Quantifying Uncertainties in Climate Science.

Palaeoclimate reconstructions can provide important constraints on climate sensitivity estimates because they are independent of the observational data used in climate model development. However, such assessments typically have one or more of the following limitations: a focus on one palaeoclimate era, risking state-dependent results; simplified model-data comparisons such as regional averages, risking under-use or misrepresentation of reconstructions; reduced model complexity or resolution due to the expense of simulating palaeoclimates, risking substantial biases. We present work in progress on an estimate of climate sensitivity that addresses these limitations.