. Arctic Amplification in Initial Condition Large Ensembles

Abstract
While the representation of physical processes in models has become more sophisticated over the last 30+ years, the vertical and seasonal fingerprints of forced Arctic greenhouse warming in models have not changed. Remarkably, observations in recent decades show the same fingerprints as the models: surface amplified warming in late fall as a consequence of ocean heat gain in the previous summer. Despite consistent fingerprints in both models and observations, internally generated climate variability is large and can overwhelm forced response fingerprints. The advent of multiple large initial condition ensembles provides a new opportunity to assess Arctic Amplification in the context of both the forced fingerprints signal and internally generated noise. In this presentation, we will assess Arctic Amplification in multiple large initial condition climate model ensembles. We will address: Can the models reproduce observed Arctic Amplification? What is the relative contribution of internally generated noise and forced signal to Arctic Amplification? How and why do the models differ in their internally generated noise and forced climate response?