Quantifying the Influence of Cloud Radiative Feedbacks on Arctic Surface Warming Using Cloud Locking in an Earth System Model
A method called “cloud-locking” specifies the cloud properties used in radiative transfer calculations as external data, decoupling the cloud radiative effect from the predicted atmospheric state. This paper shows an example of applying this method to a climatically important region – the Arctic – while allowing predicted cloud radiative effects elsewhere. Coupled climate change experiments with doubled carbon dioxide in CESM1 are evaluated with cloud radiative feedbacks disabled either globally or only in the Arctic. This approach allows a direct evaluation of how regional cloud feedbacks contribute to the well-known Arctic amplification of warming. Cloud and other feedbacks are diagnosed and compared to understand the relative role of Arctic cloud feedbacks.
This is the first time that the influence of regional cloud radiative feedbacks in the Arctic on Arctic amplification have been directly measured in a climate model. The results of the experiments show that the regional feedbacks have little impact on the amplification. When clouds are locked globally to a pre-industrial state, global warming is 2.43K compared to 3.17K when cloud radiative feedbacks are active; this is consistent with other evaluations that CESM1 has positive cloud feedback. When only the Arctic clouds are locked to a pre-industrial state, the global warming signal is 3.16K, nearly identical to the case with fully-interactive cloud feedbacks. This finding indicates that Arctic clouds have little impact on the global warming signal. Somewhat surprisingly, the Arctic cloud feedbacks appear to have a negligible influence on Arctic warming as well. When focusing on the amplification of Arctic warming – the extra warming experienced by the Arctic compared to the globe -- cloud feedbacks have little influence; the Arctic always warms faster than the rest of the planet. A feedback analysis using radiative kernels complements the cloud-locking approach. Non-cloud feedbacks are also shown to sum to a nearly invariant total with or without cloud-locking, but water vapor and lapse rate feedbacks are individually strongly changed when clouds are locked globally. Effects are similar under Artic cloud-locking but somewhat weaker and restricted to the Arctic region.
This study is a novel investigation of cloud radiative feedbacks. This is the first time that the cloud-locking methodology has been applied under greenhouse warming both globally and regionally, providing a new way to understand the roles of local versus non-local cloud feedbacks. This complements other methods of diagnosing feedbacks, such as radiative kernels. For CESM1, the results indicate that the local Arctic cloud feedbacks have little influence on either the global or Arctic warming. Global cloud feedbacks, on the other hand, are strongly positive and contribute to strong Arctic warming. The Arctic amplification of warming, however, is little impacted by cloud radiative feedbacks.