Report of 67th meeting of the CFHT SAC

Paris, 30-31 May 2005


Recommendation #1 Image Quality
Recommendation #2 Length of Megacam Runs
Recommendation #3 Megacam Queue observations
Recommendation #4 Commissioning of WIRCam
Recommendation #5 Libre-Esprit
Recommendation #6 MOS & GECKO
Recommendation #7 CADC
Recommendation #8 CFHT Future
  CFHTLS Mid-Term Review Report
  Supernova Survey
  Deep Survey
  Wide Survey
  Very Wide Survey
  Observing-time requirements
  Summary of principal actions taken by the CFHT Board of Directors in response to the May 2005 SAC and CFHTLS Mid-term Review reports
  R3 Length of MegaCam runs
  R6 Decommissioning of MOS and Gecko
  R9 Terminating the Very Wide component from the CFHTLS
  R10 Allocation of 4+4 nights per semester
  R11 Extension of survey end date by one semester


The meeting was attended by SAC members D. Bohlender, J. L. Beuzit, D. Elbaz, C. Ftaclas, J. F. Gonzalez, H. Hoekstra, G. Mitchell, P. Petitjean (chair), N. Saint-Louis (vice-chair) and B. Tully. Jim Hesser for the Board and Korean representative Dr. Ho-Il Kim were present as observers. C. Veillet, J.-C. Cuillandre, P. Martin, P. Puget and D. Salmon, participated in presentations and discussions. L Simard from CADC was invited.

1. Technical Activities Report

Derrick Salmon reported on the technical activities of this semester, including the commissioning of Espadons and the integration of WIRCam. Recent developments of the Wircam detectors and their controllers were also presented.

Different maintenance and development activities of the building and telescope were described: building air handling progressing slowly, soil erosion issues, the telescope runaways occurrences, the “drop into manual” problem, time jumps in the TCS master clock.

In anticipation of WIRCam data, the online storage capacity has been increased to about 32 Terabytes.

Problems with the telescope associated with the DIMM have interrupted the project. It should be pursued on a manpower-availability basis.

2. Instrument and Project Reports


The number of technical failures of the camera has decreased substantially after an engineer from CEA came to CFHT and provided some sense of where CFHT needs to pay more attention. Some problems with the CCD controller electronics, the shutter, the guide probe and the fiber connectors have been investigated.

A teleconference between the CEA hierarchy and CFHT resulted in the clarification of some issues of concern. This has helped to solve the cooling problem and CEA will be providing electronic spares. Two people, including the former project engineer will spend one week at CFHT in July to help find a solution to the problem with the flanges. It should be understood that the camera is a complex instrument and is fragile.

The Image Quality in the central half of the field is much better now that L3 is flipped. Given the mean seeing, 85% of the surface is within specification and the remaining problems are limited to two corners of the image only. SAC believes it is worth trying to understand the origin of the image quality problems. Models have been used to investigate possible answers to this question. The wavefront of the primary mirror does not seem to be involved. However, something is affecting the power of lens L1.

A model of the focus is in place and substantially decreases the amount of overheads related to focussing the camera. The residuals from the model are approximately 73 microns while in the best seeing the accuracy of the focus position is about 50 micron.

Based on this report SAC makes the following recommendations:

SAC Recommendation #1:

Although the improvement of the IQ has been clearly demonstrated, it is very worrying that the optical performance of the system was improved by accidentally installing a component upside-down. Therefore SAC recommends that efforts to understand the origin of the degraded IQ of Megacam should be continued but with a lower priority than WIRCam.

Back to top

SAC Recommendation #2: Length of Megacam runs

(Recommendation has been removed - deferred to recommendation #3 contained in the CFHTLS Mid-term Review report, attached to this report.)

Back to top

SAC Recommendation #3:

SAC is worried about the low efficiency of the Megacam Queue observations and the remaining technical problems that may impact further this efficiency. A visit by CEA engineers is scheduled in July 2005 at CFHT. SAC requests a written report on the results of this visit based on input by both parties. This should be discussed at the next SAC meeting.

Depending on the outcome of this visit and the report, SAC recommends that the Board be prepared to appoint an independent technical committee to review difficulties with the instrument and to help CFHT to implement the appropriate actions.

Back to top


All science detectors have been delivered by Rockwell. Their efficiency is better than specifications (70-80%) but the readout noise is slightly higher than specifications; overall they are therefore considered better than expected. ASICs are still not available but controllers are being developed by CFHT. The integration of the optics in the cryovessel was done in March 2005 and engineering runs take place every month since March. The image quality is excellent (~0.55² over the whole field) and no major problem has been found so far. The camera should be ready to be offered to the community in 2005B as expected. 15 nights of PI programs plus 8 nights of engineering time as well as 2 nights of DDT are foreseen.

All observations should be performed in Queue mode. CFHT has devoted an important effort to develop a pipeline for the pre-reduction of the data.

SAC Recommendation #4

SAC is very pleased with the progress of the commissioning of WIRCam and congratulates the teams involved. However, SAC is still concerned about resources for data reduction.

CFHT should finish the development of the Elixir-IR pipeline. Terapix is encouraged to provide advanced WIRCam data products (stacked images and catalogs) but this work should have a lower priority than the CFHTLS releases. In any case proposals for Large Programs should provide a clear plan for data reduction.

Back to top


The introduction of the instrument into CFHT observing has been proceeding very smoothly. The Cross-talk between the linear and circular modes that has been discovered will be investigated next semester. The observing runs on the sky were very successful scientifically. Observers were pleased by how easy it is to use the instrument and the pre-reduction package. Queue observing is foreseen but not before two years from now.

SAC Recommendation #5

SAC is pleased with the instrument performance but is concerned by the proprietary status of the Libre-Esprit package. CFHT should offer remote data reduction capabilities if requested by users.

Back to top

Fly-eyes, Pueo-Hou and `Ohana

No work has been done on Fly-eyes this semester.

In order to stimulate instrumental developments at CFHT, a concept for an adaptive optics system with a laser guide star that would provide diffraction limited imaging in the optical is being investigated. The so-called Pueo-Hou instrument could possibly provide CFHT with a unique capability on the 2010 horizon.

The Phase II of `Ohana is under way. The goal is to demonstrate the potential of an 800-meter fibered interferometer with the 3 10-meter apertures existing on Mauna Kea. The first baseline to be used is the CFHT-Gemini baseline planned to be combined by the end of 2005. The corresponding delay lines will be shipped during the summer and installed in the first floor Coudé room. Discussions to install a fiber between CFHT and Gemini have begun. One night was spent with the Keck interferometer to validate the fibered coherent transport using single mode fibers. It is suggested that an engineer from the instrumentation group be assigned to the project.

Other instruments

SAC Recommendation #6

The pressure on MOS is steadily decreasing. In addition, MOS is no longer a unique instrument and similar observing capabilities are accessible to all communities. SAC therefore recommends MOS be decommissioned at the end of semester 2005B. Because of the availability of Espadons, SAC recommends the decommissioning of Gecko at the same time.

Board Action 6

Back to top


SAC Recommendation #7

SAC is impressed by the amount of data already retrieved by users from CADC and encourage CADC to pursue the development of tools to facilitate access to the data. SAC would like to monitor the impact of the CFHTLS in the community and therefore wishes to receive detailed statistics on the retrieval of CFHTLS data at the CADC every semester.

Back to top


The mid-term review of CFHTLS took place just before the SAC meeting. Specific comments and recommendations have been sent to Board as a separate document.

SAC had a more general discussion about how to improve the management of possible future surveys.

Given that the LS utilizes about $1M/yr in telescope time and many times this amount in data reduction, archiving and analysis resources it is believed that programs of this magnitude require a project coordinator who is not tied to any particular component of the science and can advocate to the SAC/Board, telescope and data components to see the project is completed with the best outcome for the corporation.

For any future survey or large program work, image quality and observing efficiency goals should be set as part of the survey definition and that survey work should not begin until these goals have been demonstrated. Survey commencement should require approval of the board based on recommendations from the SAC and telescope director.

Given that the telescope currency used in execution of the LS is telescope hours and that observations are made in queue-mode the SAC believes it would be more realistic that all obligations, mandates and allocations associated with the LS and all future surveys be cast in terms of telescope hours rather than nights.

5. Telescope Use and Scheduling and DDT

Queue observations are working very well and the validation rate is now above 85%. Back-up programmes to fill the queue in case of bad weather conditions (seeing above 1 arcsec) are needed.

Ten nights per semester are put aside for DDT. Note that at the moment, 3 nights are allocated to the CFHTLS. The use of DDT should be reviewed by SAC. This has not been done seriously so far. SAC has asked to the CFHT director to provide details on how DDT is used at the end of each semester starting next semester. About 15 nights of engineering time will also be spent next semester.

6. CFHT Future

Christian Veillet presented a plan for the next 5 years that has been discussed in house with the aim of maintaining the high level of motivation required for a world-class science facility in the 2005-2010 timeframe. R&D developments could be centered on the refurbishment of the AO system, a possible laser guide star system and the Ohana project.

Christian Veillet indicated that the long-term future of CFHT is partly linked to the possibility of other partners joining the corporation. The site itself could be of interest for specific projects.

SAC Recommendation #8

SAC is pleased of the efforts made by the CFHT director to prepare for the future. The substantial investments made recently to install unique new instrumentation (Megacam, Wircam, Espadons) will produce world class science at CFHT. There is a continuing need to support efficient operation and to bring these instruments to full functionality. It is also important that CFHT maintain a healthy level of technological development to attract and retain high-quality staff.

The Ohana project is definitely to be pursued and the Pueo-Hou concept seems an interesting possibility to be investigated in more detail. SAC believes that any further plans should be based on a review of resources and how they could be adapted to the future.

Back to top

7. Other items

Korean cooperation

Korea will end its collaboration with CFHT soon. Korean and CFHT signed an MOU of cooperation in July 1, 2000. The first Korean paper based on CFHT data appeared in 2003. KASI was very happy of the experience gained in the operation of a world class telescope and in the management of a multi-nation facility. They were also pleased to have participated in the development of state-of-the-art instrumentation and to have given the opportunity to their young astronomers to gain useful experience. Now Korea is planning to build their own large telescope by participating in Mexico’s plan for 2x6.5m telescopes on San Pedro Martir. They expect First Light by 2011. The funding provided by the Koreans has been used for the WIRCam project.

SAC and TAC memberships

Jean-François Gonzales has been appointed as a French SAC member. The next chair of the French TAC will be P.A. Duc.

Next meeting

The next SAC meeting will be held in Waimea on November 10-11, 2005.

CFHTLS Mid-Term Review Report

Mid-term review process

* SAC asked the Steering Group (SG) to prepare an overview of the Legacy Survey. Each component was asked to submit a separate report

* Each component of the survey and the conduct of the survey as a whole were reviewed by two external expert referees. 9 out of 10 referees responded and provided very detailed reports.

* Members of the SAC were assigned to prepare summaries of the material for each component of the survey. These summaries were made available to the SG.

* A workshop was held on 26-27 May 2005 at IAP with approximately 100 registered participants. Status reports and scientific results were presented. The workshop also featured several active discussion periods.

* On the morning of the 28 May, SAC met with the SG and CFHT staff to review the major issues raised by the referees and during the workshop.

* In the afternoon, SAC went into executive session to review all relevant information and formulate recommendations.


SAC is generally impressed by the organization of the project as a whole and the effort and dedication of the different teams and CFHT staff to the overall goal. The SG report is well documented and comprehensive.

During the workshop, discussions were very informative and clarified a number of technical and scientific issues.

The results achieved by now are already impressive and SAC believes it would not be necessary to cut back the time needed to achieve the final goals if all aspects of the survey were proceeding according to the original plan. However, the low efficiency of the observations, partly due to weather conditions, technical problems and scheduling difficulties are presently jeopardizing the expected breakthrough results of all survey components.

In view of the continuing high pressure by PI proposals and the anticipated pressure from WIRCam large programs, SAC judges that it is unrealistic to recommend a major increase in the amount of time currently being allocated to the survey each semester beyond the 4+4 additional nights already allocated. Given the highly competitive international environment, there is a need to guarantee the rapid completion of key components of the survey in order to optimise the return of the observing-time investment and to minimize the risk of instrument failure.

In the view of the SAC, the recommendations below represent the most reasonable way to reconfigure the survey such that two of the original three primary science goals can be successfully completed in a timely fashion, maximizing the scientific return to the participating agencies and providing the greatest legacy value to the entire astronomical community.

SAC Recommendations

Supernova Survey

The SNLS is presently in an advantageous position to have a major and long-lasting impact on the study of the nature of dark energy. This survey has been very well thought out. It has clear scientific goals that are of the highest importance. The goal of 700 supernovae discoveries by 2008, observed in 4 colours, if achieved, will make this survey the most powerful supernova survey to date and every effort should be made to guarantee success.


R1    SAC considers that the supernovae survey should be regarded as the first priority of the CFHTLS. In view of its importance, it should be given every chance to be completed as fast as possible and therefore should have the highest priority during queue observing. The SAC therefore recommends that the SNLS be ranked above Canadian and French PI programs at the same right ascension.

Back to top

R2    The two data reduction pipelines are working very well and are a key element in the process, as they provide photometry that is checked independently. Therefore they should be maintained.

Back to top

R3    MegaCam queue-observing runs should in no case be shorter than 14 consecutive nights in order to enable adequate sampling of the light-curves.
Board Action 3

Board Action 3

Back to top

R4   Although scientifically interesting, the proposed additional z-band observations cannot be considered as a first priority in view of the limited time available to carry out the survey and should not be added to the SNLS observations.

Back to top

R5   SAC supports the plan of the SG to re-introduce the two extra g’ epochs to better sample the light curves.

Back to top

Deep Survey

The deep survey builds on the SNLS data with the addition of deep u-band imaging of the four SNLS fields. The referees note that the scientific case is not as compelling as for the SNLS and Wide Survey. However SAC recognizes the importance of the u-band images for the legacy status of the SNLS imaging data and for all the science relying on multi-wavelength information. Indeed, the fields have been observed by XMM and Galex amongst other facilities and a large community is waiting for these data. Therefore, although not a short-term priority, the data must be obtained. SAC encourages the community to organize itself better around the science to be done with the deep fields.

R6    Deep u-band observations of the D1, D3 and D4 SNLS fields should be completed but at a lower priority than the SNLS observations and the gr/2i Wide observations. u-band observations of D2 should be dropped to avoid duplication with the COSMOS survey.

Back to top

Wide Survey

The multi-colour Wide survey is a very important part of the CFHLS with a wealth of science goals from cosmic shear to Galactic structure and clusters of galaxies. However, the main driver and justification of a high priority is the study of the cosmic shear. The particular strength of this survey compared with its competition is the large area coverage, which simultaneously gives information on structure up to 5 degrees scales and cosmic variance. SAC notes that the Wide Survey group clearly demonstrated that with the flip of L3, the image quality provided by MegaPrime no longer poses a risk to this component of the survey.

R7    Among the different parts of the Wide survey, the cosmic shear has the most compelling scientific case. It should be given every possibility to achieve its goal as fast as possible. It should be considered as the second highest priority of the overall survey and should be given highest priority with respect to the other scientific goals of the wide. Its particular strength is the 170 square degree field which should not be reduced.

Back to top

R8   g, r/2 and i imaging should be done first and finished before the end of 2007B. The u/2 and z/2 band observations are not considered as urgent, should be assigned equal secondary priority, and could be completed over a more extended time period. Given the reduced sensitivity of z’, SAC recommends obtaining z’ data with the proposed integration times but the SG should demonstrate the need for the second set of u/2

Back to top

Very Wide Survey

The primary driver of the CFHTLS Very Wide component is Kuiper Belt research. This component was originally approved as a survey of about 1300 square degrees of the sky within 2 degrees of the ecliptic plane, but avoiding the galactic plane. As summarized in the CFHTLS VW mid-term review document, the goals are to explore the stable zones of the solar system from 20 to 150 AU, measure the size distribution of the Kuiper Belt down to 100 km, provide a large sample of Kuiper Belt objects for modelling the underlying orbital distribution of the belt and enabling physical studies, and search for new orbital classes as a probe of planetary formation.

As of 2004 December only 250 square degrees have had Discovery/Nailing/Checkup (DNC) observations (800 degrees were expected at this point in time) and about 120 objects have been discovered. This already constitutes the largest sample of such objects available. However, at the current rate of observation, only half of the original proposed area would be completed during the 5-year CFHTLS time-frame. Because of this, the Very Wide survey is under the greatest threat of not meeting its goals. To overcome this shortfall, the SG suggested a significant acceleration of the discovery portion of the survey at the expense of VW observations in multiple filters. This strategy severely compromises the legacy value of the Very Wide component.

It is difficult to argue for a significant acceleration of the discovery phase since, as noted by a referee, there is no clear justification for how many objects are needed and what the sample will quantitatively constrain. This prevents an accurate estimate of the observing time really needed for a significant advance in this field.

Delaying the Very Wide survey beyond 2008 is not a viable option because of the competition with Pan-Starrs.

The other main scientific drivers of the VW are the search for brown dwarfs, the search for high redshift quasars and the detection of GRBs. The latter cannot be considered as a priority as only about one GRB per year (number uncertain by a factor of 3 to 5) is expected to be detected by the survey. The two other projects are mainly based on the i-band imaging which is not a priority for the Kuiper-belt survey. Well argued i-band observations could be considered after 2008B in case the survey is extended.

Back to top

R9   Given the higher priorities put on cosmic shear and SN, difficulties with scheduling the time-constrained observations, and the lower than expected efficiency of the survey, SAC recommends that the DNC portion of the VW survey be terminated at the end of 2005A. The determination of the orbital parameters of the objects already detected should be completed by executing the required recovery and 3-year observations using the current multi-filter observing strategy.

Board Action 9

Back to top 

Observing-time requirements

The amount of time needed to achieve the reconfigured survey described above is 250 h of validated observations per semester till the end of 2008B (based on numbers provided by the SG). With a mean expected efficiency of 4.5 h per night this corresponds to ~57 nights per semester. SAC wishes to ensure the success of the reconfigured survey and does not consider it prudent to rely on the possibility of a significantly better QSO efficiency.

R10   In order to obtain the required 250h, SAC strongly recommends that the participating agencies continue to allocate the 4+4 nights per semester. This additional allocation will be subject to review by SAC every year during the fall meetings. These reviews will be based on the SG progress reports and CFHT QSO statistics.

Board Action 10

Back to top

R11   The survey end date should be moved from 2008A to 2008B. The need for a further extension of the survey after this date should be discussed by SAC two years from now.

Board Action 11

Back to top


R12   SAC is pleased with the quality of the work done at the Terapix center. However, the present delays in the release of advanced Terapix data products, especially stacked images, to CADC are not acceptable. The detrimental impact of these delays was noted by many participants in the workshop. SAC urgently recommends that delays in data releases be greatly reduced. Intermediate data products including stacked images and catalogues must be made available to the community as pre-releases at most six months after the data have been received by Terapix. The content and data quality of these pre-releases should be defined by the SG together with Terapix.

Back to top

R13   Terapix should cooperate with the SG to prepare a phased plan for the forthcoming releases of data.

Back to top


Summary of principal actions taken by the Board of Directors in response to the May 2005 SAC and CFHTLS Mid-term Review reports

12 July 2005

In response to recommendations resulting from the May 2005 SAC and CFHTLS Mid-term Review meetings, the Board has taken the actions listed below. These actions were taken after intensive discussions were conducted for several weeks on these particular items. The Board wishes to convey to both, the CFHTLS Steering Group and the Scientific Advisory Council, its appreciation for the impressive amount of work and careful examination reflected in the resulting recommendations.

1. SAC REPORT - Recommendation

Recommendation No. 6 – Decommissioning of MOS and Gecko

Deferred further discussion of the decommissioning until the December 2005 meeting of the Board, and asked the Executive Director to prepare and issue an announcement to the community advising that these two instruments are being considered to be decommissioned within the next 12 months.

Back to top

2. CFHTLS MID-TERM REVIEW - Recommendations

Recommendation No. 3 – Length of MegaCam runs

Agreed not to establish a rigid policy for the detailed number of consecutive nights during each MegaCam run at this time. The Time Allocation Committee should consider this recommendation as additional guidance in the establishment of each semester’s allocation.

Back to top

Recommendation No. 9 – Terminating the Very Wide component from the CFHTLS

Agreed to terminate the DNC portion of the Very Wide component at the end of semester 2005A.

Back to top

Recommendation No. 10 – Allocation of 4+4 nights per semester

Agreed to continue to view the addition by the participating agencies of these nights as extra, temporary allocations to the LS to be done on a semester basis and not as a permanent addition to the time allocation to the Survey. The decision by the participating Member Agencies on each semester beyond 2005B should be made early enough so that the TAC has advance notice of the allocations.

Back to top

Recommendation No. 11 – Extension of survey end date by one semester

Deferred discussion to extend the end-date of the survey observations to the Board’s December 2005 meeting.

Back to top