A model-based monitoring system for a space-based astrometry mission

Astrometric space missions like Hipparcos, DIVA, Gaia have to simultaneously determine a tremendous number of parameters concerning astrometric and other stellar properties, the satellite's attitude as well as the geometric and photometric calibration of the instrument. To reach the targeted level of precision for these missions many months of observational data have to be incorporated intos a global, coherent and interleaved data reduction. It is inevitable that a daily data reduction process is required in order to judge if the level of precision of the stellar, attitude and instrument parameters achieve its targeted level. This sophisticated data analysis is the in-depth scientific assessment of the quality of all observations within about 24 hours after its reception. It is based on the very complicated procedure "First Look preprocessing" (more known as a Great-Circle reduction from Hipparcos) that provides a one-dimensional, self-consistent and simultaneous solution of the attitude, the instrument calibration and celestial source parameters. For this purpose one needs to process all the 24-hours-data, a task which can be only performed at the Data Center with its computer resources. On the other hand, it is necessary and reasonable to process the observations at the ground Space Operations Center for a quick discovery of delicate changes in the spacecraft performance in the quasi-real time constraints ($15\div 30$~min after data reception). For this latter purpose, the concept of a model-based monitoring system has been developed that comprises activities concerning scientific data health of an astrometrical satellite which can not be guaranteed by only standard procedures applied to typical space missions. This monitoring system, called Science Quick Look (ScQL), performs the preliminary scientific assessment of the instrument and proper astrometric working of the spacecraft at the (coarse) level of precision attainable at this stage. The prototype of this software is designed in the framework of the DIVA project, providing monitoring, diagnostic and visualization tools. It performs the first scientific assessment of the geometric stability of the instrument and proper working of the spacecraft. The process of the monitoring is based on a model of the Galaxy, on the structure and behavior of the components of the spacecraft and its scanning strategy. The system incorporates a simulator of the observations of stars -- a core of our model, that allows to mimic the work of the on-board software and to simulate star transits. The results of an evaluation of our system look very promising, so we plan to pursue further studies in this area. As the DIVA project was stopped, we will adopt our approach to the next space-based astrometry mission, Gaia, which will be launched in 2012. Indeed, many aspects for the rapid assessment of payload and spacecraft health, developed in this work in the framework of DIVA project, are analogous to those in Gaia due to the fact that the basic principle and geometry of the measurements are the same. A successful completion of the ScQL prototype for the DIVA mission provides the basis for our belief that a ScQL monitoring system for the larger project -- Gaia -- is achievable in terms of the developed concept. Building a ScQL monitoring system for Gaia therefore would become a lot easier if the important steps have already been done in the DIVA project. It is evident, however, that this work has to evolve and grow, along with the concept of the Gaia satellite.