Abstract
Clouds are an integral component of our atmosphere and are involved in complex interactions that regulate the climate of our Earth. It is necessary to have a comprehensive observational record of both cloud type and cloud cover in order to look at trends, better understand atmospheric processes, and to compare with models. Unfortunately, there are problems with both manual and instrumental observations because clouds are very difficult to objectively classify into distinct categories. This paper examines the science behind cloud classification and the network of tools, algorithms, and instruments that shape our understanding of clouds. Constructing an automated cloud classification algorithm for sky images confronts problems of observational subjectivity and examines the validity of technological solutions.