As part of the input for their regional-scale photochemical oxidant model of air pollution, known as the Regional Oxidant Model, requires statistical descriptions of total cloud amount, cumulus cloud amount, and cumulus cloud top height for certain regions and dates. These statistics are used to parameterize photochemical reactions, and to diagnose vertical transport of pollutants. The purpose of this work was to derive these statistics from digital satellite data. The first stage of the work was to apply a human-computer interactive technique to satellite data to distinguish cumulus clouds from clouds of other types, and to produce the cloud amount statistics for cells within specified grids. Next, the interactive technique was applied to the infrared data to calculate the mean cloud top temperature for each cell. Those mean temperatures were then converted to heights by referring to objectively analyzed temperature versus height information derived from radiosonde data. A frequency distribution of cumulus cloud top heights was also produced for some cases.