The files and folders contain cloud cases of "typical" cloud classes extracted from the (complete 5x5000 case) Chevallier-91L dataset. They are thought as examples for such classes, however, we do not guarantee them to be representative in any sense (i.e., they do not constitute any "standard" cases). Cloud scenarios cover a following cloud classes: (high level) ice cloud, (low/middle level) water cloud, warm rain (i.e., no icy top), cold rain, and snow. When possible we selected a class case without and one with overlaying cirrus (that is, single layer and multilayer cases). In addition we added a deep convection case. As a result, following cases have been selected (numbers in brackets give the case-id in the complete set of chevallier cases, ordered as [ccol,oz,q,rcol,t]): - (high level) ice cloud [12287] - low-level water cloud [23340] - mid-level water cloud [12141] - warm rain [3296] - cold rain [2100] - snow [21923] - (low level) water cloud with cirrus [13136] - warm rain with cirrus [2756] - cold rain with cirrus [13681] - snow with cirrus [24430] - deep convection [19915] Chevallier91L_SelectCloudScenesBatch.xml contains the cases in compact batch format (all cases with all parameter profiles in one file, each as a GriddedField4 entry in an ArrayOfGriddedField4; order of cases as listed above). The folders, one per cloud case and named according to the case, contain raw gridded profiles in one file per parameter. Not all cases strictly follow the intended class definition. The model data only contained cirrus-only cases with very low ice contents, all significant amounts of cloud ice are accompanied by snow, i.e., includes precipitating particles. Furthermore, for warm and cold rain and snow cases with overlaying cirrus we could not identify cases with clearly detached high cloud layers. Also, warm rain with overlaying cirrus still has a small amount of ice at it's top, i.e., is not strictly a warm rain case. However, amount of ice is small. High/middle/low level classification follows the ISCCP cloud chart.