Recent Advances in Stray Light Modeling for Large Telescope/Observatory Systems

摘要

Stray light analyses have been a serious consideration in the development of optomechanical instrumentation from the time of Galileo. In the late 1960's and 1970's, NASA and the US Air Force realized that software modeling was the key to predictive development and analysis of optical systems and funded the first efforts to automate the analysis process with the development of the GUERAP and APART computer software programs.. The development of multi-wavelength scatterometers in that same era made it possible to more fully characterize the scatter properties of paints and other surface treatments.. These two improvements advanced stray light modeling capabilities and a well-developed modeling approach. In the past twenty years demands for modeling complex structures (e.g., individual screws and small vanes edges) coupled with requirements for ever increasing sensitivity have placed a tremendous computational burden on the stray light analyst. In this paper, we discuss the approaches to modeling today's complex telescope systems that frequently include stray light artifacts traceable to the complete observatory. Scatter modeling and data reduction has become major components of the stray light analysis task and recent work has identified instances where scatter measurements performed by skilled technicians are inaccurate and useless without additional analysis and interpretation. Finally we discuss common metrics for stray light characterization of large systems and how distributed computing, cloud computing, and GPU-enabled software allow the analyst to compute levels and uniformity of stray light in these systems to levels heretofore considered impossible.