Gaia: taking the Galactic census

Gaia spacecraft Gaia will map the structure and unravel the formation history of our Galaxy through a stereoscopic census. It will measure parallaxes and proper motions for every object in the sky brighter than magnitude 20 — amounting to 1 billion stars, galaxies, quasars and solar system objects. The astrometric accuracy will be 12–25 μas, depending on colour, at 15th magnitude and 100–300 μas at 20th magnitude. Multi-colour photometry will be obtained for all objects by means of low-resolution spectrophotometry between 330 and 1000 nm. In addition radial velocities with a precision of 1–15 km/s will be measured for all objects to 17th magnitude. Gaia is scheduled for launch in late 2013 and the processing of the data will start immediately thereafter. The catalogue is expected to be completed in 2022.

Gaia is an ESA mission and the prime contractor for the development, building, and testing of the spacecraft and payload is EADS-Astrium. The on-ground data processing will be undertaken by the Gaia Data Processing and Analysis Consortium (DPAC). More information on Gaia can be found at the links below. The following sections describe the Dutch contributions to the Gaia mission preparations.

Photometric instrument algorithms

Gaia
      instruments The last part of the optical train of the Gaia telescopes is shown on the right together with the focal plane containing the detectors for the three instruments of Gaia. The photometric instrument consists of two low-resolution fused-silica prisms dispersing all the light entering the field of view. The Blue Photometer (BP) operates in the wavelength range 330–680 nm; the Red Photometer (RP) covers the wavelength range 640–1000 nm. Raw
      photometric data The next image on the left is a simulation of a crowded stellar field as observed by Gaia's photometric instruments. The image of each star is smeared into a little spectrum. From these dispersed images one dimensional BP and RP spectra have to be extracted. Simulated versions of the latter are shown in the image below. The two panels show BP (left) and RP (right) spectra for a range of main sequence stars (O5, black lines, to M5, red lines). From these spectra all sources observed by Gaia can be classified and parametrized in terms of astrophysical parameters.

Reduced photometric data
The group in Leiden (Paola Marrese, Giorgia Busso, Anthony Brown) will develop the data processing algorithms to extract the BP and RP spectra from the raw dispersed images. In addition algorithms for deriving colour parameters describing the source SEDs will be developed. The main challenges are the treatment of CTI effects (see below), the calibration of the PSF, and the disentangling of overlapping images in crowded fields. Algorithms for the latter task are developed in collaboration with groups at the INAF observatories in Rome and Teramo. Scott Trager in Groningen is involved in the ground based spectrophotometry of standard stars, which will be used to put the photometric measurements on a physical scale.

Modelling of CTI effects

A major issue for Gaia is the increasing effect of charge transfer inefficiency (CTI) caused by radiation damage to the CCDs accumulated during the mission. Solar wind particles passing through the CCDs create `traps' which can hold electrons in the same position on the CCD for a while before releasing them again. This has two effects on a PSF image which is clocked through the (4500) CCD lines during time-delayed integration operation: (1) a fraction of the signal charge will be lost and (2) The PSF will be distorted into an asymmetric shape. The image below illustrates these effects (highly exaggerated). CTI
      effect on PSF

The flux losses will cause a decrease in the precision of all Gaia's measurements. For the astrometric data the PSF distortion will lead to systematic errors in the centroid measurement of the images from the astrometric field in the focal plane. A major complication is that the CTI effects are highly variable with the pixel illumination history. This is illustrated in the image here below. The stars travel from left to right across the focal plane leaving charge trails behind them (electrons released after having been trapped). These trails will affect the CCD state (amount of traps already filled) seen by following stars. Hence the CTI effects on a given image depend on the previous images that crossed the focal plane. In addition there will be periodic charge injections (vertical lines in the image) which are used to `reset' the state of the CCD (by filling traps) but they leave their own charge trails which will affect the stellar images and the background in the measurements.
CTI
      illumination history

The PhD research of Thibaut Prod'homme is focused on the empirical and theoretical modelling of CTI in astronomical CCDs. The results will be incorporated into the radiation damage mitigation algorithms for the Gaia data processing.

Physical modelling of spacecraft attitude

Plot of
      torque due to solar radiation The research of Daniel Risquez focuses on the incorporation of the detailed physical understanding of the dynamics of a continuously rotating space platform into the attitude modelling for space astrometric missions such as Hipparcos and Gaia. The highest quality for the attitude modelling will be needed to reach the astrometric goals of the Gaia mission, as the reconstructed attitude provides the reference frame relative to which the astrometric measurements are obtained. The results of this study will be used in understanding the dynamics of the satellite once it is in operation, and to incorporate, in as far as possible, that understanding in the core astrometric processing for the Gaia mission.

As an example of the kinds of effects studied, the plot on the left shows the expected torque acting on the Gaia satellite due to solar radiation. Photons from the sun impact on the sunshield and transfer angular momentum to the spinning satellite, thus changing the orientation of its spin axis if nothing is done to compensate. The torque is calculated over one spin period (6 hours, 'Scan Phase' shows the Sun's position with respect to Gaia's spin axis). Six different sunshields are analysed, with different surface roughness and overall misalignments.

Dutch industry contributions

TNO: Basic Angle Monitor

Basic angle monitor OMA Like Hipparcos Gaia will observe the sky in two directions simultaneously, separated by the basic angle of 106.5°. The astrometric accuracy is highly dependent on the stability of this angle. Gaia will be equipped with a Basic Angle Monitoring (BAM) system, to monitor this angle with a precision of 0.5 μas. The BAM consists of two laser interferometers. Two pairs of parallel laser bundles are sent to the two telescopes, which create two interference patterns on a detector. If the basic angle varies, the interference patterns will shift.

TNO is responsible for the BAM Opto Mechanical Assembly (OMA). With the BAM, an Optical Path Difference (OPD) as small as 1.5 picometers RMS can be measured. To fulfil the stability requirements for such accurate OPD measurements, the entire BAM OMA is constructed from Silicon Carbide (SiC). Ultra stable cryogenic mounts have been developed for mirrors and beam splitters. The BAM OMA shall have an extremely small wave front error, less than 25 nm RMS over the entire optical path. TNO has developed processes for the polishing of SiC mirrors to very low surface errors.

TNO: Wavefront sensor

Wavefront
      Sensor TNO is also responsible for the wavefront sensor (WFS) which will be used after the launch of Gaia to monitor the wave front errors of the two Gaia telescopes mounted on the Gaia satellite. The image shows the Qualification Model of the Gaia Wavefront Sensor. The WFS system is based on the `Shack Hartmann' principle and must have have low optical aberrations itself. Gaia will operate over a broad wavelength and in cryogenic conditions (450 to 900 nm wavelength band and 130 to 200 K operation temperature). For these boundary conditions a temperature independent solution of Invar was selected, with fused silica optics, with the least number of dispersive elements in the design.

Bradford Engineering BV

Bradford Engineering delivered the Gaia Fine Sun Sensors.

Dutch Space

Dutch Space developed software and hardware which was used during the extensive Gaia test campaign. In particular Dutch Space was responsible for the Real-Time Simulator for Gaia.

The activities describes above are or have been funded by NWO, NOVA, and the Marie Curie research training network ELSA (European Leadership in Space Astrometry). Figure credits: EADS-Astrium (Gaia instruments), C. Carreau (Spacecraft), A. Short (CTI figures), TNO (BAM, WFS). More information can be found at:

Other space astrometry missions

Below other astrometric missions are listed which are currently under development by JAXA. The JASMINE mission will be highly complementary to Gaia, and will benefit from the reference frame established by Gaia. JASMINE will be able to map the inner regions of the Galaxy much better as it can see further through the dust. Nano-JASMINE is a precursor to JASMINE, aiming at an astrometric accuracy of several milliarcseconds for the brightest stars in the sky. The Gaia AGIS data processing system, which will carry out the core astrometric data reduction, will also be used in the Nano-JASMINE data processing.