3 Comparison with other surveys

Some ground-based stellar variability surveys, covering most of the HR diagram parameter space, have been carried out over the last years, but with poorer time resolution than can be achieved with FGS photometry. Frequently, the discovery of new variable stars is a by-product of the search for brown dwarfs and planets. Many projects (e.g., MACHO, EROS, OGLE, MEGA, AGAPE, PLANET) are monitoring photometrically several million stars to detect gravitational microlensing effects, for example in the Large Magellanic Cloud, the Galactic Bulge of the Milky Way and the Andromeda Galaxy. With these large-scale surveys it is possible to discover an enormous amount of new variables, such as RR Lyrae, eclipsing binaries, Cepheids, Miras and $\delta$Scuti stars. As an example, the Catalog of Periodic Variable Stars in the Galactic Center (Udalski et al. 1997) contains more than 2000 objects and is one of the results from the Optical Gravitational Lensing Experiment (OGLE).

Other projects are dedicated to all-sky variable star studies like the All-Sky Automated Survey (ASAS), which monitors about 10⁷ stars brighter than 14$^{\rm th}$ magnitude over the entire sky, but with a temporal resolution of only about one hour. The Robotic Optical Transient Search Experiment (ROTSE) recently reported 1950 variable star detections; 90% of these are not present in the General Catalog of Variable Stars.

Two other very interesting surveys are the ground-based Geneva Photometric Survey (Grenon 1990) and the HIPPARCOS/TYCHO space missions (Eyer & Grenon 1997). The Geneva Photometric Catalog contains about 29 000 stars that were observed more than three times, typically separated by several days, with a precison of about 10 mmag for a 6$^{\rm th}$ magnitude star. Stars with larger than expected variance were systematically re-observed to confirm their variability.

The HIPPARCOS photometry is a complete survey of stars brighter than 8$^{\rm th}$ magnitude and contains 118218 objects with an accuracy of 0.2 (6) mmag for a 5$^{\rm th}$ (9$^{\rm th}$) magnitude star. The detectable periods range from one hour up to 1.5 years. On the same spacecraft, the TYCHO (star mapper) experiment obtained brightness measurements for more than a million stars (complete down to 10$^{\rm th}$ magnitude), but with lower intrinsic accuracy (25 mmag for a 9$^{\rm th}$ magnitude star). Spectral types are known only for a few percent of the TYCHO target stars, but the B-V values could be used to estimate the spectral classification of the other HST guide stars.

Compared to these programs, the FGS data provide a survey of stellar variability which is complementary in the sense that it is focused on short time scale brightness changes of on average fainter stars. Photometric accuracies of better than 500 ppm (0.5 mmag) are obtained for stars down to 11$^{\rm th}$ magnitude, if observed for a few hours. The extremely high sampling rate allows to detect short time scale variations such as rapid oscillations, found for instance in a certain group of Ap stars (roAp, Kurtz 1990). However, a statistical investigation on the occurence of micro-variability in the HR diagram can only be done if spectral types and luminosities (or $T_{\rm eff}$ and $\log{g}$ values) are known.