Back to main CUC page
Martin Laming (chair)
Jon Miller (by phone)
Steve Allen (by phone)
Harvey Tananbaum opened the meeting and welcomed the committee on behalf of the CXC. Members of the CUC then introduced themselves.
Edward Mattison (standing in for Roger Brissenden), gave the "CXC Manager's Status Report" for the period April - October 2009. The current budget is on track with the FY 09 operating plan. The CXC contract extension for up to 10 more years has been negotiated and is expected to be signed before the end of 2009. The feasibility of moving the OCC from Kendall Square to the Cambridge Discovery Park was studied by SAO and NGST. After considering benefits and risks, MSFC declined the move.
The spacecraft and instruments continue to operate very well, with nominal passages through the summer eclipse season. Three anomalies, all with benign impacts, occurred during the summer. The Momentum Unloading Propulsion System (MUPS) pressure reading became unreliable (discussed further below), a grating status switch unexpectedly began to operate (having "failed" earlier in the mission), and an incorrect command sequence was uploaded.
The observing efficiency remains very good, at greater than 70% for the whole April - October period, even during the MUPS anomaly. Thruster warm starts and MUPS fuel usage are also very good. In fact, during the MUPS anomaly the spacecraft was operated without expending MUPS fuel. Guest observers continue to receive their data within approximately one day of observation, and grant issuance is current with an average time to award of less than three weeks.
The CUC remains completely satisfied with the professionalism and efficiency with which Chandra operations are carried out.
Harvey Tananbaum briefed the committee on the use of Director's Discretionary Time. In the period April - October 2009, a total of 86 ks was used, with a single longest observation of 25 ks. All observations sought to take advantage of the high spatial resolution available with Chandra. Out of the yearly allocation of 700 ks, with 400 ks unused from previous cycles, there is now a surplus of around 1 Ms. It is proposed to use this, together with 500 ks from the next two cycles to increase the exposure time on the Chandra Deep Field South by 2 Ms for a total of 4 Ms, of which 1.2 Ms will be observed this coming spring, and the remainder in spring 2011. A special archival program of $500K will be made available to support data analysis. We return to this specific issue in our recommendations section below, but otherwise support the director in his use of DDT.
Belinda Wilkes summarized the Cycle 11 peer review. A total of 668 GO proposals, requesting a total of 92670 ks, were received, representing a 5.6 oversubscription in time. These break down into 455 GO, 50 LPs, 9 VLPs. Requested observing time for gratings observations was down by 32% from Cycle 10. The LP and VLP time requested was also down from Cycle 10 by about 6%. In other categories, there were 65 archive proposals (oversubscription of 4.4) and 40 theory (oversubscription of 5.3), though these oversubscriptions have been reduced to around 3.3 by augmenting the archive and theory budgets. The cost review is under way, with only 18 proposals asking for significantly more than the "fair share" proposal amount.
For Cycle 12, it is proposed to make a separate allocation of funds for the Chandra Deep Field South archive proposals, and to restore the separate allocations of time for LP and VLP proposals. We revisit these issues in our discussion and recommendations section below.
Andrea Prestwich updated us on the CXC Website overhaul. Phase 1, including the redesign of the top page and the identification of obsolete material (obsolete in the opinion of the authors) is complete. Phase 2, infrastructure improvements, is underway to remove all obsolete material, and create new instrument pages with clearer links between the calibration results and data analysis/observation.
Tom Aldcroft described the activities related to resolving the MUPS anomaly over the summer. On July 19th 2009, the pressure gauge monitoring the hydrazine fuel showed a drop, equivalent to the loss of 5 lbs of propellant. Within four hours, operator intervention closed the MUPS bank-A isovalve. The two most credible root causes for the reported pressure drop were a sensor malfunction and an actual propellant leak (potentially self-sealing due to ice build-up as observed in tests). However in the latter case the expected temperature drop and disturbance torque were not observed.
Although a spacecraft emergency was not declared, a number of measures were taken to handle the situation. The FOT operations crew went to 24 hour operations and significant support from NGST engineering staff and every spare minute of DSN communications were obtained. Surprisingly, there was no interruption of science observations during this period, although the pitch angle restrictions were more severe (to keep the MUPS tank cool; further anomalies were noticed during periods of higher tank temperature) and no momentum dumps were performed.
The anomaly was finally confirmed to be a transducer failure rather than a leak by firing the thrusters, where an efficiency was obtained consistent with that expected on the basis of no loss of fuel. The decision to do this came after a large amount of analysis had determined that a transducer failure was the most likely scenario, and the recognition that maintaining high alert operations while still not understanding the anomaly carried increasing risk itself.
Belinda Wilkes summarized changes to the program of observations instituted during July-August 2009 in response to the anomaly. In particular, recently approved Cycle 11 targets which were available for observation within the restricted observing conditions were ingested and approved early, in consultation with the PIs, to schedule them and maintain the observing efficiency. Rapid response from the many Chandra observers and PIs involved with this process was critical. The successful resolution of this anomaly is due in no small part to the dedication of the operations team and their calm response to the apparently burgeoning crisis. It also clearly shows the value of having a team of experts available at close hand to call on when necessary.
Larry David reviewed the status of the Chandra calibration. The quality and stability of the Chandra telescope pointing and angular resolution are demonstrated by a series of images showing the proper motion of Ar Lac. Work on the temperature dependent gain correction for ACIS has also been brought to completion.
The HRMA calibration has also been further studied. Ground calibration data has been reanalyzed since January. A few earlier errors have been corrected, a better treatment of pileup has been implemented, with a net 3% change in effective area, well within the uncertainty range of the values released in Feb 2009.
The ACIS filter contamination continues to be studied. The External Calibration Source is exposed just prior to and following every Chandra pass through radiation belts, although the flux from this source is decaying. Plots of opacity at Mn L and F K indicate more contamination at the edges of the filters than in the center. ACIS-I appears to be more difficult to track and model than ACIS-S primarily due to the lack of gratings data on ACIS-I. An updated contamination model that incorporates the recent trends in the contamination deposition rate as well as an empirical correction based on observations of the Coma cluster taken 10 years apart is scheduled for release in CALDB 4.2.
Finally the cross calibration of Chandra with other X-ray observatories was presented. For the O VI, O VIII, Ne IX, and Ne X lines observed by ACIS-S3, this also appears improved. HETG-MEG observations were missing from the plot.
Jonathan McDowell gave a very detailed account of the activities of the Science Data Systems Group. The CIAO 4.1.2 patch was released in April, and the CIAO 4.2 release is in test for a December release. The CIAO helpdesk has received 189 new tickets, of which 11 were still open. Initial response times and times to ticket resolution remain, as ever, entirely satisfactory.
The CUC are gratified that many of our previous suggestions have been taken up and implemented by the SDS group. The user survey seems to have been useful, and we acknowledge the "outreach" efforts that the group have been making. We also appreciate the progress in making example threads and scripts available. The thumbnail SHERPA examples strike us as likely to be particularly useful, as does the "ciao_install" script to be released with CIAO 4.2.
R&D work on the point spread function was described. The MARX ray generator has been improved to match SAOTrace better, both will next be compared with real data. Other work concerned the possibilities of better exploiting the parallel properties of modern pcs, which could benefit computationally intensive work such as fitting detailed spectral models to data. Research is continuing, and a report is anticipated.
Finally, it has been decided to phase out support for the S-Lang interface in Sherpa and ChIPS, in favor of Python. The CUC endorses this decision. Future plans include a phasing out of support for Mac PPC, with Linux and Mac-Intel being the main user platforms.
Ian Evans presented an update on the Chandra Source Catalog (CSC). Version 1.0.1 was released on July 21 2009. The CSC has also been released to be viewed in Sky in Google Earth, with 500,000 hits so far from this interface, believed to come from a few thousand distinct IP addresses. The next release for Sky for Google Earth will include pop-up balloons with primary source information, and it is possible that links to allow the retrieval of FITS data products could also be implemented.
Catalog Release 1.1 is expected early 2010, to include public HRC-I observations as well as newly public ACIS data, but otherwise to have to same limitations as release 1. In the longer term catalog release 2 will co-add multiple observations of the same source that use the same instrument, and improve background modeling. Detailed plans are not fully established, and will be reported on at the next CUC meeting.
Ken Ebisawa demonstrated two additional web capabilities using a fixed URL prefix, known as a "permalink", that allows users to access public data, proposal abstracts, and to search data and catalog results for sources, without using a GUI. As well as the "permalink", a batch interface would ease the accessing of the catalog for multiple sources. Chaser already provides permalinks via obsid, but analogous simplified links and batch access to Chandra data should be enhanced.
The CUC strongly endorses all the actions taken to resolve this issue, and congratulates all personnel on the successful outcome. Having the necessary expertise available "on call" was clearly critical.
The CUC endorses this use of Director's Discretionary Time, and the suggestion to have archive proposals compete for the fund available to analyze this dataset at the Cycle 12 Peer Review. We recognize that a number of (good) proposals to do just this have been submitted in recent years, but have not made it through the peer review process. While peer review is the best selection mechanism we have for proposals, it is not perfect, and the use of DDT to remedy such imperfections is wholly appropriate.
Tell the community as soon as possible that this is going ahead. It looks like this has already been done to some extent in Chandra bulletin #89, e-mailed in 11/5/2009, and in AAS Electronic Announcement #202, e-mailed on November 11.
It was noted in Cycle 11 that several of the LPs approved asked for substantially more than the 300 ks LP threshold, some for 700 - 800 ks, and so we felt that there is no real bias against long observations in the big project panel. Notwithstanding our endorsement of the 2Ms DDT observation of the CDFS, the CUC is unanimously opposed to restoring separate allocations for LPs and VLPs in Cycle 12.
We recommend for Cycle 12 at least, not to go ahead with separate LP and VLP time allocations. This is a difficult issue and should probably be revisited every year after the peer review. We are generally cautious about the unintended consequences of "over-engineering" the peer review, and prefer to leave this process alone to decide the relative merits of proposals. Of course in extreme cases, such as the Chandra Deep Field South, external intervention is necessary and desirable. We don't feel that the relative lack of approved VLP projects in Cycle 11 yet rises to this level of concern.
Great progress has been made. We like the new instrument sites, and the way they are linked to the rest.
The web site should be exercised externally before the Cycle 12 Call for Proposals. Some of the CUC members expressed a willingness to help in this. We also endorse better integration of the MIT sites with the Chandra web site, with e.g. lots of reciprocal links.
We endorse the choice of Python over S-Lang for the scripting language. It seems that Python is going to be the more widely used, and this change lowers the threshold for new users. We also endorse dropping support for Mac PPCs.
The CIAO survey seems to have been valuable, and we suggest doing this more often. Not every year, but maybe every 2- 3 years. We also recommend looking into providing support for the Mac Snow Leopard platform, and like the idea of keeping CHIPS and SHERPA separate from CIAO in the distribution.
In the area of long term planning, we encourage discussions between the different projects (XMM, Suzaku, etc) to coordinate approaches to issues such as platforms, operating systems, libraries, compilers etc.
The CUC feels that the 2010 Senior Review is of paramount importance, especially because Chandra will be held to a high standard as the flagship high energy astrophysics mission. Both in terms of science, and science/dollar, the CUC feels that Chandra should strive to achieve the top ranking.
The CUC urges Chandra to engage with its observing community to help develop the strongest possible proposal, and to do so on a timescale that permits revisions and reflection.
We recognize the extensive calibration effort that has been applied to the Chandra telescope, and the dedication of the calibration team, and endorse the approach of the calibration group in trying to understand the instrumentation as thoroughly as possible. Having said this there are still some aspects of the calibration that cause concern, particularly the cross calibration. We are aware that problems here might not be Chandra problems, but associated with the other missions. While the reanalysis of the ground calibration data undertaken since January appears to have improved things and validated the recent CALDB release, we note that the cross calibration plot only showed data from ACIS, where O VII and O VIII lines are blended together, as are Ne IX and Ne X. Data from HETG-MEG for these lines would be much more compelling, especially in comparison with the XMM Newton RGS.
Progress on the filter contamination looks good. We appreciate that this problem is exceedingly complex, given its temporal, spatial and spectral dimensions. However ideally the CUC would really like some more insight into what the calibration team knows and how they are dealing with it.
Concerning the Calibration website: We were told in response to our 4/09 report (see notes 1.2 and 1.3 from that report) that the Calibration team would add "new sections regarding hot topics" (like contamination and cross-calibration, we presume) and that "some prototypes for these new sections" would be presented at the 10/09 meeting. Most of these issues were discussed, but not in the context of the CXC Calibration web page, which still poses difficulties for some users.
The cross calibration between the HETG-MEG and the RGS should be completed as soon as possible. With the change of effective area in January 2009, this comparison worsened (see e.g. slides 13 and 14 of http://www.mssl.ucl.ac.uk/~gbr/workshop3/papers/plucinsky_mssl09_small.pdf), and it would be interesting to see whether the application of the new contamination model improves this, as it appears to do for ACIS-S3.
As has been mentioned before, our main concern is not so much with the technical aspects of the work, which we really cannot fault, but with communication with the user community who need advice that is as clear and well justified as possible. One possibility might be a comprehensive "Status of ACIS Calibration" memo describing the issue. Such a memo should address (a) what is known about the problem (i.e., collect and explain all the relevant data in a coherent fashion), (b) the various models proposed, and their strengths and weaknesses, (c) the near-term implementation of a contamination model(s) in the CALDB, and the rationale behind the implementation (as we understand the informal discussions during our 10/09 meeting, some sort of hybrid model was just put in place, consisting of one piece for ACIS-I and one piece for ACIS-S), and (d) the calibration team's longer-term plans for contamination monitoring and modeling.
Significant progress has been made since the last CUC meeting in terms of accessibility. The Google Earth interface is a very nice thing to have, and will probably prove to be of more use in getting non-experts involved in X-ray astronomy than for professional astronomers.
The 500,000 hits on the Google Earth interface, even if only from a few thousand distinct IP addresses, is impressive. We are prompted to suggest tracking scientific use of the catalog in this way, maybe for the Senior Review in two years time, and to encourage authors to acknowledge use of the catalog in publications. We also encourage the importation of extended sources to the catalog as soon as possible.