MAY 8-10, 1996, QUEBEC CITY

This is the public version of the SAC report for its 49th meeting. Note that some sections dealing with internal issues, and of no interest to the community, have been removed for this version. Members of the SAC present at this meeting were: Claude Catala (chair), Jean-Gabriel Cuby, David Hanes (vice-chair), Paul Hickson, Esther Hu, John Hutchings, Gilles Joncas, Pierre-Olivier Lagage, Alain Mazure, Richard Wainscoat. The open sessions of the SAC meeting were also attended by Pierre Couturier, Director of CFHT, and Derrick Salmon, Director of Engineering. The agenda for the 49th SAC meeting was as follows:


  1. Director's report
  2. Report of 1995 December Board meeting, for items concerning SAC
  3. High angular resolution
    1. status of AOB
    2. status of OASIS
    3. fiber-link between AOB and OSIS
  4. Instrumentation plan
    1. present status
    2. wide-field imaging in the visible
    3. science drivers for IR high-resolution imaging
    4. science drivers for wide-field IR spectroscopy
    5. Wide-field IR imagin
  5. Long-term future of CFHT
  6. Other items
    1. Acceptance of OSIS
    2. OSIS observation simulator
    3. Status of f/35 focus
    4. Block scheduling
    5. CNRS/PPARC meeting on IR astronomy

Note that the material under 'other items' is not included in this version of the SAC report.

SAC Recommendations

Recommendation on AOB
Recommendation on fibre link between AOB and OSIS
Recommendation on wide-field optical imager
Recommendation on OSIS IR dectector
Recommendation on a wide-field IR camera

1. Director's report (by D. Salmon, Director of Engineering)

CFHT Technical Activities November, 1995 through April, 1996

1.1 Building and Dome

In December a safety inspection of the 9 ton crane used for handling telescope upper ends disclosed that the crane is under-rated for these loads. Re-engineering is underway with the manufacturer to upgrade rated loads to 12 tons. Crane modifications and certification are scheduled for mid-May this year. We will add load meters and overload protection at the same time. The load readouts will be useful during upper end 'landing' operations, in particular to forestall handling accidents such as those recently experienced with the IR upper end.

The dome shutter continues to work with only 7 of its 8 motors in service, as it has since 1991. A new electrical control system is in place but is not currently connected. Eight replacement motors are on hand. At least one new motor will be installed this year to fill the 8 motor compliment. The schedule and plan for replacing the remaining motors is not yet clear.

An automatic electric power transfer switch was put into service at the summit. Upon failure of HELCO power the switch starts the CFHT diesel generator and switches the building to generator power. After HELCO power is re-established the switch transfers the building back to HELCO power and cycles the generator off.

The building air system responsible for primary mirror support and the Johnson Controls units used throughout the building has been upgraded to continuous 24 hour service. The air system is now cycled automatically between a pair of compressors/air dryers . This clean air source is currently used for CCD dewar window dry air flushing and will soon be used as the source for the f/8 secondary mirror support system air pressure.

The staff elevator has failed several times over the past months and has been out of service for several weeks at a time. Failures were traced to hydraulic metering valves and elevators electrical control relay switch banks. Since elevator maintenance is strictly regulated by the State, repairs have been both expensive and slow. The elevator is currently back in service. We are looking into the cost of an upgraded control system or at least the replacement of all 90 relays in the existing controller.

A new PC-based preventive maintenance software package is in use by the summit crew for generating pm work orders and maintenance logs.

1.2 Telescope

Repeated failures of the Prime Focus Bonnette control system have been traced to component failures and to intermittent cabling problems which have not yet been fully diagnosed. Access to critical control components, especially for real-time maintenance, is difficult since much of the electronics is located under the floor of the PF cage. The existing P.F. Bonnette controls will be very difficult to integrate into the TCS IV control system. As a result the technical staff strongly recommends that the controls be redesigned and rebuilt.

A logic error in the control of the field pick-off x-y stage in the cassegrain bonnette required modifications to control algorithms to prevent the stages from traveling outside safe limits during simultaneous operation of both the x and y stage motors.

Incorrect mounting of the cassegrain environment rotation encoder, probably during recent primary mirror cell handling, resulted in occasional rotation offsets which resulted, for several observing runs, in the T.O.s inability to acquire guide stars using the HST guide star tool. The encoder assembly has since been returned to correct operation.

Failure of air dryers for the support of both the primary and the f/8 secondary mirrors resulted in loss of air pressure due to ice in air lines. Both units were repaired. We hope to connect the f/8 secondary air pressure line to the same supply as used by the primary mirror later this year.

The coma mapping and correction project has largely been stalled by bad weather. We have evaluated a user-friendly wavefront analysis program from Malcolm Northcott (Laplacian Optics) which greatly speeds the reduction of defocussed imagery used for wavefront evaluation.

Upper end handling and the IR upper end accidents

The f/35 IR secondary mirror and chopper/rotator mechanism fell from their mounts on the IR upper end during an upper end-to-telescope landing operation on January 4. The fall was arrested within 1 meter by electrical cables. The immediate cause of the accident was the absence of 12 rotator bearing retaining screws which apparently had not been in place since delivery of the mechanics to CFHT in 1979. Although initially not understood, a related cause was mechanical shock resulting from an abnormally severe upper end landing during the exchange. The accident removed the f/35 secondary mirror and its associated chopping drivers from service. Damage to the mirror consisted of 3 chips on the optical face in an area normally hidden by a central cone, and the fracture of a glass mirror support ring on the back of the mirror. The mirror has since been repaired by Contraves. Interferometer tests show no significant change in the optical surface compared to original delivery. The mirror is currently being coated and should be returned to CFHT by mid-May in time for integration on the chopping system for July observing runs.

An attempt to put the (old) f/36 IR secondary mirror back in service as a backup failed on January 6 when again a mechanical shock, this time during handling operations at the 5th floor level, resulted in breakage of the glass connecting stem at the back of the mirror and in three large chips from its back face.

As a result of these accidents a technical emergency response policy was issued. A review of the upper end handling process identified a number of deficiencies. These included the need for maintenance of the upper end locking mechanisms, insufficient training of new staff, the lack of clear handling procedures, and increasing reliance on judgement in the absence of functional safety devices. For subsequent upper end exchanges we developed written upper end handling procedures and issued a list of personnel certified by CFHT to participate in upper end exchanges. We also developped a list of the functional systems required for an upper end exchange to proceed. The open and frank discussions among the staff on the then-current mode of operation was extremely helpful in identifying and resolving these issues.

Inspection of the locking system on the Upper End Handler and of the 'bird's head' locks on the top ring of the telescope showed that neither these mechanisms nor their associated safety lights were working correctly. Subsequent maintenance has returned the handler to full operation. Although they can now safely be used, further action is required on the top end locks to return them to full operation.

Primary Mirror Support System Failures

The rubber membrane on a single primary mirror support puck failed in October, 1995 resulting in the loss of mirror support and associated poor image quality. The mirror was removed from the telescope under 'emergency' conditions the next day for puck repairs. Inspections of all 24 puck membranes showed that the majority were cracked to some degree. Those membranes with cracks were replaced. The telescope was returned to service within 36 hours of the initial primary mirror removal. All membranes including spares were roughly 10 years old at that point.

Replacement membranes were not readily available from the original vendor since neither the molds nor the original rubber formulation were available to them. Because of the high quoted cost for replacements and what we felt was an overly long delivery schedule we opened procurement to other vendors. Although the final vendor was identified relatively quickly, orders for the mold and membranes were delayed on the vendor's advice pending test of the failed units and recommendation for a rubber formulation. Apparently mold characteristics depend to some degree on the rubber to be used.

Tests of failed membranes by a specialty testing lab - the Akron Rubber Development Labs - identified the failure mechanism as ozone exposure related to age. The same lab analyzed the rubber composition and recommended small improvements to the formulation. The mold and membrane orders were placed in January, 1996. A sample membrane was received and tested at CFHT in February, 1996.

A second failure of a mirror support pad occurred in mid-January. Inspection of all membranes showed continued degradation. Since we were limited then to 7 spares only the 7 most severely degraded membranes were replaced.

Sixty new membranes were received in mid-March and were put into service during rescheduled engineering time on April 2/3, 1996. The process again required the primary mirror to be removed from the telescope. We are hoping to re-plumb the cell at some point so that individual puck failures won't immediately shut telescope operations down.

1.3 Instrumentation

1.3.1 MOS/SIS

The ARGUS mode of MOS was refocussed on the sky to remove telescope focus changes between MOS direct and MOS-ARGUS modes.

Several reports have been received regarding the inability of SIS/OSIS guiding to communicate low frequency track-rate corrections to the TCS. These occurrences were followed by observing runs for which low frequency TCS correction was successful. The changing nature of the problem led us to suspect that it might be related to conditions such as a change of a cassegrain rotation angle. A thorough test of OSIS guiding on the sky during engineering in mid- April demonstrated that the system works well. The problem is now closed although work is still planned to make the system more "user- friendly".

As discovered for the Woodgate run, OSIS guiding is not available for Fabry-Perot observation when using filters mounted in the mask slide at the telescope focal plane. The facility for mounting filters in the mask slide was an operational afterthought introduced after delivery of MOS/SIS to CFHT. In order to maintain camera focus and a collimated beam at the Fabry-Perot use of these filters requires the telescope focus to be changed. The resulting defocus at the OSIS guider head renders fast guiding impossible. We have not yet devised a remedy.

Occasional crashes of the MOS/OSIS controller during OSIS guiding have been traced to a bug in the controller's communications handler and not to the fast guider algorithms per se as previously suspected. A work-around, which avoids communications while fast guiding is in effect, has removed the problem.

Two new IR grisms were tested before and during the OSIS engineering run. The grism for the J-band (R=400) worked well. The H-band grism (R=450) was fabricated with the wrong wedge angle with the result that the desired central wavelength is displaced to the field edge. The manufacturer has agreed to fabricate a new H-band grism at no cost to CFHT. Delivery is expected by mid-May. Two additional high resolution grism for OSIS J-band and H-band observations (R= 1400 and 2000 respectively) should be received at about the same time.

A pair of bi-prisms to aid in spectrograph focussing have been ordered. Delivery is expected by June. One prism is planned for the Argus grism wheel, the other as a replacement in the MOS / OSIS grism wheel.

Details of the SIS upgrade to OSIS are covered in section 1.6.2 of this report.

1.3.2 LAMA

A new x-y stage controller using the pre-existing stepper motors has been successfully put into service. The older system was replaced because of our inability to obtain replacement parts at reasonable cost on short notice. The old controller also had an awkward interface and programming language which significantly slowed data transfers and overall cut speed. The new controller, based on Galil hardware similar to that used in several other CFHT devices, together with control software upgrades permits straightforward downloading of both mask cut-files and AUTOCAD (HPGL) plot files. The new controller has provision for a third axis to be used for auto-focus. Auto-focus will be implemented as time permits.

A new laser-safety shield is now in place. Observers no longer need to wear protective eye glasses while cutting masks.

1.3.3 Coude f/4 Spectrograph

DUCK controller software has been upgraded to the latest version. The new software permits the implementation of a watch-dog timer for quick response to failures by PEGASUS. The facility for background error checking is also provided, although use of the function has not yet been implemented.

Modeling of focal plane aberrations led to the identification of a previously unidentified 1 mm position offset of the prism lens in the UV spectrograph. The studies were undertaken to determine the origin of spectrum shifts between the upper and lower portions of the spectrograph beam which limited the extent over which we could obtain good spectra to under 30 mm. An adjustment of the prism lens in its cell corrected the problem.
1.3.4 FTS

A signal cable from the preamplifier has been identified as the cause of noise in Cryostat #1. A replacement cable will be made for the upcoming runs in July.

150 mV of 60 Hz pickup remains on the InGaAs diode output lines. The noise is not present during system tests on the 5th floor mezzanine where the FTS is normally stored. The pickup appears whenever the power supply for the diodes integral Peltier coolers is plugged into telescope power. Noise levels are identical whether or not the FTS is physically attached or is isolated from the telescope structure. Current levels of pickup preclude the use of these detectors for science.

We have known for some time that the coatings on the IR beamsplitters need to be replaced. Both sets of coatings have degraded seriously over the past several years. The IR1 and IR2 beamsplitter pairs were removed early in the year in hopes of having both re-coated this semester. However, due to coating costs and technical difficulties associated with the CaF2 substrates of the IR2 beamsplitters, this pair will be returned to service this semester in its current condition. The IR 1 beamsplitters have been sent for recoating and should be returned to CFHT by early May.

The BEAR (Redeye spectro-imaging) mode of the FTS is currently limited by a 4.2 second minimum interval between successive Redeye exposures. Considerable effort has been expended by the software group to identify and reduce contributions to this inter-exposure delay. Efforts include the installation of VME memory in the detector host computer and rewriting of the BEAR software handler. Much of this effort is needed for Redeye observation in general and for the KIR camera project. These system upgrades are not currently complete.

Optical misalignment and astigmatism in the BEAR-mode feed optics together with flexure lead to field vignetting and reduced spatial resolution in BEAR. Jean Pierre Maillard will be working with CFHT technical staff in May to address these and other issues in preparation for the July observing runs.

1.3.5 Fabry-Perot

The receipt of the etalon for the Woodgate observing run immediately before the start of the run (as opposed to the generally-requested 1 week advance receipt) highlights the last-minute technical problems which visiting equipment can create. The etalon was delivered without the cabling needed to connect it to our controller. Fortunately a suitable cable was borrowed from the UH facility, at the cost of considerable confusion and delay.

The step range of the CFHT controller did not match what the Woodgate team expected. Considerable staff efforts went into tracking this problem down only to determine several days after the run that the condition was caused by differences in internal options between Woodgate.s and our controller. As a result the CFHT CS-100 etalon controller has since been modified to accept normal and high resolution step control.

To use the STIS-2 CCD on SIS a special one-time-only detector mounting plate was fabricated to adapt STIS-2 to SIS. (Note: This was the last SIS run before conversion to OSIS).

On a brighter note a new etalon mounting wedge was fabricated to support etalons which benefit from reduced ghosts when tipped.

1.4 Detectors

The thinned STIS-2 chip on long-term loan from Bruce Woodgate is currently in operation. This 2048 x 2048 x 21 micron pixel device is mounted in a Luppino dewar. The chip provides an estimated 90.6% q.e. at 700nm, a readout noise of 9 e, and deviation from linearity of <0.1% up to the 120k electron full well. Although originally expected to be an OSIS-dedicated detector because of its large pixel size, strong support from the community has convinced us to provide STIS-2 as a general facility detector. We are currently hoping - as are many observers - to make STIS-2 available on MOS for the start of the May 8 observing run. To accommodate the May and June MOS runs the OSIS grism and filter mounting boxes will be moved to MOS, an extension on the OSIS shutter assembly will be implemented to solve a mechanical interference problem with the MOS camera, and a new detector mount plate to accommodate the larger dewar front end will be fabricated.

Upon the generous suggestion of Guy Monnet, ESO has offered CFHT the long-term loan of a thinned 3-edge-buttable 2048 x 2048 x 15 micron LORAL CCD. An engineering-grade device is now in hand. The science-grade device will be available shortly after successfull tests of the engineering chip. We hope to offer the science device to the community by semester 96II. After the KIR camera project, this project is the detector group's highest priority.

Orbit 1 was returned to Mike Lesser at the University of Arizona in December for replacement of the AR coating. Except for successful use of a limited portion of the chip at the Coude f/4 spectrograph we were unable to obtain reliable UV flooding of the chip to secure the high sensitivity expected of thinned devices. The problem may be related to severe surface contamination evident on many, but not all, test exposures. The likelihood of a successful recoating operation is thought to be 50:50 at best. A thinned Loral device (Loral 5) is currently being integrated into a dewar at IfA, but its scientific utility is in doubt.

The thinned 4096 x 200 x 15micron pixel UBC1 CCD and its stand-alone controller were borrowed from Gordon Walker's group for the March coude f/4 observing runs. Ron Johnson from Gordon's group traveled to CFHT to put the system into operation. Setup and use of the detector on the f/4 spectrograph were straightforward. The device suffers from strong fringing in the red and will be difficult to use beyond 800nm. Initial comparisons with Loral 3 indicate that this device did not provide significant sensitivity gains over a thick Loral device. The absence of significant sensitivity gains may be a result of not having O2 soaked the detector after delivery to CFHT. Tests at UBC to determine sensitivities under conditions of optimal O2 sensitizing are underway. CFHT and UBC are discussing options to mount a thin and a thick device in adjacent locations in a Luppino dewar.

Loral 4, an engineering grade-device, is available in a Luppino dewar. This CCD has been a considerable help for instrument setups since it is currently the only large 15 micron device available apart from Loral 3. Loral 4 is not Lumigen coated and therefore is not well suited for work in the blue.

The order for the long awaited small-pixel CCD from EGG Reticon was canceled in early April on the advice of Reticon. They are unable to fabricate suitable devices.

The 2k x 2k Lick-2 CCD was removed from service after failing last winter. The origin of the initial problem - either in the Photometrics controller or due to the chip itself - is not currently understood. We hope to return this CCD to service using a Gen III controller this semester.

The telescope offset tool used to produce image mosaics with Redeye Wide was successfully tested on the sky in January. A series of vacuum leaks on the Redeye Wide dewar in December and January were traced to repeated failures of the fiberglass-to-metal epoxy seals at the top end of the dewar filling tube. The seal has since been stabilized and has survived several cold cycles. A new seal geometry is being pursued with the original dewar vendor. Intermittent mechanical binding of the filter wheel drive assemblies has been addressed by mechanical re-alignment of the motor assemblies.

The fanout board for the Redeye Narrow upgrade is completed. Mechanics for the new detector mount is 90% complete. Dewar wiring is expected to be completed by mid-May. The multiplexer for the Rockwell 1kx1k Hawaii array is expected to arrive at CFHT in early May. Delivery of the engineering-grade device is expected before the end of semester 96I. If Rockwell can meet this schedule we plan to be able to release the upgraded Redeye Narrow dewar using a single quadrant of the engineering device by semester 96II.

A new CCD dark lab has been outfitted in what was previously the darkroom at the Waimea headquarters. The new lab more than doubles the space available for detector development in Waimea. The lab is outfitted with a new integrating sphere and light sources to permit routine q.e. measurements, an optical test table, a leak detector with a residual gas analyzer for checks of vacuum contamination, and a dry gas manifold for dewar backfilling and sensitizing. Charge transfer efficiencies are now routinely tested using the industry- standard Fe-55 X-ray source and a replaceable Be dewar window.

An order for 3 new Luppino dewars was placed with IfA in February. Delivery of all 3 dewars is slated for August, 1996.

1.5 Computers and Software

The observer's console has been upgraded with the implementation of a full PEGASUS system on the new high speed HP-9000/747 workstation (Neptune). The system features a 3 monitor display, 128 MB of on-board memory and a 5 GB hard disk. The long-term problem with crashes of CFHT Image (a CFHT SAO Image clone) during multiple zoom actions has been solved with the incorporation of a new release of SAO Image.

Gen-III CCD controller software upgrades have improved the minimum inter-exposure overhead for FTS Bear-mode exposures to 4.5 seconds. VME memory has been installed in a detector host as one step in an attempt to further reduce overhead. These efforts are system-wide and are not limited to Redeye or Bear-mode improvements.

Variability of the overhead between Redeye exposures causes signal coming from the warm Redeye dewar shutter to differ between science and reference frames. This variability degraded the science being done with Bear. After considerable hardware/software effort variability has since been reduced from roughly 0.5 seconds to a few tens of milliseconds.

GPIB interfacing via independent network devices is being implemented as a step to removing the need for CAMAC control of the FTS and CC- 200 CCD controllers. The new GPIB net-based interfaces will allow the data logger to function independently of other summit computers and permit the retirement of MOE.

The CFHT router through which all network messages entering or exiting the CFHT network has been configured to filter network packets. This filtering provides a moderate level of security for the CFHT network. This is the first step in upgrading CFHT network security. In addition, benefits of this packet filtering will be to allow some network applications, such as ftp and telnet to be initiated directly from the summit on selected machines. In the past any connection to the Internet from the summit had to be started from a machine located in Waimea.

CFHT WEB pages have been reorganized. A regular mirroring procedure of CFHT html-based instrument manuals at the CDS (Centre de Donnees Stellaires de Strasbourg) has been initiated. This allows fast access to these manuals for the French Astronomical community. This information was, for all practical purposes, unavailable previously in France due to slow network response time. Important information from the CFHT observing log and night report as well as selected measurements from the datalogger are now available to WEB browsers.

Development of X-windows Pegasus-type forms compatible with WEB browsers has allowed straightforward development of Pegasus forms using html. This facility is particularly well suited for remote observing using the WEB for example.

1.6 Major Projects

1.6.1 AOB

The AO Bonnette arrived at the summit from Paris in excellent condition in 14 crates on 8 January. The system was assembled and initially tested in the 2nd floor instrument lab. No discernable optical mis-alignment was evident as a result of the shipping. After roughly a month and a half of lab tests the bonnette was moved to the 5th floor mezzanine for operational tests in the cold dome environment. During these initial test periods considerable effort went to verifying correct AOB operation. The scientific and technical staff put in a similar effort to become familiar with the bonnette's many subsystems.

System tests through the month of February identified two WFC to real-time computer data transmission errors. One class of error was traced to poor impedance matching in the WFS-to-computer signal cables. This problem was solved by the installation of new cables fabricated by DAO. The second error class was caused by faulty indicators of signal overflow from the WFS counters. It appears that this error is caused by problems within the databus circuitry of the WFS electronics. Software algorithms were modified to ignore overflow condition until this problem can be fixed. A high-voltage power supply transformer for the drive of the bimorph mirror failed and was replaced shortly before the first observing run. A test fit of the AOB onto the Cassegrain Bonnette in mid-March required trimming of a mating flange on the top of the AOB in order to provide a comfortable fit.

Integration of the CILAS control software into the CFHT computing environment consumed roughly 1 man month. First-pass development of the AOB User's Interface under PEGASUS was completed during the first observing run.

The AO Bonnette was mounted on the telescope in preparation for the first observing run on March 13. Alignment with the telescope optics was excellent. Universite de Montreal's MONICA IR camera was made available for 96 I commissioning tests by Daniel Nadeau. To produce an f/60 output beam, changes in lens positions inside MONICA were required. MONICA was mounted and aligned to the AOB without problem. As a very encouraging prelude to sky tests, AOB imaging performance with MONICA using an artificial star produced diffraction-limited H- band images with Strehl ratios greater than 90% over the full 90 arcsecond field diameter.

The first observations with the AOB on the night of March 28 were an unmitigated success. The first reference star (V=12) was acquired and the control loop was closed within ten minutes of the start of observations. Although seeing during this run was variable, the first image - in the K-band - was clearly diffraction-limited, with a central diffraction spike 0.12 arcsec in diameter at a Strehl ratio of 30%. Natural seeing at the time was 0.7 arcsec FWHM.

The first engineering run was used for tuning and performance evaluation. Image Strehl ratios from 5% to 15% in J, 15% to 40% in H, and 30% to 60% in K were obtained in natural seeing conditions ranging from 0.6 to 0.9 arcseconds FWHM. Connection of low frequency AOB guiding errors to the telescope control system for autoguiding was accomplished in roughly 15 minutes on the second night. Tests of mode-gain optimization algorithms which are intended to simplify AOB operation resulted in 35% improvement in residual wavefront error. We expect to add a slide-bar control of membrane-mirror gain to further simplify operation. A very straightforward image mosaic tool using MONICA was also implemented on the second night. To give you an appreciation of the ease with which mosaics can now be taken, the process of closing the loop, moving the telescope, re- acquiring the WFS reference star, closing the control loop, and starting the next exposure takes under one minute. The process is mostly automatic. On a less optimistic note, the sensitivity of the WFS appears to be a factor of 2 lower than expected. We are also currently limited to relatively short exposures by flexures inside MONICA which produce image motion of 0.5 pixels (0.018 arcsec) in 5 minutes.

The first engineering run was dedicated to IR observations. CCD images will be taken during the second engineering run to evaluate performance at wavelengths between 0.5 micron and 1.0 micron. To obtain the required image sampling we have developed a 2x image enlarger to be mounted between the AOB and FOCAM. The optics for this enlarger were Sol-Gel AR coated at CFHT.

One of the principal goals of the AOB project was to provide an adaptive optics system useable by non-specialist observers in a straightforward manner. After the first observing run we have every expectation that this goal will be fulfilled in semester 96 II.

1.6.2 OSIS

OSIS is an upgrade of SIS, extending its spectral coverage in the near IR up to 1.8 micron. Compared to SIS, the spectral coverage is more than doubled. The f/ratio is now 8, providing spatial samplings of 0.11"/pixel in the visible for 15 micron pixels (0.15" for the STIS CCD), instead of 0.09"/pixel previously. The field of view and the spectral coverage in spectroscopic mode are increased accordingly. In the IR, OSIS uses the redeye cameras. With redeye wide, the sampling is 0.5"/pixel, and the field of view is 2' x 2'.

The first run in February 96 has been almost completely lost due to bad weather conditions at the summit. However, the basic capabilities of the instrument have been tested in the IR. The transmission in the IR is ~ 60% of the direct redeye transmission at the Cassegrain focus. Two IR grisms provide resolving powers of ~400 in J and ~450 in H for 1" slit. The zero deviation wavelength is 1.252 micron in J, and 1.493 micron in H. This latter value is out of specs (the spec was 1.65 micron), due to a wrong prism angle. This however does not prevent from working if offsetting the slits by the adequate amount on the masks. The scattered light and ghosts have been measured at ~ 0.2% level in both visible and IR domains. The level is slightly more important in the visible than for SIS because the anti-reflection coatings are less efficient over the huge OSIS spectral range. It is strongly recommended for pure imagery runs to use redeye at the direct Cassegrain focus rather than behind OSIS, especially in K where the OSIS transmission is extremely poor.

A second run in April 96, both at visible and IR wavelengths, was successful. Data is being reduced. Many spectra of distant galaxies and quasars have been taken. Halpha emission line on an object as faint as J =18 has been easily detected. The observations have been made by nodding along the slits, combined with shift and add techniques. A detailed analysis of the data will indicate what are the most efficient procedures for observing and reducing the data.

Two high resolution grisms (R ~ 2000 in H and R ~ 1500 in J) are ordered, as well as two cutoff filters which will cut the upper end of the H band so as to reduce the thermal background whenever necessary and compatible with the observation program.

A simulation tool has been developed that can be used for OSIS in its IR configuration, and which may be adapted for use at visible wavelengths (for MOS and OSIS).

1.6.3 TCS-IV

The TCS-IV control system architecture is based on EPICS running under the VxWorks real-time operating system. TCS-IV group members have attended or will soon be attending training courses in both EPICS and VxWorks. A VME crate with a Sparc 5e was initially installed running VxWorks in Waimea last November. A second VME crate currently runs the primary mirror tip/tilt control, and will soon run the focus for the F/8 and F/35 upper ends.

A TCS hardware simulator which mimics the actions of the Marconi R- bus control hardware has been developed in parallel with, but separate from the EPICS-based TCS IV control code. The simulator was tested and verified at the summit in January this year and has been used to verify operational the R-bus driver in late January.

The telescope was slewed using TCS-IV hardware and software during daytime tests in mid-February. Apart from testing low level R-bus driver code these tests provided convincing evidence that the goal of being able to switch readily between TCS-III and TCS-IV control can be met.

A TCS-IV advisory group consisting of astronomers and technical staff has been re-constituted to provide guidance to the TCS-IV development team on issues of basic functionality. A second more focussed working group has been established to define and co-ordinate work on communications between the TCS system and the software used to control science instruments.

1.6.4 Coude Fiber Feed

Some progress has been made in the development of the coude fiber feed system. The cable run from the back of MOS/OSIS to the third floor has been detailed and preliminary designs for the cable wrap at the telescope have been completed. The science fiber lengths will be well under 30m. We are currently planning to run 2 science fibers, with a single fiber available at a time. The design should, however, make it easy for the OA to change fibers during the night as the observer's program requires. A second pair of fibers will provide comparison arc and flat field illumination at the telescope from the current coude f/4 comparison arc and flat field facility on the 3rd floor.

Preliminary designs of the microlens input optics are completed and a design of the acquisition camera optics and comparison lamp feed is underway. The output fiber and image slicer mounting geometries have been defined and some mechanical stages have been received. The final optical layout of the Bowen Walraven image slicers is in progress.

The technical staff has been in contact with fiber vendors and optical houses to integrate the input-end microlens and field view mirror with a commercially available fiber connector. Once the slicer design is complete potential vendors for their construction will be contacted.

1.6.5 Status of OASIS (by Pierre Couturier, Director of CFHT)

On May 17, 1996, CFHT staff and Roland Bacon, project manager of OASIS, have reviewed together the evolution of the development of OASIS.

The optics is completed and tested (except filters). The electronics is in the middle of development phase. For the mechanics, the Serrurier mounting is completed, other parts are being built. The prototype of Pegasus user's interface is completed. Development of the final version is on-going. The software package for Tiger-mode data treatment is being developped.

The final assembly of the mechanics is scheduled for mid-September, and the final assembly of the electronics, for the end of November 1996.

A request for 1.93m telescope time has been addressed to OHP (we know now that two nights only were allocated to test OASIS on the sky in mid-December). CFHT staff will be involved during the final assembly and OHP tests. The details of the acceptance process need to be finalized.

If all this goes well, delivery at CFHT is scheduled for March 1997, the engineering and acceptance will take place on Semester 1997I, and the scientific runs would start on Semester 1997II

By the end of 1996 we will formalize a contract for delivery of this guest instrument. The main points which will be be covered are the following:

  1. OASIS will be owned by the Observatoire de Lyon.
  2. OASIS will go through formal acceptance tests for compliance with CFHT requirements.
  3. If the tests are successfull OASIS will be a guest instrument operated by CFHT.
  4. Observatoire de Lyon asks to have the possibility of using OASIS later on a large telescope, CFHT considers that from the time OASIS will leave our control, it will be operated as a visitor instrument, we ask a minimum of 3 years of operation full time at CFHT to institute the training of our staff to maintain and operate OASIS .

Roland Bacon has emphasized that AOB needs a major upgrade to allow spectral analysis in the visible when the AO is guided on the scientific target itself. He asked that as part of a plan to upgrade AOB CFHT acquire new beamsplitters, in order to use part of the visible bandwith for guiding and the complementary part for transmission to the science device. Acquiring a set of beamsplitters for different wavelengths could cost $k 50.

The budget for shipment and acceptance (~ $k 100) is not covered by the resources presently available at Observatoire de Lyon, and CFHT does not have sufficient funds available in its budget except for CFHT staff travel expenditures (final assembly and acceptance tests). This issue has to be addressed soon with INSU.

Argus mode is not included in the delivery. $k 50 more will be needed.

Pytheas mode is not included because of a lack of funding, scientific supervision and manpower.

Improvement of gumball calibration will be needed. This is a CFHT project not currently in progress.

1.7 Operations

With the establishment of the new management structure in January several project responsibilities were reassigned. Scott McArthur assumed the role of project manager and project engineer for the AOB project. Together with the project scientist, he organized the receipt and commissioning of the AOB very successfully. Greg Barrick stepped in as head of the Optics Group and managed the OSIS upgrade work, again with great success. And David Bohlender assumed responsibility for the Coude Fiber Optics project which is currently stalled for lack of manpower.

A new scheduling and transportation policy has been established for Waimea-base staff working at the summit. Whereas departure to the summit previously occurred at 8:30 am, departure time is now 7:00 am. A single summit departure time for Waimea staff has been set at 5:00 pm. Travel time is paid on the basis of a single round-trip to the summit per week. Although staff are permitted to travel back and forth to Waimea each day we encourage them to overnight at Hale Pohaku if several successive work days are planned. Similarly we are in the later stages of establishing a revised schedule for the summit crew.

A fire at the Subaru telescope building on January 16, which cost the lives of 3 construction workers and sent several others to hospital, focussed our attention on the safety of CFHT personnel. As a result of a subsequent safety review meeting we are planning to implement bi-annual emergency response drills at the summit. Smoke inhalation from burning foam insulation was the cause of death in the Subaru incident. Similar fires involving burning insulation have been reported at Keck. As a result of the speed at which personnel were overcome by fumes at Subaru and Keck we are co-coordinating the purchase of low-cost breathing hoods for staff evacuation with other observatories at the summit. Several of these hoods will be deployed on each floor of the building. After the recent retirement of our previous safety officer, Peter Sydserff, we have appointed 2 staff members as safety officers for the summit and Waimea facilities respectively.

We are currently reviewing the use of the instrument storage and setup area on the 5th floor mezzanine with an eye to moving these facilities to the 5th floor proper. The effort is prompted by awkward access to and the exposed nature of the mezzanine area.

CFHT hosted the 1996 Astronomy Neighborhood Meeting at the Royal Waikoloan Hotel on March 28 and 29. The meeting was attended by 80 astronomers and technical staff from observatories on the islands.

Apart from our obligations to the Wide-Field Imaging Plan, upcoming commitments for the next semester include a major rebuild of the cassegrain comparison arc injection facility which is needed by OASIS. To significantly reduce MOS/OSIS setup times we hope to fabricate several new filter and grism wheels and to secure grisms in the wheels so that they will not be removed for each observing run. The full complement of dome shutter motors need to be replaced. The f/8 secondary mirror will be re-coated in semester 96 II. There is serious technical concern about the longer-term viability of the control system for the Prime Focus Bonnette which needs to be addressed.

1.8 New instrumentation plan

1.8.1 KIR - (1k 1k HgCdTe Hawaii Array)

The order for the 1024 x 1024 HgCdTe array for KIR was placed with Rockwell in late December. Under the agreement CFHT should receive a multiplexer by late April, 1996, an engineering-grade device with at least one science-grade quadrant by May, 1996 and the science-grade device by October 1996. We will use the multiplexer and the engineering-grade device in the Redeye Narrow upgrade. The science-grade device will go to U. de M. for integration into KIR.

We received proposals from the U de M / OMP , the IfA, Ohio State, and Infrared Labs for fabrication of the KIR camera system. The U de M / OMP proposal was selected in February with U. of M. acting as contractor. A draft contract, technical specifications, and statements of work are currently under review at U. de M. Fabrication of the optics is the most time-critical element if we are to meet an end 1996 delivery. U. de M. is currently finalizing the optical design so that orders for glasses can be placed by the time the contract is finalized in early May.

The choice of the computing environment to be provided by CFHT for software and control developments at OMP is a time-critical issue at CFHT. This choice will influence work on several other systems including future FTS and CCD upgrade. A final decision on hardware and the associated Pegasus User's Interface will be made by early May.

1.8.2 CFHT 10k (upgraded clone of LUPPINO 8k camera)

A draft contract for an augmented (8k x 10k) duplication of the UH 8k camera system, exclusive of the CCD's, has been received from the IfA. A submittal to the Contract Review Committee (COMA) is being prepared at CFHT. We hope to have paperwork to COMA by the end of May. Fabrication is expected to be complete by the end of 1996. The CFHT10k camera will contain an upgraded shutter mechanism since the unit currently on the Luppino camera has been an ongoing source of problems. A separate contract with the IfA for the delivery of S-bus interfaces for ultimate use with this camera, but for more immediate use under the loan agreement for Luppino's 8k camera, is in place.

The delivery schedule for thinned 2k x 4k CCD's from Lincoln Lab is unclear at this time. We hope to receive at least one each of the red and the blue optimized devices by this summer. An initial attempt by Gerry Luppino to provide a low-volume edge interconnect so that the Lincoln Lab devices could potentially be used in the MEGACAM project was not successful. He had hoped to mount the individual device edge-connectors at right-angles to the detectors, but this is apparently not possible because of constraints on the substrate geometry.

Upgrades to the summit and Waimea network facilities needed to accommodate the UH 8k and the CFHT10k cameras is underway. We have purchased, but not yet received, 100 Mb/s high speed switches, network hubs, and workstation interfaces which will permit high speed data transmission between the UH 8k camera system and CFHT summit computing facilities. A SUN Ultrasparc system with 512 Mb memory and 18 Gb disks will be installed at the summit by July for handling images from the 8k camera. We have already installed a similar Sun Supersparc system in Waimea. This system will be attached to a 100 Mb/s hub connected to a high-speed data reduction subnet. The Waimea hub and twisted pair net connections should be installed by July.

To archive 8k camera data, the archiving facility currently in Waimea will probably be moved to the summit since high-bandwidth communications to Waimea will not be available until the end of 1996 at the earliest. We are currently investigating the use of high-speed (2Gb/s) digital linear tapes as the primary summit archive facility.

Due to staff workload this semester no actions on the other priorities identified under the Wide-Field Imaging Plan have been taken. We hope that by July 1 a concerted effort can be started at CFHT on image quality studies and the Megacam project.

2 Report of 1995 December Board meeting, for items concerning SAC

This section has been removed, including Recommendation #1 and Recommendation #2

3 High Angular Resolution

3.1 Status of the Adaptive Optics Bonnette

See sect. 1.6.1.

After the presentation of the first results from the CFHT AOB, the SAC members were admirative of the achievements of this project. Several important improvements were also extensively discussed, and the outcome of the discussion prompted the SAC to make the following recommendation:

Recommendation #3

The SAC is impressed by the recent and very encouraging results of the AOB. The SAC congratulates the CFHT staff and the groups who developed this instrument for the success of this project, which provides CFHT with one of the most efficient adaptive optics system in the world. To further enhance the AOB capabilities, and maintain it at the forefront of the international competition, the SAC recommends an immediate upgrade of its performances. The AOB wavefront sensor should be equipped with new Avalanche Photodiodes, providing a gain of at least 2 in sensitivity, resulting in a significant gain in sky coverage. To improve the versatility in visible imaging, as well as in visible spectroscopy -- with the forthcoming availability of OASIS -- the AOB must be equipped with a set of at least 2 beamsplitter dichroic coatings, and possibly also of interference coatings.

3.2 Status of OASIS (report from Executive Director)

See Sect. 1.6.5.

3.3 Fiber link between AOB and OSIS

As pointed out by the SAC in previous reports, it is important to have a spectrographic capability with high spatial resolution, extending in the NIR, that interfaces with the AOB. The only spectrograph available is OSIS. The AOIFU is a proposed link that samples the AOB field adequately and allows full-field spectroscopy with OSIS. The proposal by Crampton and Felenbok is to build this link. The SAC was requested to approve its scientific value so that funding may be sought. The SAC agreed that this proposal is timely and of high scientific value. The extension into the NIR makes it more desirable than the ARGUS mode of OASIS, which would work only at CCD wavelengths. The SAC also felt it would be useful if the link could also accomodate the f/8 beam from the uncorrected telescope focus, to allow a wider field with sampling appropriate to the uncorrected image size. Consequently, the SAC makes the following recommendation:

Recommendation #4

The SAC strongly endorses the scientific merit of providing a fiber-optic link between the f/20 AOB and the OSIS spectrograph, and recommends that the proposal submitted by D. Crampton and P. Felenbok be pursued, with attention to be paid to the additional possibility that the coupler might serve to provide a link from direct f/8 focus to enjoy the widest possible field. In making this recommendation, the SAC is explicitely prioritizing such a fiber-optic coupler above the Argus mode of OASIS, because of its IR capability.

4 Instrumentation Plan

4.1 Present Status

4.1.1 KIR - (1k x 1k HgCdTe Hawaii Array)

See Sect. 1.8.1

4.1.2 CFHT 10k (upgraded clone of LUPPINO 8k camera)

See Sect. 1.8.2

4.2 Wide-field imaging in the visible

The wide field imaging discussion centred on the scientific questions that can be addressed, the field size of various cameras, and how CFHT can achieve a unique capability of long-term value - both in anticipated science and in comparison with other telescopes over the next 5 years or more. It was agreed that a 10 x 12 K array within the existing WFC field, as proposed by the Megacam consortium, was not an effective upgrade over the 8 x 10 k camera currently being built. If a new corrector could be built to accomodate a larger field camera, it was felt that this would constitute a capability that would allow CFHT to pursue competitive research for a considerable time. If such a camera can not be built, CFHT would not have a significantly better capability than other telescopes, and other instrument options should be considered. The list of scientific drivers in this report contains comments on which projects of current importance would benefit greatly from the very wide field (either in making a project viable or in large savings in observing time). It also notes which projects would require enhanced image quality by tip-tilt correction, which would still be important several years from now. We note that any unique facility will also always enable major research programs not foreseen at the time of its conception.

There are several possible problems that might prevent the camera from being realised: WFC performance, CCD technology, image quality, etc. SAC has delegated Cuby and Hutchings to keep the committee informed as the development proceeds.

Science drivers for wide field CCD imaging

Each program or set of projects are qualified by W, S, and L, indicating whether they benefit significantly from:

  1. Wider field than 30 arcmin (W)
  2. Seeing enhancement over ambient (S)
  3. Long term: likely to be current in 5 years (L)

(Where these are uncertain they are given in parentheses)

Note that the 3 classifications are inevitably subjective. While they are an attempt to quantify the situation, the main message is that there are many large programs that the camera will enable with long-term value and use. All represent programs of current interest that have been proposed or discussed. Some will not be done until a wide field camera is in place. Many will benefit significantly from a 1deg field by greater sky coverage or by making the total observing time achievable rather than unreasonable.

  1. Galaxy structure programs W,S,(L) (halo structure, brown dwarfs, mass function, white dwarfs)
  2. Microlensing in local galaxies S
  3. Local group stellar populations S
  4. Structure in disk galaxies S,(L)
  5. Local group galaxy haloes W
  6. Globular cluster statistics in distant systems S
  7. Low surface brightness galaxies, outer regions of globular clusters W, L
  8. Galaxy cluster luminosity functions S,(L)
  9. Galaxy cluster dynamics, morphology S, L
  10. Cosmological field galaxy studies W, L
  11. Supernova searches W, S
  12. Weak lensing studies and surveys (W), S, L
  13. QSO and AGN surveys W, S,(L)
  14. Faint galaxy surveys (high z) W, (L)
  15. Outer solar system objects W,(L)

Notes on the science programs:

  1. This covers many different programs, but all are aimed at a full understanding of the star-formation and evolution history of our own galaxy. This will be a major arena of work in the next few years.
  2. It will be possible to search for low mass components of the haloes of M31 and M33 by microlensing. This will provide a direct comparison with the results for our own galaxy halo now being accumulated, and relate to galaxy formation and history.
  3. 4 These programs will seek to understand the stellar populations and astration processes in the general galaxy population at the current (z=0) epoch.
  4. 6 There are studies of the outer stellar and cluster populations in galaxies and clusters of galaxies. These trace the history of star-formation and merging of galaxies. Current results show great promise but lack statistics of sufficient numbers.
  5. The study of diffuse stellar populations is another aspect of understanding star-formation processes and their effects on galaxy histories. (Topics 1-7 are all under the subject area of understanding how stars make up galaxies and affect their behaviour. It will be a very active area over the next few years.)
  6. 9 These studies will seek to understand the history and workings of galaxy clusters, which are the largest bound systens known, and probably the oldest. Cosmic evolution of these structures will be studied to lookback time fractions of up to ~50%.
  7. This is similar to 8,9 but for the study of field galaxies which do not belong to major structures. Thus their formation and history is different and traces undisturbed galaxy evolution.
  8. The statistics of supernovae which can be seen to large lookback times, offers a very powerful way to study the distance scale and cosmology. This is already under way, but requires better instrumentation to significant progress.
  9. Weak lensing is a powerful way to study the distribution of all matter (including invisible matter) to very large scales and distances. Excellent images over wide fields are essential.
  10. 14 These are surveys which will give us the statistics on the earliest times of galaxy formation, and the cosmic evolution of nuclear activity. Present results are very exciting but lack statistics of large umnbers.
  11. The search for objects in the outer solar system bears on our understanding of the formation of planetary systems in general. Observations of young stars are a link between our own and other systems. The outcome of the SAC discussions about wide-field CCD imaging is summarized in the following recommendation:

Recommendation #5

The SAC considers that the scientific drivers outlined in this report make a compelling case for a very wide-field CCD imager, significantly larger than the 8k x 10k camera currently being built. A full 1 degree square field of high image quality is needed to provide exceptional scientific capability with long-term value. This implies a 16k x 16k array with tip-tilt correction.

The field size requires a larger wide-field corrector of high performance. The corrector should be designed with the following goals:

  1. field diameter ~ 1.4 degree, with low distortion and less than 15% edge vignetting
  2. 0.3 arcsec image quality
  3. wavelength of operation: 0.38 - 1.1 micron
  4. coated optical surfaces
  5. flat focal surface at more than 80 mm behind the rear element.

We request a report on the corrector from CFHT by next SAC meeting. We will review the project contingent upon costs and feasibility of achieving these goals. An option of an ADC should also be investigated. We consider that this camera should perform large survey projects, which should be supported by peer-review process.

4.3 Science drivers for IR high-resolution imaging

High angular resolution InfraRed observations will allow for a better knowledge of many astronomical objects, ranging from nearby objects, such as comets, to objects at the confine of the Universe, such as quasars.

Among the various astrophysical problems which will benefit from such observations, one can mention the problems related to the formation of stars and planets.

For example, the frequency of binaries among YSO's is one of the key parameters for star formation theories. In the Taurus molecular cloud, the binarity among YSO's objects is more common than among main sequence stars. But it is not the case in the Rho Ophiucus cloud. To complete the study to close binaries and to extend it to other star forming regions (which are further away), high angular resolution observations are needed. Observations in the IR are required because of the high extinction due to the dust surrounding the YSO's.

The study of this dust at high angular resolution is also very promising. For example, while there is now some consensus that most of the infrared excess in solar-mass optically visible young stars (the T Tauri stars) arises from a circumstellar accretion disk, the circumstellar environment of their higher mass analogs, the Herbig Ae/Be stars, is still controversial. Models based on the presence of a large dust envelope can be tested thanks to high angular resolution in the IR.

Dust may eventually form planets, so that the study of the dust environment is also important in the context of planet formation. At some stages, both substantial dust residues and planets may be present. Then gravitational footprints of planets on the dust (asymmetries, void of matter..) could be observed thanks to high angular resolution. Direct observations of planets is the ultimate goal for high angular observations. But it is clear that such observations are very difficult.

If we now jump to the extragalactic studies, one can quote the study of the link between starburst formation and AGN's. There are claims that all AGN's are surrounded by a circumnuclear starburst and reciprocally. In order to resolve the starburst region, high angular IR resolution are needed. The angular resolution now achievable with adaptative optics is of the same order as the resolution achieved with the VLA, so that IR observations will complement radio observations.

4.4 Science drivers for wide-field IR spectroscopy

The three main specificities of IR spectroscopy are:

  1. the presence of spectral signatures at IR wavelengths,
  2. less extinction (one sixth in H its value in V), and
  3. redshift effects.

Rather than performing an exhaustive list of scientific programs, we highlight hereafter two examples which make the best use of these specificities:

  1. IR stellar spectroscopy in embedded regions (items 1 and 2 above)
  2. IR spectroscopy of highly redshifted objects (item 3 above)
  3. Kinematics and Spectral classification in obscured regions

Because the extinction is in H one-sixth that in V, IR spectroscopy is an invaluable tool to study individual stars in embedded and obscured regions, such as star forming regions and the Galactic center. Although spectral classification in the IR was originally made in the K band, more recent studies have clearly identified in the H band spectral indices (based on Si, Fe, CO atomic and molecular absorption lines) which allow adequate spectral type classification. More recently, from a work carried out at CFHT, a luminosity class indicator has been identified which allows to clearly discriminate giants and supergiants from dwarfs at least for spectral types later than K0.

IR spectroscopy is also the unique tool to perform kinematical studies in heavily obscured regions. When OSIS is equipped with a high resolution grism, it will become possible to measure radial velocities with a precision of ~ 30 km/s.

Therefore, OSIS allows to study e.g. star formation rates and formed mass functions in star forming regions, or to study the relationship of the kinematics with the stellar populations on large samples of stars in the Galactic center. Such studies will give important clues on the origin and evolution of young clusters and of the inner Galaxy.

Until similar instruments are available on 8 m telescopes (VLT, GEMINI) - not before 2001 or 2002 -, OSIS with its multiplex gain of 10 - 15 (current) to ~30 (near future when a larger detector equips OSIS) is a worldwide unique instrument for such studies.
Redshifts of galaxies at z > 0.5

Deep redshift surveys on large fields of faint galaxies provide information on the evolution of clustering with redshift, and on the evolution of the luminosity function and the spectral content of galaxies with look back time. They constrain cosmological scenarios of formation and evolution of structures and, as such, are indispensable to solve puzzling issues on the growth of perturbation, the value of the density parameter Omega, the relation between mass and light, or on the luminosity/density evolution of galaxies, AGNs and quasars, and the period(s) of galaxy formation.

The recent successful CFRS and CNOC surveys at CFHT are unique redshift catalogs of galaxies, and provide statistically significant trends on the luminosity evolution of galaxies and on the evolution of galaxy clustering as well. However, the low- and moderate-redshift filters U, B, and V are redshifted into J, H, K for high-redshift galaxies which need to be studied in the next decade. For these galaxies, a NIR spectrograph is necessary in order to detect emission lines highly redshifted to the infrared.

Above z ~ 0.5, and until Lyalpha shifts in the visible domain at z > 2.5, near infrared spectroscopy is actually the best - if not the unique - way to provide useful spectra of very distant galaxies in this redshift range. In this respect, a multi-object spectrograph in the near infrared is certainly an instrument needed in cosmology for the future redshift surveys. The competition with the VLT makes this instrument competitive on 4 meter telescopes only during the next 4 or 5 years. For that reason a wide field IR camera mounted on OSIS should be available rapidly.

These science cases, together with the uniqueness of OSIS in the international competition, make OSIS a unique facility for the forthcoming years. It is then urgent to provide this instrument with an efficient and dedicated camera. Because the field of view is limited to ~ 4 arcmin, a 1k x 1k camera exploiting the full field will lead to ~ 0.23"/pixel which is adequate to most of the applications. It is unlikely that even with the image stabilisation provided by OSIS the image quality will be significantly better than ~ 0.4 - 0.45", because of the intrinsic prime mirror quality.

Moreover, spectroscopic programs seldom use slit widths narrower than 0.5". A finer sampling with a 2k x 2k camera could even degrade the performances, as more pixels would sample the slits, and hence add their dark current and readout noise contributions. In addition, as 2k x 2k arrays are in a development phase, implementing such a detector on OSIS would likely delay its implementation compared to a 1k x 1k camera. Last, but not least, a 2k x 2k camera could require using a dual optics for use at another focus.

In consequence, ongoing developments on 2k x 2k arrays are to be de-coupled from the OSIS needs. The SAC therefore makes the following recommendation:

Recommendation #6

The SAC reaffirms the high scientific merit of rapidly developing an IR detector for OSIS. A 1k x 1k array will provide adequate image sampling. Consequently, we recommend that a 1k x 1k camera, with ~ 0.2 arcsec per pixel, be provided for use by the community for semester 97II.

4.5 Wide-field IR imaging

The near infrared is a key region for cosmology, since at high redshifts (z>1) most of a galaxy's light is emitted at these wavelengths, and for studying dusty regions of star formation and protoplanetary disks. In addition to the significance of this spectral regime for regions where dust extinction is a factor (e.g., directions toward the galactic center), it is important for objects whose spectral energy distributions peak at these wavelengths (brown dwarfs and low mass stars contributing to the dynamical mass of the halo of our Galaxy, planetary bodies, comets, and asteroids). There are a number of scientific programs which can take advantage of wide-field infrared imaging.

Science Drivers:

  1. Searches for protogalaxies at z>7, and studies of galaxy formation
  2. Evolution of galaxy populations and clusters in the early universe
  3. Studies of star-forming regions and young stellar objects, and their relation to large-scale molecular clouds
  4. The extragalactic distance scale and Hubble constant: infrared Tully-Fisher relation and surface brightness fluctuations methods
  5. Brown dwarf surveys
  6. Galactic structure studies
  7. Dynamical studies of the halo, and relations with gravitational microlensing by low-mass stars
  8. The interstellar medium in starburst and ultraluminous-IR galaxies
  9. Planetary imaging
  10. Exploration of the structure of the distant Solar System and Kuiper belt objects

There are a number of advantages that would allow CFHT to be highly competitive in the area of wide-field IR imaging, even anticipating the advent of 8-10 m class IR-optimized telescopes: (1) the rapid pace of current improvements in IR detector performance (lower dark current, higher QE, improved uniformity, and larger formats) can be quickly exploited with cameras designed for 4-m class telescopes -- there is an experienced instrument-building community among the three agencies which is familiar with the optics design, and a large number of accessible 2-m class telescopes which can be used to test these instruments. By contrast, such instruments for the 8-10 m class telescopes will have to be designed for the new optics, will typically be more expensive and have longer lead times to build, and will have to be tested on these same telescopes where observing time will be at a premium. This means that it may be possible to keep a technological `edge' by the faster implementation of IR arrays with both larger formats and superior performance. (2) The new 8-10 m class IR-optimized telescopes are targeting adaptive-optics optimized performance (smaller angular scale pixels, hence smaller field of view). Many science programs require both depth and field of view, to study structures that are larger than a single IR array can image at one time (e.g., star-forming regions) or because they must study a large sample of objects (e.g., galaxy evolution in the early universe). In this situation for such surveys, even a 1024 x 1024 IR array on CFHT (2.5 arcmin x 2.5 arcmin) can outperform the NIRC IR camera on Keck (38 arcsec x 38 arcsec) and compensate for the difference in collecting area. (3) IR observations must be mosaiced as short exposures, so there is a fixed overhead to read out each exposure and move to the next position in the mosaic. The larger collecting area of 8-10 m class telescopes force short individual exposures, so that the effective overheads of time not spent in actual data collection are much higher.

In addition to issues of competitivity, the superb image quality achieved at CFHT in the optical translates into even better image quality in the IR due to the wavelength dependence of the seeing function -- this is where superlative image quality can be realized, even without adaptive optics. Finally, wide-field IR imaging will allow CFHT users to work in modes which complement the upcoming second-generation Hubble Space Telescope instrument, the NICMOS IR camera.

Options exist for participation in the development of 2048 x 2048 IR arrays, and UH is trying to organize such a consortium.

Considering the importance of these science drivers, as well as the considerations listed above, the SAC makes the following recommendation:

Recommendation #7

Wide-field imaging in the near-IR is scientifically extremely promising. The SAC recommends the rapid development of a 2k x 2k near-IR camera at the f/8 focus, with a field in the range from 7 to 10 arcmin.

5 Long-term future of CFHT

In the months following the November 1995 SAC meeting, various steps were taken towards the establishment of the proposed Working Group charged with considering the long-term future of the CFHT. Principal among these was the definition of firm terms of reference, worked out through discussions among members of the Board, the Director of CFHT, and Harvey Richer, the outgoing SAC Chair. David Hanes, the SAC member charged with chairing the Working Group, had meanwhile approached various astronomers to seek their participation. From the French community, Daniel Rouan and Guy Monnet accepted the invitation; from the Canadian community, Bob McClure declined (citing Gemini responsibilities) and Jean-Rene Roy expressed a desire to consider the evolving terms of reference more carefully, with a decision to come by the start of June. The Hawaiian community has yet to suggest a representative.

Now that the terms of reference have been established, the Working Group (once its membership is finally determined) will begin deliberating, primarily by Email, the relevant issues and lines of investigation, including the broad questions of creating a new Cassegrain focus or of replacing the present 3.6-metre telescope by a large (6-8 m) telescope. Meanwhile, however, the SAC has come to the conclusion that there is no compelling need for the second working group discussed at the November 1995 SAC meeting. This working group was to consider the issue of high-resolution NIR imaging at CFHT under the chairmanship of P.-O. Lagage, but SAC now believes that the issues it would have considered will fall fairly naturally under the consideration of the Working Group to be chaired by D. Hanes.