Paul Valdes, a climatology researcher at the University of Bristol recently argued that existing climate models have been unable to simulate abrupt climate changes due to oversimplifying the factors involved. This means that we do not understand past climate-shaping events and of more immediate concern, he says this could render us unable to predict massive changes.
In his editorial in Nature Geoscience, Valdes claims that looking to historical climate shifts and the conditions that sparked them is difficult due to the number of factors involved, but if we are to react to coming changes, we require more sophisticated models to understand these events.
Historical events like the Palaeocene-Eocene Thermal Maximum shift that was marked by rapid warming could explain (and warn us) of future rapid climate changes but current models cannot simulate the climate that preceded the change.
As a discussion of Valdes’ argument in Ars Technica pointed out, “Although climate models have been accused of being overly sensitive to changes in greenhouse gasses, it seems that in some cases, the models are too stable, requiring larger perturbations to cause the actual changes seen in the past.”
Valdes argues that because of this over-stability the models are underestimating the possibility of vast, rapid climate shifts and might lead to “a false sense of security.”