Modeling the 8.2 ka event using a coupled atmosphere–ocean GCM

A series of simulations have been performed using the Hadley Centre GCM, (HadCM3) to investigate the 8.2 ka event. A control simulation with 9 ka boundary conditions showed that intrinsic variability in the strength of the N. Atlantic Meridional Overturning Circulation (MOC) was greater than in a comparable pre-industrial simulation but these natural variations were not strong enough to replicate the 8.2 ka event at any time, consistent with the idea that the event was ‘forced’ through a freshwater discharge. We therefore investigated the response of the 9 ka climate to a freshwater forcing consistent with the final drainage of Lake Agassiz. A 10 member ensemble was performed with 5 Sv of freshwater added to the N. Atlantic ocean for 1 year at various times of the control run, which led to a reduction in the strength of the MOC and the temperature, precipitation and δ18O of precipitation over the N. Atlantic region. The strength and location of the response agreed well with proxy evidence of the 8.2 ka event but the duration was too short in all cases. The response to the freshwater forcing was sensitive to the initial state of the ocean, with better agreement with paleodata when the freshwater was added at a time when the Atlantic MOC was relatively weak. The structure and duration of the δ18Op response was consistent over a large region, and appeared in better agreement with regional precipitation than regional temperature, however there were spatial differences between the maximum δ18Op response and both the maximum temperature and precipitation responses.