Principle; characteristics

Integral Field Spectroscopy with optical fibres

Optical fibres are now widely used for multi-object spectroscopy; several 4m-class telescopes are equipped with such "redshift machines". But it is less common to use fibres for Integral Field Spectroscopy (IFS).

There are basically 2 different techniques for packing on bidimensional detectors the 3 variables (2 spatial variables: , and one spectral: wavelength )that are involved in a complet spectroscopic study of small extended objects:

One technique uses an array of microlenses. This is the "TIGER" approach. The number of spectra that can be simultaneously recorded is large (up to 1500 for a 2048x2048 detector) but there is a limitation on the free spectral range, inherent to the geometry of the device, and the data reduction is quite complex.
The other method is to use optical fibres for changing the geometry of the image. This is the "ARGUS" approach. The number of spectra can be of the order of 500 with a 2048x2048 detector; the spectral range is limited by the size of the CCD and the first steps in the data reduction are similar to the reduction of long-slit data. At CFHT, the situation is as following :
Between 1986 and 1992, IFS with optical fibres has been made at CFHT with SILFID (Spectrographe Intégral à Linéarisation par Fibres de l'Image Directe). This was a visitor instrument developped at DAEC (Meudon observatory).
Since 1993, an ARGUS device can be installed in MOS for the same purpose. In 1996, it became a guest instrument offered for general use and listed in the "Observing time request" forms for CFHT. It is the aim of the present document to describe MOS/ARGUS and give indications on how to use it.
The future OASIS spectrograph will also have an ARGUS mode. For a while, both MOS/ARGUS and OASIS/ARGUS will probably be available to observers. They could choose one or the other of these instruments, depending on their requirements for field of view and spatial resolution.
A prototype IFS device with 61 long (20 m) fibres was used in Infra-Red with the visitor instrument ISIS.
A current project is to link the AOB to OSIS(R) for providing the equivalent of MOS/ARGUS or OASIS/ARGUS in the infrared with excellent spatial resolution and sampling. For a more detailled review on IFS with optical fibres, see for instance Vanderriest (1995) in the proceedings of the 149th IAU colloquium "Tridimensional optical spectroscopic methods in Astrophysics", ASP Conf. Ser., 71, 209.

ARGUS: principle ?

The idea is to transform a bidimensional area into a long slit. At the entrance side, the fibres are packed into a hexagonal array; at the exit side, the rows of the hexagon are aligned (keeping internal order within each row) into a pseudo-"slit". Then, the dispersion can be done by any spectrograph with a large field of view. The ARGUS fibre optics device is installed in MOS, whose field of view is about 80 mm. The fibre bundle collects the light directly at the Cassegrain focal plane (i.e., MOS mirror off) and redirects it, after the fibres have been re-arranged, for feeding the MOS collimator (see drawings).

The coverage of the detector can be dense. Two adjacent pixels along the slit correspond generally to adjacent points in the focal image and to the same wavelength. So, with the actual sampling (0.4" per fibre), there is no need for a careful separation between spectra; some mixing can be tolerated and taken into account during the data reduction. This is not the case with the TIGER geometry, hence the need for a sufficient separation of spectra on the detector with this instrument. On the ARGUS "slit", the condition of spatial proximity is not fulfilled when one jumps from one row of the hexagon to the next one. The different rows are thus separated by "dead" fibres, used as spacers; these few spacers are also useful for the data calibration.

The free spectral range is the same for each fibre and depends only on the grism and on the size of the CCD.

ARGUS: optical and mechanical characteristics

Mechanically, the device consists of 2 parts:

a "front optics box", which contains the focal enlarger optics mounted on Microcontrol plates and holds the front end of the fibre bundle; it is attached to the bottom plate, below the mirror carriage in MOS/SIS.
a MOS slide, to be inserted normally in MOS in front of the collimator. The 2 parts are linked with a "shower hose" style flex which contains and protects the fibre bundle. The total length of fibres is 80 cm. They are in silica.

[Top] [Prev] [Next] [Bottom]

Send comments to: veillet@cfht.hawaii.edu