Setting up a model intercomparison project for the last deglaciation

Abstract:

The last deglaciation (~ 21-9 ka) presents a series of opportunities to study the underlying mechanisms of abrupt climate changes and long-term trends in the Earth System. Most of the forcings are relatively well constrained and geological archives record responses over a range of timescales. Despite this, large uncertainties remain over the feedback loops that culminated in the collapse of the great Northern Hemisphere ice sheets, and a consensus has yet to be reached on the chains of events that led to rapid surface warming and cooling during this period.

Climate models are powerful tools for quantitatively assessing these outstanding issues through their ability to temporally resolve cause and effect, as well as break down the contributions from different forcings. This is well demonstrated by pioneering work; for example by Liu et al. (2009), Roche et al. (2011), Gregoire et al. (2012) and Menviel et al. (2011). However, such work is not without challenges; model-geological data mismatches remain unsolved and it is difficult to compare results from different models with unique experiment designs.

Therefore, we have established a multidisciplinary Paleoclimate Model Intercomparison Project working group to coordinate transient climate model simulations and geological archive compilations of the last deglaciation. Here, we present the plans and progress of the working group in its first phase of activity; the investigation of Heinrich Stadial 1 and the lead into the Bolling warming event. We describe the set-up of the core deglacial experiment, explain our approach for dealing with uncertain climate forcings and outline our solutions to challenges posed by this research.

By defining a common experiment design, we have built a framework to include models of different speeds, complexities and resolution, maximising the reward of this varied approach. One of the next challenges is to compile transient proxy records and develop a methodology for dealing with uncertainty and error in model-geological data comparisons.

Through this global and interdisciplinary initiative, we combine multi-proxy records with a suite of different modelling techniques to test hypotheses for abrupt climate changes and reconstruct the chain of events that deglaciated the Earth 21-9 ka.