Chandra Users Committee Meeting, 5-6 April 2006

Attendees:
Mark Dickinson, Vicki Kalogera, Nobuyuki Kawai (telecon), Julia Lee, Smita Mathur (telecon), Chris Mauche (telecon), Chris Reynolds (chair), Maria Santos-Lleo, Steve Snowden (telecon), Lisa Storrie-Lombardi, Allyn Tennant (ex-officio), Leisa Townsley

• Belinda Wilkes and Harvey Tananbaum welcomed the committee on behalf of the CXC

• Roger Brissenden summarized the current status of the CXC and the observatory. The CXC experienced a 5% reduction in its FY06 budget mid-year (February); steps are still being taken to absorb this reduction. The CUC is grateful for the CXC's clear commitment to protect the GO funds but acknowledges that some reduction to GO funds may have to be tolerated in order to keep the operational component of CXC intact. The CUC strongly supports the CXC position that the current proposed FY07 budget be maintained and not reduced any further. The observatory itself appears to be operating well and the CXC are efficiently and effectively investigating and solving anomalies as they arise. The thermal regulation of the EPHIN and the fuel lines remains the only significant technical issue. The CXC recently decided to increase the upper temperature limit for the EPHIN, thereby relaxing scheduling constraints and increasing efficiency. The CUC appreciates the extensive efforts made by the CXC to maintain observing efficiency and flexibility and applaud this latest initiative.


• Harvey Tananbaum described the Cycle-7 Directors Discretionary Time (DDT) to date. The CUC continues to support the Director's general approach and philosophy to DDT.


• Belinda Wilkes presented the activities of the Chandra Director's Office and, in particular, the up-and-coming Cycle-8 review. The general over-subscription rate is holding steady at about 6 to 7 (by time), demonstrating the high-quality of the science that continues to be performed by Chandra. The most significant changes to the Cycle-8 review from previous cycles concerns the Big Project Panel (BPP). The review now incorporates a significantly longer "reading-period" during which time members of the BPP can review the LP and VLP projects that have been most highly rated by the topical panels. The need for such a reading period has been highlighted in past CUC reports and the CUC applauds the CXC for initiating the new review timetable.
  Previous CUC meetings have strongly recommended that GO grants be awarded for an initial period of two years as opposed to the standard one year. Starting in the Cycle-7 GO round, the CXC is now allowing the GO the option of a two year initial grant. The CUC thanks the CXC for this change.


• Larry David presented the latest calibration update. In direct response to comments from the last CUC meeting, Larry discussed the spatial regions over which the ACIS has been calibrated. The CUC continues to support the good work of the calibration group towards systematically achieving the calibrations goals of Chandra.


• Hermann Marshall summarized ongoing efforts to cross-calibrate Chandra and XMM. A potential new methodology for formal cross-calibration based on the application of "deviation functions" was presented. The CUC generally supports this important effort. The CUC did, however, have some concerns. Firstly, it is not apparent how the cross-calibration work meshes with the work of the Chandra calibration team, i.e., will results from cross-calibration be fed back into the primary Chandra calibration? Secondly, it is not clear that specific user needs have been carefully considered when planning the cross-calibration analysis. Cross-calibration has different practical meanings for a user who wants to compare spectral parameters derived from different epochs of Chandra and XMM data than it does to a user who is performing a genuine joint spectral fit. [see cross-calibration recommendations below].


• Jonathan McDowell discussed several major items related to software and analysis. Firstly, he presented the results and recommendations of the review panel that convened in Feb 2006 to examine plans for the Chandra Source Catalog (CSC). The CSC review made a set of 11 recommendations, 6 of which were flagged as urgent. The spirit of these recommendations was to encourage the CXC to produce a formal requirements document, descope the first version of the catalog in order to expedite its release, and to aim the catalog at the multi-wavelength astronomer. The CUC strongly endorses these recommendations but remains concerned about the apparent low-priority of this project and the uncertain timetable for completion of the first release. [see CSC recommendations below].
  Secondly, Jonathan presented an update on CIAO and related software. In response to a request by the CUC, the issue of the testing and debugging of CIAO was given special attention. Given the complex nature of this software and the limited resources available, the CUC feel that the testing and debugging of CIAO is being performed well.
  Thirdly, in response to a request by the CUC, the analysis of grating data with a piled-up zeroth order image was addressed. It was noted that warnings about these kind of data plus a link to the (imperfect) "correct zeroth order source position" thread do already exist on the CIAO web-pages, although it was acknowledged that they might not be sufficiently clear. The CUC were informed of work on a new algorithm which corrects the zero-order position using the ACIS read-out streak.
  Fourthly, an update on Repro-3 was presented. This latest reprocessing is now underway. The CUC are pleased to see the posting of the "how will this affect my processing?" web-page, although it was noted that this page was not accessible from the top-level pages.
  Finally, in response to a request by the CUC, the issue of combining multiple observations for spectral analysis was discussed. It was acknowledged to be an important but technically challenging issue which is currently under investigation by the CXC. The CUC urge the SDS to continue work on this important issue.
  Informal discussion at the CUC meeting revealed that several people have concerns with the current ACIS bad pixel table, which was changed in 2005 to include a much larger number of bad columns than in previous versions. These bad columns exhibit extra events at low energy (0.2--1 keV) and "turned out to cause problems for some sensitive analyses" (cxc.harvard.edu/cal/Acis/Cal_prods/badpix/badpix_2005_02_21.html). Since the bad column plus its flanking columns get included in the bad pixel table, these added bad columns non-trivially affect the ACIS sensitivity and are not nessarily "bad" for all analysis situations. The CUC noted that it would be helpful for many users to have a convenient way to access a bad pixel file appropriate for the data (e.g. having appropriate bad bias values noted) that does not have these "warm" columns included, and/or to have a tool to allow quick and easy manipulation of the bad pixel files to remove unwanted bad columns (and their neighbors) or to include new bad pixels/columns discovered in the analysis. At the same time, the CUC feel it is important to maintain the simplicity and robustness of the bad pixel analysis for the general users.


• In the final presentation of the day (late addition to the agenda), Mike Nowak presented statistics related to the fraction of awarded GO targets and GO observing time that is allocated to grating observations. Concern was expressed that the "expensive" nature of spectroscopic grating observations results in a bias against such observations being awarded by the peer review. While the CUC is sympathetic to this issue, it does not recommend any solution that side-steps the peer review process (e.g., we do not recommend a special block of time set aside for spectroscopic observations). We do, however, strongly support efforts to increase the community awareness of the importance of high-resolution X-ray spectroscopy through, for example, special sessions at October's HEAD meeting.

   

  DISCUSSION AND RECOMMENDATIONS
  

  The CUC continues to be extremely satisfied by its interactions with the CXC. The CXC's support of the Chandra observatory and associated programs is consistently of the highest quality.

  The CUC is also extremely pleased by the responsiveness of the CXC to previous CUC recommendations and requests. Most recently, the implementation of the 2-year initial GO award is most welcome and the CDOs office is to be commended for overcoming the administrative hurdles required to make this happen. Furthermore, all of the specific CUC queries send to the CXC prior to this meeting were addressed thoroughly and professionally.

  On the basis of the presentations and subsequent committee discussion, we have a number of specific suggestions and recommendations for the CXC.

  1) Prioritizing the Chandra Source Catalog : The CSC is an important part of the scientific output of the Chandra observatory and we commend the work that has been performed to date. We believe that the Level-3 review panel was extremely useful and strongly endorse the recommendations of this panel. The CUC remains concerned, however, about the timetable/schedule for the first release of this catalog. It seems like this important project has been relegated to that of a "spare time project" for the SDS group resulting in large uncertainties about the release date.

  The CUC urges the CXC/SDS to take a serious look at the prioritization of the catalog in relation to the other on-going projects pursued by the SDS group. While the CUC fully acknowledges (and entirely supports) that spacecraft operations and basic data analysis needs must take priority, it is unclear that some of the other projects consuming the SDS group's time (e.g., the Sherpa/CHIPs update) demand a higher priority than the source catalog.

  2) Communication of basic technical information to the peer review panel : The concern has been raised that some members of the Chandra peer review committees may be applying incorrect preconceptions about the capabilities of Chandra, especially in comparison with XMM and Suzaku, and that this may affect the review of some proposals. It was felt that this is especially a concern for spectroscopic (grating) proposals. The CUC recommends that, starting in the Cycle-8 review, the CXC produces a simple one-page technical description showing, for example, plots of the effective area and spectral resolutions of the various instruments/modes of Chandra, XMM and Suzaku. This sheet should be distributed to the proposal reviewers along with the proposals, and should be widely available (in hard-copy format) at the time of the review. The CUC requests that a draft of this sheet be provided to the CUC at the time of the written CXC response.

  3) Announcement of future CUC meetings via the Chandra bulletin : Based on the small number of community complaints received by the CUC, we believe that the community is basically satisfied with all aspects of the Chandra program. However, the CUC would like to maximize the feedback received from the community. Thus, the CUC recommends that an announcement of future CUC meetings be made to the Chandra community (most likely through the Chandra Bulletin). The announcement should be at least 4-6 weeks prior to the meeting thereby allowing new agenda items to be added as necessary. The announcement should list the names and email addresses of the current CUC members as well as provide a link to the CUC website.

  4) Better definition of the cross calibration efforts : As stated above, the CUC is pleased to see continued and systematic cross-calibration activities with XMM. However, based on the concerns raised above, the CUC recommends that the calibration team and the cross-calibration team formulate a strategy for the (eventual) integration of their respective works and that this strategy be included in the Fall 2006 presentations to the CUC.

  Furthermore, it is unclear that the current cross-calibration efforts are being guided by a consideration of what is most useful for the "real" user. The work currently being performed is most relevant for those users who are performing a direct joint fit of XMM and Chandra-grating spectra. A more common occurence, however, will be a user who wishes to compare the flux and spectral fit parameters from XMM and Chandra observations taken at different epochs. It is important that the cross calibration efforts serve this type of user as well by quantifying the systematic errors on the flux and spectral parameters.