CFHT, Instruments, Spectroscopy, OASIS, Instrument Overview
 

Welcome to the world of OASIS and... Enjoy!

• A Brief History
• The TIGER 1 prototype
• OASIS at CFHT
• Overview
• Observing Modes
• The Imagery mode
• The TIGER mode

• The TIGER 1 prototype

  The Integral Field Spectrograph TIGER 1, first described by G.Courtès, was initiated by a collaboration between Observatoire de Lyon and Observatoire de Marseille. The prototype was initially dedicated to the study of the central regions of galaxies. The first observing run was conducted at CFHT in 1987. Numerous important programs were then undertaken with TIGER 1 and it was rapidly demonstrated that the list of scientific targets for such an instrument should not be limited to galaxy nuclei only but should also include a large number of other astronomical phenomena. Nevertheless, the need for a better spatial resolution (1.2 arcsec in 1987 - 0.7 arcsec in the 1990s) and a larger amount of information (from 80 spectra in 1987 to 350 in 1990) lead to deep modifications of the instrument, and thus great efforts of the teams. Also, due to the very complex structure of the TIGER spectrographic exposures, data processing is a demanding task and had necessited a full-time engineer specially dedicated to the development of the software for several years. In spite of the clear success of TIGER 1, however, the development of the adaptative optics bonnette at CFHT would offer another opportunity for this kind of instrument: gathering a large number of spectra on a small field but with very high spatial resolution. A new integral field spectrograph (IFS) was needed. The result: OASIS!

 

• OASIS at CFHT

  The new IFS for CFHT has been built by Observatoire de Lyon under the supervision of Roland Bacon. OASIS (Optically Adaptive System for Imaging Spectroscopy) was designed to make use of the superb image quality provided by the Adaptative Optics Bonnette (AOB). At the beginning of 1997, the instrument was completed and tested at the Observatoire de Haute Provence. Excellent results were gathered and the instrument was then shipped to CFHT.

The first engineering run at CFHT (photo1, photo2) was completed in April 1997 although very bad weather conditions did not allow any astronomical observations. But on August 10th, OASIS saw its first light at CFHT! During this run, numerous mechanical and optical tests were performed for the f/20 mode and the software for data acquisition was implemented. In January 1998, two engineering nights, used to install a more elaborated user's interface, were followed by four nights of commissioning by a team of external experts (photo3). Tests performed with the upgraded Gumball calibration system revealed a clear improvement in the exposure times needed to obtain good calibration frames. During that period, the data reduction software was also locally implemented and a preliminary reduction performed by the observers was very successful. This software, entirely supported by Lyon, became available for different platforms worldwide in March 1998.

Many observing runs were scheduled for the end of March 1998. Although the observations were again affected by periods of very bad seeing, good science data have been obtained. Two nights of engineering were also used to test the instrument in the f/8 mode and a new beamsplitter for the AOB. This beamsplitter considerably improved the total efficiency of the instrument in the I-band. OASIS has been used in November 1998 (again during bad whether conditions) and in December 1998. During this last run, atmospheric conditions were finally very good and spectacular data were gathered on diverse objects (T Tauri stars, proto-stellar disks in the Orion nebula, M31 nucleus, AGNs, etc.). In 1999, many runs with OASIS were again compromised by bad weather but some nights were also extremely productive. At the beginning of 2000, a long 20 nights run has been very successful.

Despite the fact that OASIS was designed to support six observing modes, only two (imaging and TIGER) will be available for the current year. Further implementation of the other modes is under study at the moment.

In time of writing (June 2000), six new filters have been delivered allowing more spectral configurations. New beamsplitters (V and R bands) have also been delivered and are available for the observations. These devices result in an increase of the performance of OASIS by a factor of 2 in most spectral ranges! Finally, a new 2k x 4k CCD, EEV1, is now available for the visible configurations of OASIS (below 7000 Ang.). This device again improves the performance by a factor of 2, making OASIS a superb instrument for 2D spectroscopy of very faint structures.

OASIS is the result of many years of dedicated work by the Lyon team, supported by the CFHT staff. A list of all the people involved in this project can be found here.

OASIS is a multi-mode spectro-imager, working in the 0.43 µm to 1 µm range and using a CFHT 2kx2k 15µm pixels CCD (i.e. Loral3) or 2kx4k 13.5µm pixels (EEV1). It is operated as a CFHT "guest" instrument. OASIS normally make use of the corrected 90 arcsec diameter F/20 field given by the CFHT Adaptative Optics Bonnette. It can also be used at the direct F/8 Cassegrain focus as a backup mode and for science programs necessiting integral field spectroscopy but with a coarser spatial sampling defined by the natural seeing. (Note: This mode has now been fully tested. It is then VERY important to bring a backup program in case of seeing worse than 1.2", value for which the AO correction becomes inoperative. ).

For now, OASIS can be operated in the following modes:

• The Imagery mode :

It is used primarily for accurate pointing on the objects, for further study in a spectrographic mode, and to provide a pointing "chart" with a known PSF. Image quality has been optimized so that high spatial resolution (0".1) images can be obtained. This offers the flexibility of combining imagery and spectrography programs during a single night.
• 
  The TIGER mode :

It offers a 3D (alpha,delta,lambda) spectrographic capability with low-to-medium spectral resolution, and a wide range of spatial samplings performed by an array of hexagonal micro-lenses. Depending of the configuration employed, the spectrographic field diameter varies from 1.5" to 10" with 0".04 to 0".41 sky samplings. This gives ~1100 spectra of 400 resolved spectral elements; the spectral resolution is R < 3350.

This multi-mode/configuration approach has been chosen:

• To make optimum use of the widely different image qualities delivered by the AOB
• To open an unique observing capability over a broad range of scientific domains.

Due to the high complexity of the instrument, a special user's interface has been implemented in the current Pegasus sessions at CFHT. Instead of selecting the optical elements needed to achieve a specific optical configuration, the observer defines instead an observing ``scenario'' for which he/she specifies the specific instrumental requirements for the science program (e.g. spectral range, sky sampling, resolution)(form1, form2). Different optical elements are then moved automatically by software to achieve an instrumental configuration fulfilling these specifications.

Pointing Limits: OASIS being a long instrument, there are some pointing limits to be taken into account while planning your observations. These limits are implemented in the TCS software so collisions with the telescope pier will be automatically avoided but some objects in your program might not be accessible at all with OASIS! With Loral3 as the detector, the pointing limits in declination are +67d59'15" and -35d54'49". With the EEV1 detector, the limits are more severe because the dewar is longer: +62d and -28d.

Detectors: Two detectors, Loral 3 and EEV1, are available for OASIS, depending of the spectral range required for a science program. Loral3 is a thick 2k x 2k device with a very moderate quantum efficiency but remains the detector to be used for programs requesting the red spectral domain (that is, over 7000 Ang). Its characteristics can be found here. The EEV1 is a thinned 2k x 4.5k device with very good quantum efficiency and low readout-noise, ideal for the visible range of OASIS (i.e. 4200 - 7000 Ang). It should not be used for the red domain because fringing is very important at about 35% at 8500 Ang. Its characteristics can be found here.

Another significant difference between both detectors is the size of the pixels. Loral3 has 15µm pixels while EEV1 has 13.5µm pixels, that is, 10% smaller. This results in a different spectral resolution in the spectroscopic configurations, smaller field in one direction for the EEV1, and different pixel scales for the imaging mode. These are noted in the configuration tables below. The readout time for Loral3 is about 165 seconds. For EEV1, it is about 50 seconds (raster of 2k x 2.5k, useful for OASIS).
 

1) Note that the dispersion is smaller by 10% if the EEV1 is used.

2) Note that the resolution is lowered by about 10% if the f/8 focus - 0.4" sampling is used.

Back to top of the page.

Back to OASIS home page.

Last update: 09/01/1998. Send comments to martin@cfht.hawaii.edu