Clouds and aerosols are key limitations to our ability to project future climate. This study seeks to assess the current state of knowledge of clouds and aerosols,with an emphasis on how they are represented in climate models, and determine future research directions. Specifically, the committee will: 1. Convene a community workshop to assess the state of the knowledge in cloud and aerosolresearch; 2. Determinea research road map that lays out the most fruitful research directions in thecoming 10-20 years to improve the representation of clouds and aerosols inclimate models. The committee should address the following questions: How are the properties of aerosols and clouds measured in the atmosphere? What are the current and future measurements that can be made from satellites? What aerosol and cloud properties can be measured in situ or via remote sensing? What measurement technologies are likely to be developed in the next 2 decades? Which ones are most needed? How are aerosols and clouds treated in major climate models? What limits our understanding of the climatic effects of aerosols—both their direct effects and indirect effects on clouds? How does the treatment of aerosols in models compare with what has been learned from observations? Are models taking full advantage of existing measurements of aerosol and cloud properties? How are regional impacts of aerosols dealt with in climate models? How do these affect regional issues such as melting sea ice? What steps are needed to improve treatment of aerosols in climate models? What measurements would be most useful? What should bedone to improve the certainty in our knowledge of aerosols and clouds inclimate models? Why are the uncertainties in the radiative forcing terms associated with still so large? Is radiative forcing the most appropriate way to frame our understandingof the climatic effects of clouds and aerosols? What should the research priorities be in the next 5 years? In the next 20 years? What information do modelers need from process studies and observations? How do we transition knowledge from process studies and observations to models?