ASPCS
 
Back to Volume
Paper: Spectrographs for the Measurement of Radial Velocities
Volume: 185, Precise Stellar Radial Velocities, IAU Colloquium 170
Page: 1
Authors: Baranne, A.
Abstract: A radial-velocity measurement derives from a shift in position of spectral features at the focus of a spectrographic instrument. We do not often think about how small these shifts are. It is not generally appreciated that the accuracy to which this shift must be measured is a tiny fraction of a pixel. Or, if we prefer to calculate in microns a surprising minuteness. What precautions should we be taking for the measurement of such small shifts? It is true that, thanks to computers, modern reduction methods allows us to correct for a wide variety of pertubations, provided that these are foreseen and understood; but such reduction procedures will give the best results if such pertubations are kept very small. We must therefore analyse these pertubations and think about how we can control them. The correlation method initiated in its modern form by Roger Griffin, and which we developed further with an optical mask in CORAVEL twenty-five years ago and more recently with a numerical mask in ELODIE, has demonstrated its power. In terms of these methods, the problem of high precision is to improve the correlation peak. Can this be done? Does the correlation method allow us to distinguish the overall radial velocity of the object from possible distortions of the lines? This is certainly a major problem which must be solved. The luminous efficiency of high-precision spectrographs is low. If the use of an optical fibre with scrambling for feeding the spectrograph seems inevitable to us today, it seems to me that the transmission of this system can be considerably improved by a better choice of the F-ratio of the image beam of the telescope which is to be matched with that of the spectrograph. This problem, common to all spectrographs, could be resolved with a specialised focal-plane instrument, giving a much greater than usual F-ratio, resulting in a simplification of the spectrograph optics, and hence an improvement in transmission and a serious decrease in size (which is most important in the case of high precision instruments which tend to be very large). Can we dream then, of a multiple instrument which could monitor several objects simultaneously and independently in a relatively large field?
Back to Volume