Do Phases of the Moon Affect Phases of Science Software Development?

This paper is based on observation of a number of satellite remote sensing missions and their associated data production efforts. Projects typically transition from initial concept through pre-launch development, launch and early operations, initial release of data, release of "validated" data and into final reprocessing and end of life. Different drivers affect development of the science code during each phase. Science teams tend to create new algorithm versions in distinct patterns during each project phase. One result of this is that a multitude of software and data set versions are created for each data product. This paper will delineate typical project phases, examine the main drivers behind algorithm change, and illustrate typical patterns of science software versions for each phase. Understanding these phases and the nature of the data prodceed in each can improve estimates of both the time it will take and the magnitude and schedule of resources needed to develop, produce, archive and distribute data. It can also help users understand when they will be able to obtain and use data from new projects and how the data sets will change with time. By the way we have not found any connection between moon phases and data set development phases.