Re-Evaluating Sources of Inter-Model Spread in Climate Sensitivity

Abstract:

Determining which feedbacks are responsible for the large spread in model predictions of global warming allows us to better focus our research towards reducing this spread. This study shows that there is no unique way to partition temperature change into components associated with individual feedbacks. Alarmingly, different partitioning approaches lead to conflicting conclusions about which processes are important in driving inter-model spread in climate sensitivity. Moreover, covariances between feedbacks is found to play an important yet largely ignored role in determining inter-model spread in climate response. We find, however, that the feedback definition of Held and Shell (2012), which assumes a fixed relative humidity for the baseline climate response, produces results that are more consistent across partitioning methods and are less dominated by covariance terms.