Cloud feedback drives most of the inter-model disagreement in how much the climate will warm in response to human activities.  Climate models produce a negative cloud feedback in middle latitudes over the oceans and especially strongly so over the high latitude oceans in the Southern Hemisphere.  Latitudes poleward of 45 degrees are the only regions where cloud feedback is consistently predicted to be negative in climate models. Different models do not agree on the strength of this feedback, however, which has a large inter-model spread of -0.5 to -3.0 Wm-2 in the regions from 40-60S.  In all models tested the clouds in this region reflect more solar radiation as the climate warms, thus reducing the warming that would otherwise occur and also causing changes in the wind patterns that can affect weather and the ocean circulation.  Models also disagree on the amount of solar radiation that clouds in this region reflect back to space for the present day climate.  Another interesting aspect of this region is that low clouds, which have their tops within a few kilometers of the surface and are the most common type of cloud, are very cold and consist of a combination of liquid water that has been cooled below the freezing point (super-cooled) and ice.

The objectives of this investigation are to provide a deeper understanding of the brightening of middle and high latitude clouds that occurs in many climate models as the climate warms, and to better understand why models give varying estimates of the strength of this negative feedback.  The investigation may also provide new information on why models are deficient in their simulation of the present day clouds in these regions.  The analysis will reveal which models simulate current clouds most realistically in the regions of interest, and relate this to the strength of the cloud feedbacks in the models.  This will improve the uncertainty quantification of climate simulations and will provide theoretical understanding potentially leading to increased reliability of climate change projections.  This will enhance the predictive understanding of the Earth system by analyzing the natural and anthropogenic components of global climate change in models.

The primary methods involve the study of the clouds in the suite of CMIP5 model climate simulations and comparison of clouds in climate simulations with data from satellite measurement systems, with a particular emphasis on mixed-phase cloud physics, for which ice and water are present simultaneously.  State-of-the-art climate model feedback analysis will be combined with new and more comprehensive observations of clouds available from modern measurement systems. Novel use of model simulations in combination with observations enables a deeper understanding of the role of clouds in climate change.  Metrics that facilitate model-to-model and model-to-observation comparisons related to mixed-phase cloud physics will be developed and used to highlight challenges and opportunities for improving the state-of-the-science models. This investigation is a collaboration between the University of Washington, where the data work is concentrated, and LLNL where the model data and the tools and expertise to do cloud feedback analysis are located.