparing the inferred sea level changes with independent reconstructions based on marine records (corals and sediments). We use stochastic inversion of a seasonal resolving climate model (Schmittner and Stocker, 1999, J. Climate), composed of a zonally averaged ocean model (Wright and Stocker, 1992, J. Geophys. Res.) coupled to an energy and moisture balance model of the atmosphere, to infer the changes in ice sheet meltwater required to explain the Greenland paleotemperature record from 39 ka to 34 ka which include DOIs 8, 7, and 6. We found a compelling consistency between modeled and reconstructed sea level changes for DOI 8 (a particularly long interstadial following Heinrich Event 4), which tends to support the B/P hypothesis. On the other hand, the large estimated uncertainties in the inversion solution and sea level reconstructions do not permit a firm test of the B/P hypothesis for DOIs 7 and 6.