On the climatic influence of CO2 forcing in the Pliocene

The Pliocene (~3 million years ago) is of great interest to the palaeoclimate community as a potential palaeoclimate analogue for future climate change. It was the most recent period of sustained warmth above pre-industrial levels, was recent enough to have similar-to-modern continental configuration, and had similar-to-modern atmospheric CO₂ concentrations around 400 ppm. If we are to use the Pliocene as a palaeoclimate analogue for our warmer future, it is important to consider the drivers of Pliocene climate change as well as its comparable large-scale climate features.  

We implement a novel, simple linear factorisation method to assess the relative influence of CO₂ forcing in Pliocene climate change compared to non-CO₂ forcings such as changes to ice sheets, orography and vegetation. Outputs of this method are termed “FCO₂” and reflect the relative influence of CO₂, where 1 represents wholly dominant CO₂ forcing and 0 represents wholly dominant non-CO₂ forcing.

The accuracy of the FCO₂ method is evidenced by comparison to an energy balance analysis using a method previous used in Pliocene climate research, and the energy balance analysis also adds nuance to the FCO₂ results and highlights feedbacks that arise from CO₂ forcing.

We apply the FCO₂ method to seven models from the PlioMIP2 ensemble (CCSM4-UoT, CESM2, COSMOS, HadCM3, IPSLCM5A2, MIROC4m and NorESM1-F) which are found to be representative of the ensemble in terms of the modelled climate sensitivity and global mean surface air temperature anomaly.

CO₂ forcing is found to be the most important driver of surface air temperature change in six of the seven models (global mean FCO₂ of ensemble = 0.56), and five of the seven models for sea surface temperature (global mean FCO₂ of ensemble = 0.56). CO₂ forcing is also the most important driver for precipitation change (global mean FCO₂ of ensemble = 0.51), but spatial patterns in precipitation change are predominantly driven by non-CO₂ forcings and the effects of these must not be overlooked.

CO₂ forcing being the most important driver of change in the climate variables considered here suggests that the Pliocene is a relevant analogue for our warmer future, but attention must also be paid to the significant effects of non-CO₂ forcing in the Pliocene which may be less analogous to the present and near-term future.

Our results also have implications for the interpretation of Pliocene proxy data and data-model comparison. For example, by assessing FCO₂ in regions with large data-model discord it could become possible to highlight where the implementation of boundary conditions is largely responsible for the discord and, hence, where model boundary conditions should be modified. Given the spatial and latitudinal patterns seen in the FCO₂ results, it may also be possible to suggest new sites from which additional proxy data would be most useful.