The Warm Winter Paradox in the mid-Pliocene Warm Period - a focus on model parameterisations.

Abstract

Modelling results from PlioMIP2 (Pliocene Model Intercomparison Project Phase 2) are in strong disagreement with terrestrial proxy data over the high latitudes for the winter season.  This disagreement is large:  models simulate winter temperatures ~20°C cooler than the data suggests.  We term this the ‘warm winter paradox’.We have shown that the warm winter paradox cannot be easily resolved.  For example, changing model boundary conditions to account for orbital and CO2 uncertainty have only a small effect on winter temperatures.Here we use the Hadley Centre General Circulation Model, HadCM3, to investigate whether accounting for uncertainties in model parameterisations could improve the model data agreement for the Pliocene winter.  A new set of parameters for HadCM3, which improve model-data agreement for the Eocene, will be used to investigate the Pliocene climate.  We will show that the new parameters in HadCM3 lead to additional winter Pliocene warming at some locations, although a large model-data disagreement remains.   The new model parameters do not improve the Pliocene data-model comparison as much as they do for the Eocene.  This may indicate that finding a single set of parameters capable of producing an optimised simulation of warm climate states in general is not possible, and that further exploration of model parameter uncertainty is warranted; or that the cause of model data disagreements in the high latitudes may be time period specific.