Instrument performance issues are the major risks to successful data validation and improvement. Instrument performance subtleties, although usually correctable, are often overlooked and corrupt final results. Our software development strategy requires that algorithm development teams participate in the design of ground testing and the analysis of calibration data. This "hands on" experience provides data analysts with first-hand knowledge of instrument performance. It also promotes communication between instrument and software development teams. This communication is vital during data validation and analysis, when experts from different fields must work together to efficiently diagnose and solve problems.

GATS engineers use the level 0 through level 1 processing system (commonly used only during post-launch) to support pre-launch testing and calibration. This thoroughly tests data processing codes using "real" instrument data, without the need for a costly data simulation effort, while taking advantage of instrument engineering expertise during the building and testing of the data processing system. The result is software development proficiency and a data analysis team with a thorough understanding of the instrument, which equates to prompt and reliable post-launch results.