Solutions for Quality Control of multi-detect or instruments and their application to CRIRES and VIMOS

摘要

Quality Control (QC) of calibration and science data is an integral part of the data flow process for the ESO Very Large Telescope (VLT) and has guaranteed continuous data quality since start of operations. For each VLT instrument, dedicated checks of pipeline products have been developed and numerical QC parameters to monitor instrumental behavior have been defined. The advent of the survey telescopes VISTA and VST with multi-detector instruments imposes the challenge to transform the established QC process from a detector-by-detector approach to operations that are able to handle high data rates and guarantee consistent data quality. In this paper, we present solutions for QC of multi-detector instruments and report on experience with these concepts for the operational instruments CRIRES and VIMOS. Since QC parameters scale with the number of detectors, we have introduced the concept of calculating averages (and standard deviations) of parameters across detectors. This approach is a powerful tool to evaluate trends that involve all detectors but is also able to detect outliers on single detectors. Furthermore, a scoring system has been developed which compares QC parameters for new products to those from already existing ones and gives an automated judgment about data quality. This is part of the general concept of information on demand: detailed investigations are only triggered on a selected number of products.