Given their role as the primary driver of uncertainty in climate sensitivity, it is a vital that better constraints on cloud feedbacks be determined. This includes accurately quantifying them in observations, interrogating climate models to determine whether they can faithfully reproduce them, and establishing the degree to which skill in simulating feedbacks on observable timescales translates to skill on climate change timescales with relevance for climate sensitivity. In this talk I will highlight recent work on these three fronts. Several novel diagnostic techniques are being applied to the roughly two decade-long satellite cloud and radiative flux record to better establish how clouds and their radiative properties depend on both the mean and pattern of surface warming. This, coupled with satellite simulator output, allows for a more rigorous evaluation of individual cloud feedback components in climate models, exposing important areas of model bias. It also highlights feedbacks that may be important in nature but are not well understood and likely to be poorly represented in models. Additionally, we will present evidence that cloud feedback components beyond the oft-discussed marine low cloud feedback exhibit dependence on surface warming patterns, suggesting that caution is needed before directly relating feedbacks derived from short records to the long-term feedback. Key knowledge gaps, observational uncertainties, and model biases will be discussed, along with their implications for cloud feedbacks determined through recent expert judgment assessments.

This work was performed under the auspices of the U.S. Department of Energy by Lawrence Livermore National Laboratory under Contract DE-AC52-07NA27344. It is supported by the Regional and Global Model Analysis Program of the Office of Science at the DOE. IM Release # LLNL-ABS-852580.