----- Ann Fridlind ----- Cloud-resolving models: How good are they and what can they do for you? My work at GISS using cloud-resolving models (CRMs) has two primary goals: (i) generate the best possible simulacra of mixed-phase cloud regimes in the atmosphere and (ii) use the results to improve cloud parameterizations for general circulation models (GCMs). Working with Andy Ackerman and Bastiaan van Diedenhoven (our new post-doctoral associate), our approach is first to focus on case studies from aircraft-based field campaigns, which generally provide uniquely detailed data on cloud structure and the background atmosphere in which clouds form---such as aerosol size distributions and ice nucleus concentrations above and below cloud layers, droplet and ice concentrations and size distributions. We then aim to maximize use of ancillary data (soundings, reanalyses, and ground-, aircraft-, and satellite-based remote sensing data) to initialize, drive, and evaluate the quality of the CRM simulations. I will give examples of how the resulting data-constrained simulations can be used to study outstanding, unresolved problems of cloud physics (such as ice formation), and how they may also be used to study phenomena that are important to parameterize in GCMs (such as cloud-scale fluxes and associated tracer transport, aerosol processing, and so on).