Cloud feedbacks are one of the greatest uncertainties in climate modeling and climate change predictions. Sorting cloud effects and cloud changes by large-scale dynamical regimes (e.g. 500 hPa vertical velocity) has proved to be useful in understanding cloud variability and identifying discrepancies among models. CloudSat provides the first global vertical profiles of cloud liquid and ice water contents. By sorting the CloudSat-observed cloud water content profiles by a variety of large-scale variables (mostly satellite observations), we provide a set of new observational metrics for evaluation of cloud simulations in general circulation models (GCMs). These large-scale parameters include sea surface temperature (SST), SST gradients, surface divergence, precipitation, water vapor path, convective available potential temperature and lower-tropospheric stability as well as 500 hPa vertical velocity from reanalysis. We found that vertical structure of clouds is clustered in various large-scale parameter space. These physics-based, phenomenon-oriented cloud-sorting techniques shed lights on the relationships between clouds and large-scale environment. Examples of the regime-specific comparison of modeled and observed cloud profiles will be presented.