... much^1.1

A short review of the most significant steps in the history of astrometry is given in Appendix B.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... area^1.2

The angular resolution may alternatively be increased by means of interferometric techniques, e.g. by combining the signals from two or more telescopes. Ideally, one should thus obtain a resolution equivalent to that of a telescope with an aperture equal to the separation of the telescopes, or baseline. This is however fairly difficult to realize from a technological standpoint.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

...35 mas^1.3

Hereafter, mas and $\mu$as stand for a thousandth and a millionth of second of arc, respectively. The abbreviations and conversion factors of angular units used in this study are summarized in Appendix A.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... community^1.4

A short history of ESA space science activities is given in Appendix C.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... Hipparcos^1.5

Hipparcos is both an acronym for HIgh Precision PARallax Satellite and a dedication to the second century BC Greek astronomer Hipparchus, whose name is generally associated with the first ``modern'' catalogue of star positions.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... Tycho^1.6

After the 16th century Danish astronomer Tycho Brahe (1546-1601), who discovered the nova in Cassiopeia in 1572 and compiled the most accurate pre-telescope era catalogue of positions of stars.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... accuracy^1.7

Following the glossary at the end of Volume 1 of [ESA 1997a], in this work accuracy indicates the uncertainty of a measured quantity due to accidental and systematic errors, whereas precision denotes the uncertainty of a measured quantity due to accidental errors only. Accuracy and precision are often used synonymously with external standard error and (internal) standard error, respectively. Finally, bias indicates the uncertainty of a measured quantity due to systematic errors only.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... publication^1.8

To obtain a quantitative estimate of the impact that Hipparcos data had on astronomical research the Astrophysics Data System (e.g. http://cdsads.u-strasbg.fr) was searched for articles published from January 1994 (i.e. about 5 months after the end of the satellite's operations) to December 1999 containing the 'hipparcos' word in the title; 542 articles, i.e. about 90 articles/year, were found.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... imaging^1.9

GAIA was the Greek goddess of Earth worshipped as the universal mother who had created the Universe. More recently her name was taken by James Lovelock for his theory on the interdependency of the Earth's biosphere and biological organisms. It is therefore perhaps appropriate that it is now the name given to an ambitious project to unravel the structure, origin and evolution of such a complex entity as our Galaxy. Unfortunately the original meaning of the GAIA acronym, ``Global Astrometric Interferometer for Astrophysics'', explicitly referred to the interferometric option, so that the mission currently lacks an ``official'' name.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... aperture^2.1

Henceforth, when referring to the size of (a portion of) a surface of the instrument such as the aperture or the focal plane, $x \times y$ will indicate a size of $x$ along scan and of $y$ across scan.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... Mapper^2.2

The Photometric Sky Mapper is often referred to as the last column of the Astrometric Field, but, in consideration of its different function, it was decided to clearly distinguish it in this study.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... readnoise^5.1

The total noise arising from the CCD reading process is usually referred to as read-out noise. However, according to the nomenclature adopted by [Vannier 1999] in the discussion of instrumental noise in GAIA Astro CCDs, the read-out noise is only one of the contributions to the total noise. For the sake of clarity, hereafter the total noise will simply be referred to as readnoise.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

...sample^5.2

Note that in the following sample will be used to indicate both a sample size, i.e. the area of the sky region corresponding to more electronically binned pixels, and sometimes a sample value, i.e. the charge obtained from the readout of the sample area. Thus a sample value is obtained through A/D conversion of the charge accumulated at the serial register of the CCD.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... photometry^5.3

Note that in the following it is assumed that the astrometric calibration, i.e. the satellite attitude and the geometric parameters of the focal plane are available from a previous calibration process, which yields the position of each sample with an accuracy better than about $1~\textrm{mas}$.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... telemetry^5.4

The original proposal actually suggested that the detection of high-surface-brightness regions all over the sky would have naturally brought to the observation of Galactic nebulae as well. This possibility, however, has not been investigated in this study, owing to the suspected serious difficulties in the detection at very low Galactic latitudes originating from the crowding of the focal plane.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... regions^7.1

Simulated flux maps based on HST WFPC2 PC images of three other galaxies chosen following the same criterion are presented in Appendix E, together with a bigger version of the HST image and the simulated GAIA flux map of M100 presented here.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... force^B.1

Actually, the detailed evidence necessary to show that this force was of gravitational nature would not be available for another generation.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... subsampling^F.1

Actually, the more refined version of the drizzling algorithm calculate the overlap between shrunken samples and flux map elements exactly rather than dividing each shrunken sample in a given number of square subsamples, but this does not significantly change the overall results.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... SAG\_CUO\_68^G.1

SAG_AUT_NN refers to the NNth technical report presented to the European Space Agency Science Advisory Group for the GAIA mission by the author specified by the AUT abbreviation. The abbreviation CUO, in particular, stands for Copenhagen University Observatory.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.