Current work on Gaia:
Lund Observatory has a major long-term commitment in the development and implementation of software for part of the Gaia data processing, in particular the so-called Astrometric Global Iterative Solution (AGIS).
This is a central part of the science data analysis for Gaia where the reference frame for the observations is established together with the corresponding instrument calibrations and attitude parameters. Lund Observatory provides the basic algorithms and some of the software for AGIS, which are implemented at one of the Gaia Data Processing Centres, located at the European Space Astronomy Centre (ESAC) outside of Madrid. The scientific team in Lund collaborates intensely with the technical teams at ESAC to build up a software system that will eventually provide the astrometric solutions (together with instrument and attitude data) for about 100 million stars. Details of the methods and algorithms have been published (Lindegren et al., A&A 538, A78, 2012) including the results of test runs with several million stars.
The development effort will continue during and after the mission in order to cope with (as yet unforeseen) complications in the real data. Other on-going research concern the combination of astrometric catalogues, for example Hipparcos and Gaia (PhD student D. Michalik) and the reference frame and fundamental physics aspects of Gaia (PhD student R. K. Bachchan). Starting in 2014 the group will also develop algorithms and software for improved error analysis in the future Gaia Archive. A theoretical foundation for this was developed in the thesis work of B. Holl (A&A 543, A14, 2012; A&A 543, A15, 2012).
The historical involvement of Lund Observatory in the mission:
Gaia was originally proposed in October 1993 by Lindegren, Perryman (ESTEC) et al. as a concept for an ESA cornerstone mission. Called GAIA, an acronym for Global Astrometric Interferometer for Astrophysics, it envisaged using two optical interferometers with 30 cm apertures and 3 m baselines, capable of measuring some 50 million stars brighter than V = 15-16 mag to an accuracy of about 20 microarcsec. (The 1993 proposal can be found here.)
Soon after, it was realized that a much better performance could be achieved with a direct imaging system, using filled apertures and a mosaic of CCDs operated in TDI mode, rather than the complex fringe detection method described in the 1993 proposal. These ideas had already earlier in 1993 been developed by E. Høgren and others for the Roemer mission proposed to ESA as a medium-size mission. After further studies by scientists together with ESA and industry, a mature concept was presented in 2000 to the ESA advisory bodies and eventually approved as a future mission in the ESA programme. Gaia was then no longer an interferometer, and in fact incorporated many features from the Roemer concept, but it retained its proper name Gaia - which is therefore not an acronym. A summary description of the project as it was at the time of approval in 2000 can be found here. Since then the project has continued to evolve considerably, and the reader should consult the official Gaia information sheets for up-to-date information.
Early mission studies
Studies performed by the Lund group during the Concept and Technology Study phase (up to 2000) include predictions of star counts and the detection of binaries, brown-dwarf and planetary companions, and the determination of orbits and masses. It has been shown, for instance, that Gaia will measure the individual stellar masses for many thousand stars, and that the mission will provide a nearly complete census of brown-dwarf companions to stars within 100 pc from the Sun. The impact of Gaia for the determination of astrometric radial velocities has been examined. The main mission accuracy analysis was performed in Lund, using semi-analytical tools developed here and subsequently refined as a consequence of the industrial studies. Numerous technical studies of optics and detector performance have been carried out, e.g. modelling the effects of charge transfer inefficiency due to radiation-induced charge trapping. Aspects of the data analysis for Gaia have been studied, including data-base architectures for an observatory-type facility.