The Signal-to-noise Error in Decadal Regional Climate (Invited Presentation)

Relative to observations of the last century, climate models tend to underestimate the amplitude of regional climate changes. There are three possible sources of this error: (1) models fail to imagine a diverse enough array of counterfactual climates, (2) models underestimate the role of external forcing in regional climate, or (3) we live in an extraordinarily improbable realization of climate. Here, we will show results from large ensembles of multiple climate models and forcing scenarios that all suggest that models underestimate the magnitude of forced variability in major decadal climate modes (e.g. NAO, AMV, PDO) and their impacts. We will examine the potential sources of this underestimation across model components and horizontal resolutions. This newfound, larger role for external forcing in these modes suggests that some recent multidecadal variations in regional climate may also be a response to aerosol and greenhouse gas forcing. Our results also imply that regional climate variability may be both more accurately and precisely predictable than previously assumed.