The Relationship Between the Present-Day Seasonal Cycles of Clouds in the Mid-Latitudes and Cloud-Radiative Feedback

We show that the seasonal cycles of clouds over the mid-latitude oceans in the Northern Hemisphere are predictors of the responses of clouds to increasing sea-surface temperatures globally. These regions are therefore “natural laboratories” in which the processes responsible for low-cloud feedbacks on global scales are observed as seasonal changes in local cloud properties. We use an ensemble of configurations of a global-climate model to show that the sensitivities of cloud-radiative anomalies to surface temperature and lower-tropospheric stability in the “laboratory” regions predict the models' global cloud-radiative feedbacks. Models with greater changes in low-clouds between seasons are shown to have stronger negative feedbacks in the mid-latitudes, and stronger positive feedbacks from the subtropical stratocumulus. The biases in the simulated seasonal cycles, compared to observations, imply that both feedbacks are too weak in the model. The consequences of this for configuring our model to have a lower climate sensitivity are discussed.