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

Observation procedures

Overview

An observing session with MOS/ARGUS is quite like a normal MOS session, except that you need to switch between imaging and IFS modes by moving the MOS/SIS mirror and that the pseudo-slit is at 90° from normal orientation of slits in MOS. Hence the idea of having a dedicated "ARGUS" wheel with grisms properly oriented.

There is no separate "ARGUS" session under Pegasus; just open a normal MOS session. A command has been added in the MOS window for dealing with the mirror motion, in addition to the motions of filters, grisms and MOS slide. Be aware that, at the present time, this motion is quite slow (it takes about 3 mn) so you should try to adjust your observing procedure in order to avoid losing time (for instance, switch from IFS to imaging position while pointing the telescope to a new field).

For the set-up:

The internal focus for imaging mode can be adjusted as usual with a pin-hole in position 2.
The alignment of grisms can be done directly on the spectra of halogen lamp through the ARGUS (for instance by following the "dead" fibres used as spacers).
There are two more things to check at the beginning of a run: centering and focussing.

Centering

The centre of field must be checked for each new set-up because mounting is presently not reproducible for a given CCD. And it differs, of course, from one CCD to another. It is probably wise also to check the centering after a rotation of the Cassegrain bonnette; this has not yet been investigated.

For centering, take a direct image of a rather bright star (from guide star catalogue); put it, with the MOS OFFSET command, near the center of the CCD (i.e. X=100, Y=100 if you used a 200x200 subraster centered on 1024, 1024). The star should be in the field of view of ARGUS; for instance, we found the center at X=111.5, Y=110.0 for the last ARGUS run with Loral3, and X=93.0, Y=114.5 for a run with STIS2. Centering is checked by looking, in IFS mode, at the image of the "slit" without grism and with an adequate subraster, like X=100 x Y=2048 pixels. It is not difficult, using telescope offset or MOS offset commands, to have the star centered on the middle fibre of the middle row.

Some useful indications for this operation:

increasing the X coordinate of the CCD will move your object up within the same row of fibers.
increasing the Y coordinate will make it jump from its present row to upper rows along the slit.
the middle fiber of the middle row is usually located close to line Y=1026 on the image of the slit with a 2048x2048 CCD (this could change slightly with each new set-up).
for 15µm pixels, the scale in imaging mode is 1 pixel 0.31", which is a little less than 1 fibre diameter. When you are centered at your convenience with this method, switch back to imaging mode and measure exactly where is the object, using the "centroid" option in IQE, for instance. This will be your centre of field on the CCD for the whole run. Mark it also on the TV screen.
If you need an accurate centering, take an image for each new field, measure the position of your object and apply an OFFSET before the spectroscopic exposure. This insures an excellent centering (within 1 fibre). If your object can be seen on the TV and you do not need a very precise centering, just center it on the TV; this saves the time of an imaging exposure and of a MOS mirror motion (which is not at all negligible, about 3 mn). With this procedure, the precision is about 1", i.e. something like 3 fibres (0.4" each) along a row in X or 3 rows (" each) in Y.
An overlay showing the ARGUS field of view on a SAOimage window has been made by Jean-Paul Kneib. The command should be available soon.

Focussing

The imaging and IFS foci were conjugated (with the Z motion of the focal enlarger) during an engineering night in December 1995, when the seeing was not optimal (

0.8"). For better seeing, a slight difference could be detected, needing a finer adjustment. In May 1996, very good seeing was enjoyed by observers ($\simeq$ 0.5"-0.6") but the detector was STIS2, with 21 µm pixels (i.e. 0.44" per pixel in direct imaging); it was just possible to confirm that the imaging and IFS foci are conjugated within 7-8 units of the Cassegrain mirror motion. Next time we encounter such a good seeing with a CCD which provides a better sampling, we will try to further improve the matching between the 2 foci.

Anyway, it is wise to check these 2 foci separately after a new set-up. This takes only a few minutes and will be done by your support astronomer at the beginning of your run. Also, if you experience very good seeing in imaging during your run, you can check at that time whether there is any significant difference with the IFS focus, then correct for it when doing spectroscopic exposures.

How to do that ?

In imaging mode, use the normal procedure. It would be helpful for that to have a standard bi-prism device in the ARGUS wheel for running CAF. This isavailable since semester 96II.
In IFS mode, with the same bright star and raster used for centering, try to concentrate the light on a minimum number of fibres. This is quite easy to do, maybe with the help of a GRAPH procedure displaying cuts along the slit. Note the difference (if any) between imaging and IFS foci and apply systematically the correction for IFS exposures by changing the telescope focus.

Exposures

Imaging: In MOS window, set ARGUS slide to position 1 and MOS mirror in place. Choose filter, no grism, put Ztel at "imaging" value, and expose.
IFS: Set ARGUS slide in position 3 and MOS mirror out. Choose the grism, put telescope focus to "IFS" value (if there is any difference from "imaging" value, as explained in the previous section) and expose.
Calibrations: Spectral calibrations are done with the "Gumball" calibration unit, but in the IFS configuration. Flat-fields on the dome or with the halogen lamp in "Gumball" are also needed for flat-fielding as well as for calibration in the spatial direction (using the spacers).
Take also spectra of a few spectrophotometric standards.

Multi-slit spectroscopy with the ARGUS slide

Although it is not its main purpose, the ARGUS slide allows also to make "classical" multi-slit spectroscopy, with the MOS mask installed in position 2 and MOS mirror in place. This could be interesting for some programmes; for instance studying the environment of a quasar may include IFS on its host galaxy and multi-slit spectroscopy of companion galaxies. The mask is prepared as in a normal MOS session, except that the slits should be cut at 90°. This is obtained by permutation of the slit "width" and "length" in the LAMA software. The focus for multi-slit spectroscopy is, of course, the same as in imaging.

Be careful, if you use this function, when putting the mask-holder in the slide: this has to be done with the whole device still "tied" to the telescope, which is less easy than in the 5th floor lab.

Also, we must know in advance if you intend to use this mode (which is not normally expected with an ARGUS run), because some preparation has to be done during the setup (image of a cross, etc...).

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

Send comments to: veillet@cfht.hawaii.edu

Copyright © 1997, CFHT Corporation. All rights reserved.