Chandra User's Committee Meeting -- October 19,20 2005

Attendees -- Julia Lee, Smita Mathur (telecon), Chris Mauche, Chris Reynolds, Maria Santos-Lleo, Lisa Storrie-Lombardi, Steve Snowden, Greg Taylor, Leisa Townsley, and Knox Long

Summary

• Belinda Wilkes and Harvey Tananbaum welcomed the committee on behalf of the CXC. 
• Ed Mattison described the general status of the observatory, which has passed its sixth year of successful operation in July.  Basically, the spacecraft is in good shape, with an abundance of propellant.  The main operational challenge for the CXC has been due to the degradation of MLI which has caused certain on-board systems (particularly the EPHIN) to be susceptible to overheating. To combat this, the CXC has had to place limitations on the amount of time the spacecraft can spend at various orientations. If the EPHIN failed then it would be more difficult to protect the ACIS from problems associated with radiation damage.  The CUC fully understands that as the spacecraft ages there will be restrictions on operations, but are convinced that the CXC has taken good engineering decisions regarding them. In particular, the additional scheduling constraints appear to be well thought out and not overly restrictive, given the spacecraft pointing issues.
• Harvey Tananbaum reviewed the ways in which Director's Discretionary time had been used since the last CUC meeting, pointing out that he was well within the allocation of DD times.   The CUC continues to support the approach that the Director has taken in this area.
• Belinda Wilkes briefly described the experience the CXC had had with the last TAC, which was generally good.  Cycle 7 comprised 747 proposals, 12 panels, and 101 reviewers, including 10 VLPs, 56 LPs 78 Archive proposals. It was oversubscribed by a factor of 6.5.  The CXC did not, in the end, have to exclude any proposals from selection because of the limit on constrained observations of 15%, although the final allocation is almost precisely at the limit.  In general the review went quite well, although the CXC has some concerns about the time period allocated between the conclusion of the panels and the beginning of deliberations by the merging panel.  [CXC Update: The final number of constrained observations approved by the peer review was 22%, higher than our goal but lower than in previous cycles. Steps will be taken to reduce this number in Cycle 8.]
• Belinda also described the new policy for  when observations are considered complete.   The new rules increase the minimum exposure time to 90% from 80%.  This change was motivated in part by the fact that more and more observations are being broken into shorter chunks. Without the change a higher proportion of observations would end up with considerably less observing time than proposed.. The new policy also stipulates that observations of 200 ksec or greater must have an exposure time within 20 ksec of the approved time to be considered complete.  Very short observation (<5 ksec)  will only be repeated if the observation wias completely missed.  The CUC welcomed the increase in the mimimum completenesss of 90% and felt the  rules for short observations were appropriate in view of the overhead of scheduling short observations.
  
• Paul Green described responses of the TAC panel chairs to questions regarding the balance between GO, LP, and VLP programs.  This survey indicated that at least from the perspective of panel chairs the balance of programs is appropriate.  The CUC supports the CXC's assessment that the mixture of GO/LP and VLP programs is approximately correct.
• Mike Juda described the ongoing problems associated with EPHIN cosmic ray/charged particle monitor and possible alternatives if it continues to degrade.
• Paul Plucinsky updated the committee on the progress understanding the likelihood that a bake-out would reliably remove material from the ACIS blocking filter and improve the low energy response of ACIS.  The basic problem is that because of uncertainties in the properties of the blocking material one cannot accurately predict the rate at which material would be removed from the window and exit to space.  As a result, bake out will not be considered further for the foreseeable future.  The CUC fully supports this decision.
• Larry David described progress toward improving the calibration of Chandra.  The CUC was pleased at the progress that had been made toward better calibration, and that the calibration group seemed to be establishing priorities for calibration and then working off the problems in a reasonable manner.   It was especially pleased to see the improvement in effective area files, which largely eliminate the problems with the iridium-M edge
• Herman Marshall described a substantial interaction with the XMM data center on cross calibration of XMM and Chandra (which the CUC welcomes).
• Jonathan McDowell described plans for updating CIAO over the next 18 months and areas that the CXC recognizes needs improvement in calibration software.  The plans include distribution a portable version of SAOSAC/MARX that will allow users to make custom psf’s easily and the creation of encircled energy fraction tables in CALDB.  The CUC is pleased to see progress in this area, since it has regarded this as a significant problem for observers. Jonathan also described the plans for creating a Chandra source catalog, and an associated web-based GUI and VO-aware interface to the catalog

Discussion and Recommendations

Overall, the CUC is very satisfied with its interactions with the CXC.  Chandra is performing remarkably well and that to a significant degree is due to the efforts of the CXC staff.  While there are areas which we believe might be improved, as would be expected for any operation of this complexity, the number of users who have complained to the committee is actually quite small, which reflects the general satisfaction of the community

Areas meriting additional discussion:

Calibration -- The CUC was quite pleased with the presentation given by Larry David outlining the progress being made by the CXC with regard to calibration.  It's clear that most if not all of the calibration objectives embodied in the CXC's calibration plan are being met. This is to be commended. 

The only question posed by the committee for which the answers were less than clear had to do with the region (of ACIS) in which the calibration goals associated with effective area, gain and region were valid.  Many, if not most, ACIS observations involve a significant portion of the area of the detector and therefore ACIS science depends on an accurate characterization of as much of the detector as possible.  The CUC therefore requests that the CXC clarify this issue.  At a minimum it should make clear in documentation (and at the next CUC meeting) where calibrations apply.  If, upon investigation, the region is smaller than hoped, the CXC should consider what would need to be done to improve the calibration away from the aim points of the detectors.

Cross calibration -- In the report from the previous users committee meeting, we urged that the CXC pursue cross calibration more seriously.  Hence, the CUC was pleased to hear that regular interchanges are now taking place between the CXC and the XMM-Newton Science Operations Center concerning cross calibration of the two observatories, and even more pleased to begin to see preliminary results.   Since the capabilities of the two observatories are complementary in many respects and since many fields have been observed with both observatories, it is clear that cross calibration is of practical importance to observers.   So that observers can make use of this information, it is necessary for information being developed to be accessible to scientists using Chandra. 

Therefore, we request that the CXC document the current state of knowledge, both of the self-consistency of calibration between the Chandra instruments as well as the crsos calibration with XMM,  as soon as is possible.  We expect the cross-observatory comparison to improve over time and so we renew our earlier request that the CXC create a web page for cross calibration and start posting information there on cross calibration issues.

The CUC was also pleased to hear that some cross-calibrations with Suzaku are underway.   We note that it is imporatant that this cross-calibration include both hard and soft sources. We look forward to hearing of concrete results concering cross-calibration of Suzaku with Chandra and XMM, hopefully at the next CUC meeting.

Catalogs -- At the last CUC meeting, the CXC described the progress it was making researching the prospects for creating a catalog of all Chandra sources.  At that time, the CUC reiterated its request "to be consulted regarding the characteristics of the catalogue and on other Level III issues as they arise."  At this meeting, the CXC announced that it was reaching the point where it was considering beginning production of the catalog late next year.   While the CUC is pleased that the CXC is making progress toward this goal, the CUC was disappointed with the vague nature of these plans and the lack of any set of requirements that could be used as a measure both of the catalogue's utility and for determining when one should go forward with production of the catalog.  The catalog will, we hope, become an important resource for the community, and certainly seems to us to represent a significant effort by the CXC.  

We therefore renew our request that a systematic review of the goals and plans for the generation of a Chandra X-ray catalog be held.  If the CXC feels that the CUC is not the right body for such a review, it should constitute a committee to carry out the review, but the committee should include some external experts and at least one member of the CUC.  To be useful, this review should be held early enough that plans for catalog generation could be affected.    As a general principle, the CUC would argue that the CXC should adopt a phased approach to developing the catalog, and not be overly ambitions in terms of goals for the catalog in the hopes that catalog production could begin earlier rather than later.

Grant duration and extensions -- The CUC noted at the last meeting that the lengths of Chandra grants were short compared to other great observatories.  The Spitzer Science Center issues grants for 3 years (with no extension possible).  STScI issues initial grants of 2 years, and provides for a 1-year extension without requiring any special justification.   By contrast, nearly all Chandra grants are issued for one year, even though it is clear that essentially no projects actually go from receipt of data to publication in a year. Extensions for a second year on Chandra grants are possible with justification, and any further extensions must be "exceptionally well justified". 

Notwithstanding the comments by the CXC that they could not remember when an extension was not granted, the CUC believes that the CXC needs to attempt to bring its policies more in line with the other observatories.  The ideal situation would be a two-year initial grant with an "easy" option for a third year extension.  If that is not possible, then it needs to be made clear not just to observers but also grant administrators that the extension process is straightforward through the second no-cost extension. The crucial factor is that it be possible to straightforwardly extend a grant for 3 years.

Time allowed for budget preparation -- In the budget process now being completed for Cycle 7, the fair share allocations for approved programs was sent to PIs in early August with a deadline date of September 1st.  This is too short a period of time during the latter part of the summer when many PIs and grants people are on vacation.  If the fair share values are announced in early August then, the CXC needs to allow a minimum of six weeks for the proposal due date.  At a minimum,1 full month should be allocated, even if it is a time of year when few people are on vacation.  The CXC needs to recall that during this period PIs not only have to finalize their own budgets but in many cases must work out budgeting arrangements with their various co-investigators.

Conduct of the TAC -- Based on comments from the Director's office, the last proposal review went well. This was consistent with the experience of the members of the CUC who had served on the committee.  The CXC noted that they had received some comments that a few of the panels did not have a sufficient number of experienced X-ray or senior astronomers, and that as a result they were attempting to get more senior astronomers with X-ray experience to commit to serving early.  The CUC supports this activity; the quality of Chandra science is clearly dependent on the experience of the TAC. 

In addition, the CXC asked the CUC's advice regarding the time allowed for the various portions of the review itself, and especially the length of time between the topical panels and the merging panels. Currently the TAC is scheduled so that panels are on Tuesday and Wednesday. There is an organizational meeting of the merging panel on Wednesday night, a reading period on Thursday morning, and the merging panel takes place on Thursday afternoon. 

It is the CUC's opinion that 2 full days is the appropriate amount of time to allocate to the panels.   This is the current practice, and the CUC should not change this.  However, by the end of the second day, the panel chairs are by in large quite tired, and, based on the experience of those of us who have served on the merging panel, not able to do useful work.  Therefore the CUC recommends avoiding any activities associated with the merging panel at the end of the second day (Wednesday night).  Given the importance of the merging panel selections, the CUC believes this schedule is too abbreviated to allow a comprehensive review of (arguably) the most comprehensive observations that will be taken with the observatory, namely the LPs and VLPs.  The CUC believes that it is appropriate to allow a full day of preparation for the merging panel.  It suggests that the organizational meeting of the merging panel be held in late afternoon of that day (Thursday ), but the real deliberations of the panel take place on the following day (Friday), and that a full day be allocated for this process.  The CUC recognizes that this will add a day to the total commitment of the panel chairs, but also notes that this is no longer than the other great observatories. Given the location of the TAC at the airport, nearly all panel members should be able to participate until 4 pm or so on Friday afternoon and still make evening flights to the west coast or Europe.

Preparation for the meeting -- In the past, most if not all of the presentations were available on-line early enough that committee members could download the presentations to their laptops before leaving for Boston.  This allowed committee members to read the presentations so that they could spend more of the time during the meeting exploring issues, rather than hearing about them for the first time.  For reasons that are unclear, very few of the presentations were available for this meeting ahead of time.  As a result, committee members were less prepared for the meeting than desirable.   .

We urge that  the CXC DO exercise its authority and assure that presentations are available two working days in advance of the meeting.  This will give committee members (and the CXC Director’s Office) a chance to review the presentations and help to assure that the committees time is well spent.

Written response to this report: -- The CXC provided a written response to the recommendations made by the CUC following its last meeting.  This was extremely useful.  The ostensible reason for the written response was that the time between the January meeting and this meeting was exceptionally long.  But in fact this provided a clearer and more complete response to the committee’s recommendations than had been provided previously, and made it easier to develop an agenda for this meeting. 

The CUC believes that the CXC should make the production of the written response a standard part of its interactions with the committee.  We believe it is reasonable to attempt to produce such a response within a month of receiving the CUC's report.  The committee further believes that this response should be placed on the web alongside the committee report.  The committee understands that the DO will not always know, within a month, whether something is actually practical and so that some of the responses will be tentative, but nevertheless believe this approach will foster communications between the CXC and the CUC, and perhaps more importantly between the CXC and the community, especially those members of the community who have made suggestions to CUC members.