How did the North American ice Saddle Collapse impact the climate 14,500 years ago?

Friday, 19 December 2014
Ruza F Ivanovic, University of Leeds, Leeds, United Kingdom, Lauren J Gregoire, University of Leeds, Leeds, LS2, United Kingdom, Andrew D Wickert, University of Colorado at Boulder, INSTAAR and Department of Geological Sciences, Boulder, CO, United States, Paul J Valdes, University of Bristol, Bristol, United Kingdom and Natalya A Gomez, New York University, New York, NY, United States
Around 14.5 ka, global sea level rose by around 15 m in less than 350 years (e.g. Deschamps et al., 2012) during an event known as Meltwater Pulse 1a (MWP1a). Modelling work by Gregoire et al. (2012) suggested that around half of this ~50 mm yr-1 sea level rise was caused by accelerated collapse of the ice saddle between the N. American, Cordilleran and Laurentide Ice Sheets. Sea level records place MWP1a in the Bolling-Allerod period, a time of Northern Hemisphere warmth, but dating uncertainties make it difficult to determine the sequence of events and their drivers, leaving many fundamental questions. For example, did the abrupt ice saddle collapse and melting from other ice masses have any detectable climatic impact, or were there no feedbacks? Was melting from the Northern Hemisphere ice sheets responsible for the Older-Dryas (Menviel et al., 2011) or other cooling events? And how were all these signals linked to changes in ocean overturning circulation?

To evaluate ice sheet impacts on climate during the Bolling, we integrate ice sheet, solid earth, drainage and climate modelling. We examine the effects of changing orography and meltwater inputs by combining ice-sheet model results, including the significant Laurentide–Cordilleran ice saddle collapse (Gregoire et al., 2012) and a new model of Antarctic ice sheet evolution (Gomez et al., 2013), with the ICE-5G reconstruction (Peltier et al., 2004). We route meltwater to the oceans using a high-resolution drainage calculation (Wickert et al., 2013). These modelled changes in global ice sheet topography and meltwater routing are then used to force the HadCM3 Ocean–Atmosphere–Vegetation general circulation model (using the 15 ka set-up of Singarayer and Valdes, 2010).

We compare the climate model results to proxy records for temperature and ocean circulation changes during the Bolling warming and the Older Dryas to evaluate the different hypotheses on the link between MWP1a and climate change.