CXC response to CUC Recommendations following 2008 Oct CUC Meeting
> 1. Readjustment of LP/VLP time. The CUC was surprised that no VLPs
> were approved in Cycle 10, and that LPs were more oversubscribed
> than other GO proposals in general. We discussed whether any change
> in the next CfP was warranted by this, and concluded that in the
> light of "small number statistics" and the inevitable year-to-year
> variation in proposal reviews, we should wait at least another cycle
> before recommending any change.
>
> RECOMMENDATION: No change for the time being, but possibly revisit
> this question in a year or two.
CDO will as usual tally statistics and trends for presentation
to the CUC at the next Fall meeting.
> 2. Budgets for Joint Time Cost Proposals. Proposers to other GO
> programs incorporating joint time with Chandra generally receive a
> lower "fair share" Chandra budget than they would if they were awarded
> the time in the Chandra peer review. While the amount of work involved
> in analyzing the Chandra data is not reduced because of this
> arrangement, the CUC felt that some economies of scale exist with
> these joint time proposals and the present arrangement seems the
> most satisfactory procedure.
>
> RECOMMENDATION: We therefore recommend no change, other than that
> this arrangement should be made explicit in the CfP.
The CfP for Cycle 11 and henceforth highlight the following in Section
8.4 (Evaluation of Budgets):
For a project with a foreign science PI or for a joint proposal
where Chandra is not the primary facility, the fair share will be
reduced.
> 3. Einstein Fellows Institution Conflicts. Now that the fellowship
> program administered by the CXC is expanded from five to ten fellows
> per year, we discussed whether any amendment to the "one fellow per
> institution per year" rule would be appropriate. The CUC
> generally supports the aim of this rule to spread fellowship holders
> around to a variety of institutions. We discussed whether the CfA should be
> treated as two institutions - Harvard and SAO - as is apparently done by the
> other NASA Fellowship programs.
>
> RECOMMENDATION: We recommend a somewhat less forceful change in that
> all NASA fellowship programs should treat large institutions such as
> the Harvard-Smithsonian CfA similarly. We recommend that the CXC
> discuss this with the institutions managing the Hubble and the Sagan
> Fellowships and work with them to formulate a common policy.
CDO has been in touch with the administrators of the Hubble and
Sagan Fellowships on the question of host institutions. We are putting
together a memo which will be circulated within the CXC as well
to the other programs.
> 4. Senior Review. We congratulate the CXC on achieving 2nd place in
> the senior review, very close in fact to the 1st placed mission
> (Swift).
>
> RECOMMENDATION: The CXC believes that a high ranking in such reviews
> is vital to the health of the mission. The CUC stands ready to
> assist in future reviews by submitting material, acting as "red team
> reviewers", or in any other fashion needed.
The CXC and CDO in particular, greatly appreciate the support of the
CUC.
> 5. Gratings. The CUC are very excited about the developments with
> TGCat. Most of our discussion centered around threads we would like
> to see. The threads presented during the meeting appear to be at the
> "graduate level", i.e. treating fairly complex problems
> that will be of interest mainly to users already familiar with
> grating observations.
Until Verification & Validation ("V&V") is complete, TGCat is
still in beta, though the link is public and we encourage feedback
(http://spacebase-alpha.mit.edu/tgcat/). We expect official
release by mid-February at the latest to be available for proposal
planning purposes. Before that release we plan to include an
interface via source class or object name. We have added Chandra
Source Catalog (CSC) ACIS bands to our flux table as suggested by the
CUC, and we will provide cross-links with the CSC and XMM/Birds
catalog.
> RECOMMENDATION: As examples of "freshman" level threads, to help new
> users get started, the CUC suggests the following:
>
> How to find the flux in a single isolated line?
>
> How to find the line centroid (to determine Doppler shift or other
> shift) and width (for example to determine non-thermal mass
> motions)?
>
> These simple threads could of course also be expanded to the
> treatment of blended lines. We are thinking here of optically thin
> emission lines, but another complexity would be to consider also
> absorption spectra.
An analytic formula to estimate the flux in a line is addressed in the
Proposers' Observatory Guide (POG; section 8.5.4), at a pencil and
paper level, which is adequate for proposal planning. The procedure
for grating data at the "freshman level" is the same as for fitting
ACIS/CCD spectra (for example, fitting Fe K plus a continuum using
standard response files), but perhaps it is not obvious to new grating
users that this is so. For many users, the web and CIAO threads are
the first and only resource. We will take the information from the
POG and translate it into a thread. Finding the centroid, line width,
and confidence limits naturally follow when determining a flux; these
will be incorporated into the thread. This should provide a basic
introduction to users unfamiliar with gratings or with X-ray
spectroscopy, for both proposal planning and for analysis.
We will also build an introductory thread for determining when it is
appropriate to use the gratings. For instance, for determining flux
of moderately bright continuum sources, gratings may be superior for
some applications than CC-mode or off-axis pointing.
> Would it be possible to provide a simple tool like PSEXTRACT to get
> grating analysis started with new data? We note that such a
> capability is already provided in TGCat for archive data. Also a
> link to this project from the CXC portal page could be considered.
The TGCat end-to-end script is, already available for use on
(and has been used on) user data. One need only download
the user's data from the archive which creates directories with the
CXC standard file organization, and to have the required software
installed.
As soon as we do have a substantial volume of TGCat V&V'd we will
provide prominent links.
We will be pleased to provide updates at the next CUC meeting.
> 6. Calibration. The CUC is surprised that calibration issues such as
> the effective area of the HRMA are still not completely solved, but
> understands the difficulty of the problem. We are satisfied at the
> thoroughness and rigor being applied to the problem, but are left
> with the impression that the team working on this are waiting until
> the problem is completely resolved in all its detail before making
> any sort of fix available. Given the length of time this is evolving
> over, we feel this approach does not best serve the user community.
>
> The low energy filter contamination is a harder problem. The problem
> of the characterization of the contamination, and specifically its
> time dependence, was mentioned as item 6 in section B Calibration
> Requests of the calibration/ciao wishlist the CUC recently
> submitted, and appears not to have been acted upon.
>
> RECOMMENDATIONS: We strongly urge the calibration team to release
> (as part of the CALDB) the current best estimate to the corrected
> HRMA effective area that resolves the main issue as soon as
> possible. We understand that this approach may require further
> updates and some duplication of effort as more work is done on the
> problem, but feel the tradeoff for users is worthwhile.
>
> We also ask the calibration team to review the CIAO/Calibration
> wishlist and report back to the CUC about the ability of the team to
> address these issues. In particular, we emphasize the continued
> investigation of the low energy filter contamination as a matter of
> urgency. We are concerned that H column measurements are incorrect,
> and suggest that at the least, some estimate of the uncertainty in
> these results coming from the contamination should be made public.
>
> Through the electronic announcements and website, the community
> should be advised of the progress and given recommendations for how
> to proceed in the interim before the full calibration is complete.
> Completion of the calibration should be given very high priority.
1) An updated HRMA effective area model is currently scheduled
to be released to the public by Jan 15, 2009.
2) The AO-11 version of the POG has a section that discusses
the discrepancies between the current ACIS contamination model
and recent gratings data. This section contains estimates of the
over-predictions for various energy bands and spectral models, given
the underestimate in the optical depth at the C-K edge in the current
version of the ACIS contamination model. There is also a statement in
the POG explaining that values of NH derived from the spectral
analysis of ACIS data are uncertain by about 1.0e20 cm2 due to the
systematic uncertainties in the current version of the ACIS
contamination model.
We plan on releasing an improved version of the ACIS contamination
model by the spring of 2009. We will be pleased to provide updates at
the next CUC meeting.
> 7. Catalog. This has been a major push of CXC for past 7+ years, and
> the CUC is very pleased to see these efforts finally bear fruit. The
> Chandra source catalog should be an important part of the legacy of
> this mission.
>
> RECOMMENDATIONS: At a high level the committee recommends that the
> highest priority continue to be toward providing and improving the
> content of the catalog, and that lower priority be given to
> improving the GUI interface. There are a number of specific
> suggestions as well:
>
> (1) Provide a VO cone-search service. We encourage the catalog team
> to look at the cone search tools in HEASARC and IRSA's GATOR. If
> they choose not to adapt this software, they should be able to
> reproduce its basic functionality.
>
> (2) Implement a VOTable option for the returned catalogs
>
> (3) There are so many different parameters in the catalog that a
> novice user will have a hard time figuring out which ones to
> request/use. Provide some tutorials, and perhaps also some 'canned'
> collections of parameters that could be selected for vanilla
> searches.
>
> (4) Make it clearer that the current release of the catalog is
> primarily for compact sources. The statement on the main CSC homepage
> that "The Chandra Source Catalog (CSC) is the definitive catalog of
> all X-ray sources detected by the Chandra X-Ray Observatory" is
> misleading.
>
> (5) Seek feedback on the catalog from community (beyond the CUC),
> including those who advised in its development. In the last CUC
> meeting report it was suggested that the group of proposers who
> submitted Cycle-10 catalog-related archive projects might form
> a useful beta testing team.
>
> (6) We urge the catalog team to develop a method/protocol to
> reference the Chandra source catalog from external systems not going
> through the Chandra proprietary catalog interface. In this manner,
> for example, XMM or Suzaku data centers in ESA or Japan can
> cross-reference the Chandra catalog sources, and popular
> multi-mission tools like "Aladin" may display the Chandra sources.
> It is recommended that an existing framework such as VizieR be
> adopted rather than developing a new protocol.
>
> (7) Will there be a refereed paper describing the catalog? Currently
> there are conference proceedings (which is fine), but if people
> start using the catalog,a proper reference should be available.
The Chandra Source Catalog team greatly appreciates the comments and
recommendations made by the Chandra Users' Committee. The team agrees
that improving the content of the catalog, and providing that content to
users with excellent supporting documentation, continue to be the
highest priorities moving forward.
We note that the specific functional suggestions of the Committee are
fully consistent with our existing plans for the future development, and
address them individually below.
1. VO cone-search service. As described in the Catalog presentation
to the October 2008 meeting of the CUC, a VO standard cone-search
is a part of our high priority plans, and we plan to provide this
capability in the short term.
2. Implement a VOTable option. The addition of VOTable output is also
part of our high priority plans, and this capability will also be
provided in the short term. We will initially support a TABLEDATA
serialization of the data content, but also plan to support a FITS
serialization.
3. Tutorials and 'canned' parameters. We are continually improving
the catalog user documentation, and are currently developing a set
of catalog threads (similar to the CIAO threads) that will provide
tutorial examples for the catalog user.
The next version of the CSCview GUI (that will accompany the
official first release of the catalog) will include a limited number
of 'canned' sets of parameters that can be selected for common
searches. In addition, the user will be able to use the same
mechanism to define their own parameter sets for common searches.
4. Clarify content of the first release. The catalog team has revised
the wording on the CSC homepage to clearly indicate the content
of the first catalog release.
5. Seek feedback from the community. The catalog team agrees that
feedback from the community is useful and appropriate for guiding
further development of the catalog. Once release 1 is complete, we
will work to solicit feedback from a broad audience that includes
the general multi-wavelength community as well as experienced
Chandra users.
In response to the second part of the recommendation, we did
distribute a beta version of the catalog contents to the PIs of
Cycle 10 catalog-related archive proposals in response to the
previous recommendation of the CUC. We did not receive responses
from any members of that group, so we will pursue alternate
approaches to garner community input.
6. Access from external systems. The CSC contents are currently
accessible directly via a web URL (using eg, cURL or wget) without
going through the CSCview GUI interface, and we will continue to
support this capability. In the short term we also plan to
provide VizieR TSV (tab separated value) formatted output for the
tabular data, which will be readable by VizieR-compatible
applications.
7. Catalog paper. The main catalog paper (Evans, I. N. et al.) is
currently in an advanced stage of preparation, and should be
submitted (most likely to AJ) within the next two months. This
paper will describe the catalog, contents, and algorithms. A
followup paper (Primini, F. A. et al.) will describe in detail
the statistical characterization of the catalog sources.
> 8. Revised Chandra Website Design. The CUC reviewed the new format
> for the Chandra web page, and generally approved it.
>
> RECOMMENDATIONS: The various issues we noticed were:
>
> There is no link to the calibration pages on the front page.
>
> The picture of Einstein (for the fellowship program) looks a little
> out of place. We would have expected to see a picture of Chandra
> (either the satellite or the astronomer) taking pride of place
> here.
>
> There is no link to instruments and calibration on left hand menu.
>
> Some of the topics on menu should be reordered/renamed. For example,
> "communications" could become "news", and "data analysis" should
> probably be above "public outreach".
>
> If possible, we suggest retaining the style of the link map in the
> fixed header on the CXC pages, since we believe those are regularly
> used. We also suggest using a web tool such as Google Analytics to
> discover in detail how users are currently accessing the CXC site.
>
> We also recommend live user testing. For example, how long does it
> take a random postdoc to find something in the new Chandra web
> pages?
>
> Finally, although we do not think the cxc.harvard.edu page should
> have the same look and feel of the chandra.harvard.edu page, we are
> impressed with parts of the design on that site and suggest
> consulting with the group responsible for that page as the new CXC
> site is developed.
We are working on the front pages and related links, and will
have a new version public before the next CUC meeting.
Einstein will instead appear closer to links to the Einstein
fellowship. Links for Instruments and Calibration will be added to
the front CXC page. We are accumulating data from Google Analytics on
who accesses our pages, from where, and when, which will be used as
input to our decisions about the structure and hierarchy of CXC links.
Elements of - and content from - the EPO site chandra.harvard.edu will
be shared with the new front CXC pages.
