97th Meeting of the Canada-France-Hawaii Telescope

Scientific Advisory Council

13 - 15 May, 2020 Remotely due to COVID-19

After deliberation during its July 2020 quarterly meeting, the CFHT Board of Directors, in consultation with the Executive Director, endorse the following SAC recommendations:

Recommendation 1 Importance of DRS Updates
Recommendation 2 Internal Risk Assessment and Mitigation Task Force
Recommendation 3 Scheduling of SPIRou Runs
Recommendation 4 Allocation of Additional Nights to VESTIGE
Recommendation 5 Full Allocation of Available Nights to LPs
Recommendation 6 Posting of Validated time for LPs

Report of the 97th meeting of the Canada-France-Hawaii Telescope Scientific Advisory Council,
May, 2020

The 97th Science Advisory Council meeting was held remotely on May 13–15 2020. SAC members Anthony Boccaletti, Sylvie Cabrit, Stéphane Courteau, Emanuele Daddi (chair), David Lafrenière (vice-chair), Eugene Magnier, Nicolas Martin, Roberto Mendez, Laura Parker, and Tracy Webb attended the meeting. The CFHT Executive Director Doug Simons, Director of Engineering Andy Sheinis and Director of Operations Daniel Devost gave presentations and participated in discussions together with several CFHT staff member.

Maunakea Spectroscopic Explorer (MSE)

SAC was presented with an update on the status of MSE. The project now involves 7 participating countries/agencies, along with 2 countries/agencies with an observer status. Participation in numerous international conferences was planned before travel bans related to the COVID-19 shutdown were put in place. Most of these presentations were switched to remote participation, ensuring good global visibility of the project. The Project Office continues to be involved in community surveys (Astro 2020 and Snowmass 2021 in the US; the Canadian and the French community surveys). Finally, SAC was pleased to see a significant amount of effort put into the Program Execution System Architecture that defines the MSE science platform.


The CFHT staff completed several activities to fix some parts or improve performance of SPIRou over the past few months. The motor of the second cold head of the instrument, which failed in December, was replaced with one purchased following the failure of the first cold head last summer. The failure and repair occurred during the hydraulics shutdown and did not directly cause any loss of observing time with SPIRou. Progress was made on a new design for the focus mechanism of the Cassegrain unit which, once implemented, should lower sensitivity of the fiber-to-field-mirror alignment to temperature changes and impacts. A sensing system for the fiber agitators was installed and provided encouraging results, but some problems remain; a plan to fix them was developed and is moving forward. Rhomboid prisms newly refurbished by the company Winlight were received and installed in SPIRou. These rhombs have a better alignment between their entrance and exit faces and should reduce spurious RV motions observed with the previous rhombs when moving them. The neutral density wheel of the calibration unit went out of alignment for an unknown reason. The wheel was realigned and a system was put in place to identify this problem more quickly should it reoccur. Operation of the Laser Frequency Comb (LFC) has improved but remains unreliable and the system displays a drift for several hours after being turned on. At this point, diagnosis and maintenance of the LFC still requires non-negligible attention from the CFHT staff, who work in cooperation with Menlo Systems.

The SPIRou team appears to be determined to execute a major upgrade of the spectrograph that will require opening the cryogenic vessel and incur a shutdown of a minimum of 2 months. The main goal of this upgrade is to replace the octagonal fiber-slicer combination, which should improve fiber mode scrambling by a factor of 10 and consequently improve RV stability. Another goal is to find the source of RV jumps occasionally observed, that are thought to come from a mechanically loose part in the spectrograph. If a solid case for this intervention is made and approved, this upgrade would not happen before 2021A. It would be preferable for CFHT to perform this upgrade during A-semesters when LP pressure for MegaCam is stronger.

A major upgrade of the SPIRou DRS, now called APERO, was completed and provided to CFHT by the instrument team. The instrument and CFHT teams have been working closely together to test this new version and improve it further. The new DRS version implemented several important changes to improve the quality of the calibrated spectra and the precision of the RV measurements, including a self-consistent determination of the spectral trace shapes across all orders, removal of contamination from the reference channel in the science channel and removal of detector persistence effects, as well as many other smaller changes. This upgrade represents a big step in approaching the RV precision goal of 1 m/s and SAC congratulates the teams on their good progress. However, the new DRS has not yet been implemented by CFHT and the wider CFHT community has not yet benefited from it and its associated improved data products. There are several reasons for this but an important one is that the new DRS introduced many incompatibilities between its data products and the software used at CFHT to operate it and prepare the data for archiving and distribution at CADC. As a result, no PI data has been processed with the new version at CFHT, and no new PI data has been processed at all since last December as processing with the previous version is now deemed obsolete. The current situation is problematic for PI's, as they can only access either sub-optimal data products or no data products at all. In a field where competition is high, where the RV precision depends critically on the quality of calibrated spectra, and where the ability to quickly analyze data is fundamental, the current situation is not acceptable.

RECOMMENDATION #1: CFHT and the DRS team should work even more closely together on DRS updates, with a first priority that the current and subsequent versions of the DRS can be readily used by CFHT to process new and reprocess old data, so that CFHT can always provide its users with optimal data products. Looking ahead, the instrument team and CFHT should consider adopting a shared development version of the DRS which can be operated both at CFHT and by the DRS instrument team and wherein both teams can contribute directly. Both teams should also consider updating the DRS in such a way that it would directly adhere to the data product specifications appropriate for packaging and distribution of data to CADC.

Back to top


SAC notes three areas of investigation for SITELLE. First, the variation of modulation efficiency noted in previous reports is still ongoing. However, the impact on science is negligible and this investigation is deemed a low priority at the moment. Second, a significant source of background hits on the detector has been discovered, most likely due to radioactive (alpha) impurities from the two new corrector lenses. This situation is being addressed with the purchase and implementation of two new field corrector lenses to be surfaced with a low-emissivity coating. These new lenses are expected to be installed in late 2020. Finally, connectors for cryo-cooler lines need to be replaced. This is viewed as routine maintenance.

SITELLE will not be mounted on the telescope during the COVID-19 shutdown. Operations could however resume even if the current lenses are still in place. Science results will be moderately affected given the added noise from the radioactive hits, let alone the normal occurrence of cosmic rays, and caution must be taken by the remote observers and resident astronomers to limit the impact of those “hits” on the main target spectra.

ESPaDOnS/SPIRou co-mount

CFHT staff have continued to explore the design options and trade-offs of a possible ESPaDOnS/SPIRou co-mount implementation, including sample coating studies from a possible vendor, Iridian. SAC commends the CFHT staff on the progress of the exploratory studies to quantify the impact on both instruments from a co-mount design using a dichroic beam-splitter. The throughput reported by Iridian for the sample coatings is encouragingly high for both the ESPaDOnS and SPIRou sides. SAC agrees with the plan to complete studies of the impact on the polarization precision and overall efficiency to have a clear understanding of the scientific impact of the co-mount implementation before approaching the community for scientific usage cases.


The banked CFHT allocation for GRACES continues to grow but is still below the threshold of 14 nights overall deemed necessary to make a useful call for proposals for France. SAC agrees with the decision to wait until enough time is available to make a call for proposals in consecutive semesters.


SAC was updated about the ongoing problem with the SLINK boards and the effort to replace them with a GigE interface. Much of this work has focused on the software development and testing and the team is now ready for the switch, which could be undertaken at any time. SAC also heard about the fabrication and testing of the GigE interface boards. This is progressing well with the majority of the boards passing the tests. Efforts are on-going to characterize and understand the readout noise in the new system, which is slightly higher than for the SLINK interface.

SAC appreciates the ongoing efforts to mitigate this noise source and encourage the Observatory to provide a characterization of the impact of this extra source of noise on low-background programs.

Dome Vents

SAC was pleased to hear that 10 out of the 12 dome vents are now released after the installation of a system preventing water freezing inside the vents and blocking them. The vents are currently only raised to two-thirds level to avoid any danger of them getting stuck in an inaccessible position while the man-lift is not available. The installation of a heating system is planned to further assist with melting ice after observing, before door closing. SAC was presented with encouraging first results of a Machine Learning study performed by CFHT to optimize the use of dome vents on the fly and improve image quality as a function of on-going conditions of observing throughout the night. The current modeling exercise enables the inference of the image quality at the 0.1" rms level, better than the current measurement with the Mauna Kea Atmospheric Monitor. The Machine Learning exercise for dome vents optimization is interesting, and worth pursuing if undergraduate labor is available. However, SAC discourages CFHT staff from spending too much further attention on this modeling exercise, given all other priorities.

Operation Status

Upgrades on the bridge crane following last year’s failures are continuing. The access platform has been designed, manufactured in-house by CFHT, coated in the Fall of 2019 and finally authorized for installation. Because of other priorities it will be installed in the Fall of 2020. Several other technical issues were addressed on the crane this year and a number of upgrades that remain to be done are identified. However, because of budgetary concerns, CFHT is considering postponing major upgrades, including the acquisition of the Variable Frequency Drive costing $207k. A risk assessment for this deferral has been triggered but is delayed by the COVID-19 shutdown.

CFHT experienced a major failure of the Telescope Hydraulic System at the end of 2019, which resulted in an extensive 45-night shutdown. It is temporarily possible to bypass the problem owing to the persistence of the summit crew and external expertise (Atlantic Hydraulic System). A definitive repair requires designing a new Hydraulic Power Unit, the cost of which is $145k. Integrations should start this summer and run until the Fall of 2020.

SAC is pleased to hear that the implementation of the astrometric camera on the southeast truss of the telescope was successful, allowing sub-arcsecond astrometric precision. However, the location of the 8-inch camera is impacted by occultation from the dome and involves a 1 arc minute deflection between the camera and the telescope due to flexures that varies with time. Modeling of the camera position indicates that the North side of the telescope is a better location, avoiding occultation and reducing the deflection to <16’’. This however requires redesigning the camera mount. Manufacturing will occur in the next few months. The field mapping software was upgraded to include the 2MASS and Gaia DR2 catalogs, which are useful for narrow field instruments, namely SPIRou and ESPaDOnS. This camera is also intended to provide a reference to compute the mean sky extinction in multiple bands. Another application of the astrometric camera is to provide SPIRou users with a photometric monitoring of their targets to assess variability in order to detrend radial velocity or transit measurements.

The atmospheric monitor (MKAM) will have its 10-year-old shroud replaced during the next scheduled maintenance. The all-sky monitoring camera (ASIVA) failed a year ago. Both the visible and infrared cameras should be replaced, the latter being more challenging. The pixel size of current detectors does not match the ASIVA one. The weather hatch is to be replaced too.

The replacement of the electrode tips in the aluminizing chamber is being completed but the final test-fire was not yet possible due to the COVID-19 shutdown. As a regular problem, one of the gauge controllers failed, and has been replaced. With the goal to produce more consistent coatings, the vacuum controller will be upgraded with a Raspberry Pi to monitor thickness and rate of the aluminum deposition on the mirror.

Cleaning of dust on the primary mirror resulted in zero-point improvements as measured with MegaCam and WIRCam. The precise improvement and appropriate frequency of cleaning are still to be quantified. However, this procedure does not clean water spots left by condensation. Another procedure is being tested to limit condensation. To that end, dedicated sensors are being developed to more accurately and quickly react to the presence of condensation.

In order to access the telescope structure and service the dome vents, a new man-lift is to be purchased, the cost of which is $60k but leasing options are being considered.

The software activities focused on the development of a new QSO platform already in place for SPIRou, and possibly re-used in the future by MSE. Collaboration with CADC is effective to improve several aspects like user interfaces, data quality assessment, availability of metadata, etc.

SAC is concerned that the aging CFHT facility is at increasingly greater risk of severe malfunction. As a stop-gap measure and an alternative to hiring an expensive external consulting firm, SAC recommends that an internal committee be struck to revisit and update the latest risk assessment report. Experience from colleagues at other “aging” observatories (e.g. Mt Wilson, Palomar, KPNO, Lick, ESO, AAO) could be most beneficial along those lines. Should local labor be lacking for this, former CFHT employees could be consulted in order to expedite a detailed comprehensive risk assessment in the next few months.

RECOMMENDATION #2: Implement an internal risk assessment task force and proactively work on risk mitigation in the next few months.

Back to top

Large Programs


CFIS is actively pursuing exploitation of the data collected so far, with a number of publications coming along on different science topics, and internal data release to the large UNIONS collaboration. Unfortunately, the data collection rate continues to be slow, increasing the risk that the overall science goals might not be reached despite the extra allocations as a result of the LP completion policy implementation in 2019. SAC is actively monitoring the situation. CFIS is still in need of some fraction of B-semester allocations, and plans to concentrate observations around the North Ecliptic pole to optimize contribution to the Euclid survey needs.


This project was finished in 2019B. The completion level was 82%. The scientific goal was measurement of trigonometric parallaxes of low-mass stars and brown dwarfs using WIRCam. As good astrometry requires good time sampling and extended baseline, the UH TAC has continued to support this program through regular PI proposals in 2020A and 2020B, allocating about 18 hours per semester.


Last semester, the Large Program completion review sub-committee recommended the allocation of additional nights for VESTIGE in 2020A and 2021A to help the LP reach 80% completion. Unfortunately, this extra allocation for 2020A could not be used because the COVID-19 shutdown aligned with when Virgo is optimally visible. VESTIGE therefore continues to fall far short of the number of allocated nights and is at risk of not being able to complete its scientific goals. The VESTIGE team has been publishing regularly with the data collected thus far, but the full potential of the survey requires coverage of Virgo to the virial radius.


The SIGNALS project enjoyed very successful completion rates in 18B and 19A but these plummeted below 50% in 19B and 20A largely because of the crane malfunction, bad weather, and the COVID-19 shutdown. With the COVID-19 reduced mode of operation, SITELLE will likely not gather any science data until September 2020. The SIGNALS team is implementing a Machine Learning spectral-line fitting procedure to HII regions that will greatly enhance their efficiency in their analysis of SITELLE data. The goal is to make the fully-calibrated spectral-line fitting software available to the entire SITELLE community. The SITELLE consortium has published one refereed paper so far. The current data should allow for the publication of at least one additional article in 2020 and a few more in early 2021. SAC is looking forward to a more extensive SIGNALS LP report in coming semesters.


The SLS LP was allocated 300 nights spread over 7 semesters. However, over the first 2 semesters (2019A and 2019B), progression has been significantly slower than expected with only 50% of validated nights out of the 88 allocated in 2019A and 2019B. This low rate reflects the combined impact of weather, technical issues and TMT protests. Even assuming an unlikely 100% validation by 2020A (39 nights), a high 90% validation rate over the following four semesters would be required to reach the 80% overall completion rate for this LP. Therefore, it is quite likely that an extension of the SLS beyond 2022A will be necessary. This will be discussed at and after the mid-term review.

In terms of science, promising results have been obtained on a few targets with already known planets, with several publications on SLS data submitted to refereed journals. However, much more data, typically 50-60 visits per star, are required to enable a significant detection of planet candidates. On younger T Tauri stars, series of at least 40 spectra within one season are necessary to investigate the potential presence of young close-in giant planets. This has not yet been possible so far.

This LP faces an additional specific scheduling problem in that its science goals — detecting planetary systems by RV studies, studying magnetic activity of low-mass stars, and removing the associated RV jitter evolving on timescales of only a few weeks, — require a dense time sampling over long periods of time with as small gaps as possible: ideally each target would be followed once per night for at least 10 consecutive nights each month.

RECOMMENDATION #3: CFHT should endeavor to schedule SPIRou runs of at least ten nights whenever possible. Transit observations carried out over consecutive nights in a similar RA range should be avoided.

Back to top

LP time loss impact and mitigation strategies

SAC is aware that a seriously problematic situation is building up for the LPs, as a result of exceptionally large night losses accumulating over several years: extended periods of weather losses in 2018, crane failure and summit protest in 2019, hydraulic system failures and COVID-19 shutdown in 2020. SAC is determined to react decisively in order to preserve the scientific output of ongoing Large Programs as much as possible and to minimize adverse consequences.

Given that a large fraction of the new nights recently allocated to VESTIGE as part of the LP completion policy were almost immediately lost in 2020, SAC decided to activate again an LP completion review for VESTIGE during this meeting, with the LP completion review sub-committee consisting of the whole SAC (including members of the International TAC). In fact, the remaining nights for VESTIGE are not sufficient to allow reaching a minimal 80% completeness for this project.

RECOMMENDATION #4: The sub-committee recommends the allocation of 10 additional nights to VESTIGE: 1.5 nights during 2020B and 8.5 nights during 2021A.

Back to top

Within the time frame of the current LP allocations, a buffer of roughly 35 nights remains available and could be allocated to LPs in the coming 4 semesters (2022A included). SAC expects to allocate this time to LPs with lagging completion rates over the coming semesters, following the guidelines of the LP completion policy.

RECOMMENDATION #5: CFHT should ensure that the entire pool of nights available to Large Programs are fully allocated in coming semesters, starting already with 2020B.

Back to top

Time Allocations

SAC commends CFHT for its handling of the many challenging changes to the schedule in 2019B and 2020A due to numerous unexpected interruptions (TMT protests, equipment failures, and global COVID-19 pandemic). SAC is also pleased to see the effective use of Director's Discretionary Time to support outreach programs with high school student projects. These projects yield large benefits in the community for a modest investment of a fraction of a night.

SAC appreciated the breakdown of PI time requests into scientific areas.


The QSO system continues to perform well, in spite of the challenges due to the unexpected schedule changes.

The intended QSO schedule had enough nights for each instrument to accommodate the accepted A&B programs, including LPs. However, 51 nights of 2019B were lost due to TMT protests, bridge crane repair work, and the hydraulics failure. Ultimately, MegaCam lost 13 nights of dark time (33/49) but received 16 nights of bright time (with some useful data), SPIRou lost 20 nights (54/74), WIRCam lost 11 nights, ESPaDOnS lost 3 nights, and SITELLE lost 5 nights.

Overall completion rate (validated / allocated) in 2019B for A programs was between 31% (SITELLE) and 56% (MegaCam and SPIRou); and even lower for A+B: 18–19% for SPIRou and SITELLE, 26% for ESPaDOnS, and 36–40% for WIRCam and MegaCam.

The overall fraction of observed time that is validated remains at healthy high levels, generally higher than 90%.

SAC noted differences in the assessment of validated time between the LP teams and CFHT. These differences remain small but can amount to multiple nights over the length of an LP.


RECOMMENDATION #6: SAC recommends CFHT provide a definitive accounting, per semester, of the validated time for the LPs on their website that will serve as the reference. The accounting should be on a publicly accessible page, and updated in real time or as frequently as possible. CFHT should also clarify which criteria were used to define validated time.

Back to top

Operational and Development Priorities:

  1. Normal operations
  2. SPIRou and DRS
  3. Observatory risk assessment and mitigation
  4. MSE
  6. ESPaDOnS-SPIRou co-mount

CFHT agrees with these priorities and anticipates being able to make progress in all of these areas over the course of 2020.

Next SAC Meeting

The next SAC meeting will take place at the CFHT Headquarters in Waimea (HI) on November 5 & 6, with a tour of the summit organized on November 4, if international conditions permit safe traveling. A remote meeting will be held on the same dates, otherwise.