1 Introduction

The DIVA satellite was proposed in 1996 by a German consortium of astronomical institutes (Bastian et al. 1996; Röser et al. 1997) and is currently foreseen for launch in 2006. DIVA will measure the positions, brightnesses and proper motions of some 35 million stars. The scientific goal is to study the Milky Way and to improve the calibration of stellar properties and parameters. This mission follows up on the HIPPARCOS satellite which measured parallaxes for 100 000 stars. For about 20 000 of these stars, the accuracy in parallax was better than 10%. With the DIVA satellite this number of stars will be increased by at least a factor of 25 (Röser 1999).

DIVA will perform an all-sky survey with a limiting visual magnitude of $V \simeq 15.5$ mag. Note that every observed star will be measured about 120 times in the course of the mission. The stated magnitude limits refer to the combined images of all single measurements. The measurements include the precise determination of positions, trigonometric parallaxes, proper motions, colours and magnitudes. For about 13 million stars, spectrophotometric data will also be obtained down to a visual magnitude of $V \simeq 13.5$. An additional UV telescope will perform photometry in two spectral ranges adjacent to the Balmer jump.

Figure 1: Wavelength versus position on the spectroscopic CCD (SC) for the first, second and third order. The zeroth order which is made up of undispersed (white) light would lie at about pixel position 122, but is not shown.

The DIVA survey represents a large scale and deep astrometric and photometric survey of the local part in our Galaxy. The importance of these data to modern astrophysics will be significant, with applications ranging from stellar structure and evolution to cosmological aspects. Examples are a precise determination of the luminosity function in the solar neighbourhood, a better understanding of the structure and formation of our galaxy, the estimation of the amount of dark matter as well as a better calibration of the cosmological distance ladder (Röser 1999).

After the mission the photometric and spectrophotometric images will be used to obtain the brightness, the colour and the DISPIs for the stars. The DISPIs will allow to derive the astrophysically relevant parameters $T_{\rm eff}$, $\log g$, [M/H] and E(B-V).

Especially the derivation of $\log g$ is of importance for objects too distant to result in an accurate parallax. With these objects in mind, we have carried out the present study. We will demonstrate that the essential parameters of the stars can be retrieved with a reasonable level of accuracy from the DISPIs alone. We will show that astrophysical parameters can be well derived down to the survey limit, perhaps even adequately for stars 1 to 2 mag fainter. There are good scientific arguments to reach fainter in selected fields, see e.g. Salim et al. (2002).