Ex-officio members present: Fred Seward and Allyn Tennant.
NASA HQ personnel present were: Alan Bunner and Don Kniffen
Harvey Tananbaum, CXC Director, reported on the status of Chandra. The launch of Chandra is currently scheduled for July 20, with the two subsequent days as a back up. After that time, the Eastern Range will be closed for maintenance for several weeks. If the shuttle launch occurs as planned, Chandra would be deployed early in the mission, on the first day. If all goes well, first light for Chandra will be 17-20 days after launch.
On April 9, 1999, a Titan rocket carrying the same Inertial Upper Stage (IUS) which will lift Chandra to its final orbit failed. The problem appears to have been due to the failure of an electrical connector to disconnect as required. The connectors on the IUS for Chandra have been thoroughly tested, and do not show this problem.
About one year ago, the actuator for the door to ACIS failed during a test. The ACIS door on Chandra has been tested extensively since that time, and has worked every time.
Tananbaum discussed the effect of the Chandra launch delays on funding for
the CXC and for the Guest Observer Program. The committee noted that the
delays have made it difficult for observers to hire graduate students and
post-docs, and suggested that the funds be made available as quickly as
possible within the rules of NASA, and the period for the AO-1 grants be
extended, again if possible.
Roger Brissenden (CXC) reported on the status of the CXC. There have been no significant changes in the scope of the CXC since the last meeting. The GO Grants Program personnel are now in place.
The Operations Readiness Review occurred in December 1998. The second Flight Readiness Review will occur on June 15. The OCC has been involved with training and simulations of operations in a variety of different modes.
The long term observation schedule for Cycle 1 is nearly done, with
observations blocked into weeks.
Martin Elvis (ASC) presented the schedule for releases of the data analysis software. The software system is now called CIAO (Chandra Interactive Analysis of Observations). Pre-release 1 occurred on February 15, 1999 to the Science Working Group, the Users' Committee, beta test sites, and the CXC. It supported Sun Solaris only. In addition to pipeline functions, it included software from the data model, software for source detection, fitting software (called Sherpa), and plotting software (now called ChiPS). Pre-release 2 occurred on June 1, 1999 to the same test groups. In additions to improvements and new documentation, it supported Linux (Redhat and Slackware).
The first public release of the software will occur in August 1999, modulo some uncertainty due to the uncertain launch date. It will include a GUI interface to the tools. The second public release is expected in late fall 1999. Functions for analyzing grating data will be wrapped as tools into this release.
The committee was happy with the progress that had been made on the data analysis software, and generally pleased with the pre-release versions 1 and 2 software and documentation they tested.
They asked that barycentric correction software be included in the first public
release of the software system, even if the routines are not integrated
into the rest of the software system. The committee also suggested that all
of the software manuals should contain a cookbook showing typical data
analysis commands and walk-throughs. They suggested that the CXC maintain
a searchable web page under topical groupings with frequently asked questions
about the software.
Fred Seward (CXC) described the review of targets from AO-1 which was done by the CXC with the PIs of the proposals. These included correcting errors in coordinates, which ACIS chips were turned on, roll constraints, etc. Because of the number of errors associated with ACIS modes, the committee recommended that the user documentation include clear instructions for the typical choices of ACIS configurations. The typical default choices should be built into RPS as defaults. Similarly, the correct frame times should be automatically built into RPS for default observations.
In mid-June 1999, the long term observation schedule (blocked by weeks) and
parameters of the observations will be put on the web, which will allow
additional checks by the proposers. The web pages will show overlays
of the instrument on the optical and (if available) the X-ray field.
The committee recommended that the position of the X-ray source being
observed be marked on the optical and X-ray overlays.
Belinda Wilkes (CXC) described the plan to allow coordinated Chandra and Hubble Space Telescope proposals during the next AO of each of the observatories. The purpose of this plan was to facilitate such observations by allowing observers to submit a single proposal to whichever observatory was viewed as primary for the science. In the next AO, STScI and CXC have agreed to exchange time. The HST TAC would be allowed to allocate up to a total of 400 ksec of Chandra time to coordinated proposals which were granted time on HST, and the Chandra TAC would be allowed to allocate up to 100 orbits of HST time to coordinated proposals which were granted Chandra time. The time on the secondary observatory would be allocated based on the relative rankings of the proposals by the primary TAC. HST proposals with coordinated Chandra observations would have to contain enough information on the Chandra observation to allow a preliminary technical review. Successful HST proposals for coordinated Chandra observations would be subject to a final technical review. The HST TAC membership would be determined in consultation with the CXC to insure sufficient X-ray expertise in the panels. The targets assigned this way would be included in the lists of previously selected targets to avoid conflicts.
This program might eventually be extended to SIRTF.
The committee felt that this program was an interesting test, but that it should be viewed as an experiment, rather than a permanent change. The committee asked that the proposals which were awarded Chandra time by the HST TAC during the next AO should be compared to the lowest ranked accepted proposals for the same amount of total time which were granted Chandra time by the Chandra TAC. The committee asked that the abstracts of these two sets of accepted proposals be made available to the committee at its first meeting after the selection process is complete.
The committee also wanted to make sure that the TAC for the primary observatory
would have the options to turn down the entire proposal, approve just the
primary observatory time, or approve both the primary and secondary
observatory time. They should not be allowed to approve only the secondary
Nancy Evans (CXC) reported on the second year selection of Chandra Fellows. She said that a Chandra Fellows Symposium would be held on October 25, 1999 at the CXC. The committee expressed the hope that the symposium would be well-attended by CXC scientists and other local astronomers.
During the second year, one person who was offered a Chandra Fellowship turned it down, but asked if they could retain the title of Chandra Fellow. The CXC named this person as a Chandra Fellow-at-Large. The committee worried that it was confusing to have different classes of Chandra Fellows, and voted to recommend against having any future Chandra Fellows-at-Large by a margin of 9 - 0 (1 abstention).
There was a discussion of the institutional rule for Chandra Fellows, which had allowed only one Fellow per institution total (i.e., 1 out of 15 in three years assuming all Fellows remain in the program for the full term). The committee voted 9 for, 0 against, 1 abstention to recommend that the rule be changed to allow at most one Fellow per year per institution (i.e., up to a total of 3 Fellows at any institution at any one time). This would also have the effect of allowing applications for any US institution in every round of the selections.
The committee was concerned about the relevance of the research of some of
the selected Fellows to CXO, but felt it would be ineffective to narrow the
focus of the Chandra Fellows program. It was noted that the reports sent
in by Fellows for renewal should highlight the relevance of the work to
Nancy Brickhouse (CXC) reported on the progress in the Emission Line Project, an effort to improve the data and models for X-ray emission lines both for and from the Chandra spectra. The project is organized by a core working group led by CXC. They plan to have a website for distributing and acquiring information. Addition external (to CXC) involvement will be solicited in a "Dear Colleague" letter. This will request proposals of two types. Level 1 proposals would be for external involvement, but no funding. Level 2 proposals would compete for a total of $150K of available funding; approximately 10-15 proposals, each about $10-15K, might be funded.
The committee suggested that the project website include a mechanism
for collecting problems which people encounter during the analysis of
Chandra spectra. The committee expressed its continued support for the
emission line project, but voiced a concern that it be managed actively
to support the greatest scientific progress from Chandra spectra.
The committee heard two reports on efforts to understand and possibly deal with pile-up in the ACIS ccds. Pile-up occurs when two photons are received in the same area of the ccd before it is read out. It degrades the sensitivity, spectral response, and spatial resolution of Chandra and other X-ray observatories.
Brian McNamara and Mike Wise (CXC) presented the results of an effort to understand and partially "fix" pile-up by "forward modeling." They used models for the actual spectrum and spatial distribution of X-ray photons from the sky. These models were propagated through the response of Chandra using MARX. The resulting signals in the ACIS chips were determined, including their "grade" (the geometric pattern of charge in nearby pixels). The resulting spectra and spatial distributions were compared to the observed spectra. The initial efforts have been aimed at understanding the ground calibration data. A crude agreement has been obtained, but requires an arbitrary choice of the grade of the events selected and the detection cell used to accumulate spectra. A better physical model of the ccd chips and their X-ray interactions was needed.
This "forward modeling" method would lead to a technique to model spectra and correct for pile-up by comparing the predicted and actual observations (using chi-square or another statistic) and iterating until the best fit was found. This forward modeling technique would be computationally slow. It could be speeded up considerably if the initial guesses for the source spectra and other properties were moderately accurate. It might also be possible to formulate a faster, approximate version of the technique which would provide the initial models for the accurate modeling. The committee was very impressed by this effort.
Aneta Siemiginowska (CXC) also described efforts to apply Bayesian
statistical techniques to the pile-up problem. These efforts also
involved statisticians from Harvard University. The committee felt
that these techniques were very interesting, but it was still unclear
whether they would really prove useful for actual data.
Harvey Tananbaum reported briefly on the ChaMP, the Chandra Multi-wavelength
Project. Following some earlier E-mail correspondence with the committee,
the service and science functions of the project have been separated.
There will be no direct CXC financial support for the science portion.
The service portion will consist of providing Chandra observers with
existing optical data (from the Digitized Sky Survey) for their fields.
Eventually, Sloan survey data will be provided. The science portion of
ChaMP is being done by individual CXC scientists using their science work
time and by external observers. Proposals have been made to NASA and NOAO
for funding and observing time, using normal proposal channels.
Fred Seward (CXC) presented the plans for the second round of Chandra GO proposals. Assuming there are no further launch delays, the NRA would be released on December 10, 1999, and the proposal deadline would be March 10, 2000. The scientific review would occur at the end of May 2000, results would be announced in mid-June 2000, cost proposals for successful observing proposals would be due about August 1, 2000, and the cost review would be complete in early September 2000. The committee recommended that the deadline be made 23:50 (11:50 pm) on the due date to avoid the confusion about which date contains midnight.
In Cycle-2, approximately 74% of the time will be allocated to GO proposals, and 26% of the time will go to GTOs. The committee was asked if the GTO targets could be listed on the web, rather than being included with the hardcopies of the NRA. Hardcopies will be available by request using mail or fax. This will save some of the time involved in printing the hardcopy materials. The committee felt this was acceptable.
For Cycle 2 proposals, the proposal text would be submitted electronically. There was a discussion of formats for this (i.e., Postscript vs. PDF vs. other formats). The proposal text submission would probably occur via web up-loads. The committee suggested that some means be provided for proposers to test that accepted proposals would print correctly. The CXC also said that it be careful to have sufficient printing resources available.
Changes to the NRA for Cycle 2 will include the discussions of Large Projects, a description of Director's Discretionary Time, and the discussion of Coordinated CXO/HST proposals (see above),
The committee discussed the problems posed by the likely discovery of changes
in the response of the instruments or errors in the documentation during
the proposal writing period. The committee suggested that, if major
changes in the instrument behavior are discovered after the release of the
NRA, they should be posted prominently on the web site. This should be
limited to significant changes which might affect the feasibility of
proposed observations. The NRA should include a warning sentence,
indicating that proposers should check the web site to get the latest details
on instrument performance
As agreed previously, during Cycle 2 there will be a separate class of Large Projects, which are defined as those requesting at least 300ks. At least 20% of the observing time will be reserved for such proposals. Such proposals will be indicated as such by the proposer during the submission.
The committee discussed the length of the science justification text in Large Projects. In other proposals, this is limited to 4 pages. The committee voted 6 - 4 in favor of allowing 6 pages (rather than 4) for science justification in Large Projects.
During the review, the Large Projects will be graded by the individual
panels on the same scale as the regular proposals. The selection of the
approved Large Projects and targets will be made during the merging process
following the individual panel meetings. The committee discussed whether
this should be done by a merging panel consisting of the chairs of the
individual panels, or by a separate smaller panel which would hear
presentations of the Large Projects by the chairs of the panels. It was
unable to reach a conclusion concerning this.
Fred Seward (CXC) presented the changes for Cycle 2 to the scientific review process for GO proposals. Unlike the Cycle 1 review, CXC, GTO, and IPI scientists would be allowed to participate as panelists. There will be an effort to increase the foreign participation in the panels. The panel chairs will be selected early, and will be encouraged to participate in the panel selection and the assignment of proposals to panelists.
To improve the overall coordination and coverage of the science program, each panel will receive copies of the titles, abstracts, and target lists of all the proposals. The panel chairs will meet on the morning of the second day to compare the science coverage of proposals within different panels.
New proposal review forms are being prepared with check-offs to simplify the writing of reports. There will also be a check-off of the level of effort required to do the science which will be used during the subsequent cost review.
There was a discussion of the advantages and disadvantages of making the
names of reviewers public after the review. The committee voted 4 - 4
(2 abstentions) on this. Given the ambivalence, the names should remain
There will be an automatic, computerized technical review as in Cycle 1. There will also be prior human technical reviews. However, the committee reiterated its concern over the reliability of the human technical reviews. They again noted that these review should be limited to major technical problems. They should not discuss the science. The technical reviews should not simulate the proposed observations in any detail. Most human technical review should simply say "no problems found."
The committee asked that the human technical reviews be screened by a senior CXC staff member, who would officially authorize the release of the reports.
There was a discussion of CXC and NASA efforts to avoiding using panelists
who have acted irresponsibly in past reviews.
Wallace Tucker (CXC) discussed the plans for press releases associated with the Chandra launch, first light, and subsequent observations. The staffing and graphical capabilities were described. An effort is being made to provide media training for key CXC and other Chandra-related personnel who are likely to deal with the press. Standard strategies for preparing press releases and materials are being developed. Related materials (observations by other instruments, animations, and explanatory diagrams) are being prepared for early Chandra observations which might yield press-worthy results.
Tucker noted that NASA has the "first right of refusal" for press releases associated with any Chandra data. A note discussing this policy and press relations will be sent out to all Chandra observers. The committee suggested that the legalistic aspects of this policy be be stated but de-emphasized, and that the opportunity associated with NASA's access to and knowledge of the press be emphasized.
There was a discussion of the conflicts between the Chandra policy of releasing all science verification and calibration data to the public archive immediately, and the need to embargo press materials before a press release.
Kathy Lestition (CXC) described the current efforts and plans for
educational outreach associated with Chandra. The Chandra webpage
has been very popular, and has won a number of awards. Materials
are being prepared for use by school teachers. An "ask a Chandra
scientist" interface exists for questions from the public, and common
or generally interesting questions and answers will in added to a
FAQ page. The committee was happy with these efforts. They expressed
an overall concern, not associated with Chandra both rather with the
overall EOP programs within NASA and NSF, that there seemed to a lot
of duplication of effort in such programs. Because the EOP programs
tended to be associated with individual observational programs on
different observatories, there is almost no coordination of efforts.
Harvey Tananbaum and the committee discussed turnover in the membership. Because of a desire by the CXC to get early input into important decisions being made long before the launch of Chandra which would eventually affect users, the committee started work in 1993. Of the current members, six are from this original group. In addition, one of the original members who resigned last year has not yet been replaced. Several members of the committee expressed a desire to maintain their association until the launch of Chandra. However, everyone felt that it was important to have a regular turnover in the committee membership once Chandra was an operating observatory. The nominal number of members of the committee (without the ex-officio members) is 12. The suggestion was that, in steady state, 4 members be replaced each year. and that the normal term of members be 3 years.
In the past, the committee has generally met only once a year. An exception was in 1997, when a second short meeting was held to approve the rules for the Cycle 1 NRA and proposal review. The committee agreed to meet approximately 6 months and also one year after launch. After the meeting six months after launch, at least three of the original members would be replaced with new members. The frequency of future meeting would probably be once a year, but this may depend on experience.