Minutes of HEASARC Users Group Meeting
2007 October 15-16
Monday October 15
8:30am: Coffee/snacks 9:00am: Session start: Introduction, Logistics, etc 9:10am: The Future of the HEASARC (Nick White) - NRC NASA Data Centers Report - Restructuring of Archives/GOFs at GSFC and Refocussing of the HEASARC to be Science-based rather than Wavelength-based - Upcoming (Spring/Summer 2008) HEASARC proposal to Senior Review 10:10am: NASA HQ Perspective (Jeff Hayes) 10:30am: Break 10:45am: Status of HEASARC Activities at GSFC: Archives Status (incl. GLAST and Suzaku), usage statistics, and Response to Action Items from Previous HUG Meeting (Steve Drake) 11:30am: AGILE Archive (Lorella Angelini) 11:45am: Status of HEASARC activities at SAO: CDA, CalDB, ds9, Level 3 products, etc. (Roger Brissenden) 12:15 (-12:45pm): EPO Activities: Cosmic Times, Student Hera, General Status (Jim Lochner) [Working Lunch (sandwiches/drinks) will be provided at noon] 12:45pm: Break 1:00pm: FITS standards, FTOOLS, Hera inc. heraSPEC (Bill Pence) 1:30pm: XSPEC status and Future Plans (Keith Arnaud) 1:45pm: Current and future data access at the HEASARC (Tom McGlynn): - Browse Developments - NVO Developments - SkyView - Other Possible Future Initiatives 2:15pm: General Discussion on the Strategy for HEASARC's 2008 Senior Review Proposal (Nick White) [We'll take a 15min break sometime in this discussion] 5:00pm: End of First Day
Tussday April 16
8:30am: Coffee/Snacks 9:00am: General Discussion on Strategy of HEASARC's 2008 Senior Review Proposal (continued) 10:30am: Break 10:45am: HUG Executive Session 11:30am: HUG Initial Response/Recommendations (Frits Paerels) 12:15pm: End of HUG Meeting
Minutes and RecommendationsHUG Members Present: Derek Fox, Eric Gotthelf (day 2), Julia Lee, Nancy Levenson, Frits Paerels (Chair), Paul Ray, Gordon Richards, Masao Sako, Rita Sambruna, Jon Miller. Sarah Gallagher was absent, with notification.
Note: specific recommendations and requests are printed in italics.
The HEASARC Users Group (HUG) met on October 15-16, 2007, at GSFC. It had been eighteen months since the last HUG meeting; a meeting in the spring of 2007 turned out to meet with severe schedule constraints and could not be held with a sufficient number of HUG members. Even though it had been more than a year since the last meeting, the meeting (and therefore this report) was dominated by discussions concerning the future of the HEASARC, and less than usual by discussion of technical development and related issues.
First, a few general remarks. We want to express our appreciation for the excellent job the HEASARC is doing. As usual, the members agreed that the HEASARC performs a vital service, and performs it extremely well. We also much appreciate its forward-looking attitude, which now extends to initiatives related to an expansion of its role, both along the traditional wavelength bands as well as in the direction of greater intellectual cohesion in astrophysical space research.
We note with pleasure the response to our previous recommendations and queries. One specific request (that the RXTE Good Xe data be made more easily accessible) was considered, but was put off pending recommendations by the RXTE Users Group, since the reformatting involved would exceed the available RXTE GOF manpower. The HUG understands and agrees with this approach.
We first heard a presentation by the director, Nick White, concerning the general status of the HEASARC, a reorganization of the Astrophysics Science Division at GSFC, and the (re-)institution of a theme-based structure to astrophysics at NASA. This raised questions of a general nature, and comments were invited by Nick, which we will discuss below after the discussion of the archive and software- specific presentations.
We then heard a presentation by Jeff Hayes, on organizational and funding issues related to astrophysics at NASA HQ.
With respect to the redesign of the ADP (Astrophysics Data Program), we recommend that it has an appropriate mix of grant sizes, enabling both specific, targeted investigations that typically run for one year and require on the order of $50 K to complete, as well as multi-year, comprehensive investigations. Members are concerned that the small projects may not receive adequate funding if their importance is not explicitly recognized by management and peer reviews. The members also endorse the removal of as many restrictions as possible on the proposed investigations (such as requirements that a project be based on data from at least N1 observatories from the list of N2 missions; or that it not use data from certain missions).
We then heard presentations on specific archives and software, and on the HEASARC-related work at SAO (see agenda). Our comments and recommendations follow:
The Suzaku archive has opened, but we are concerned about the status of the analysis software. Numerous important analysis tasks appear to be not yet ready for use. The HEASARC is aware of this, and the issue will be taken up by the Suzaku Users Group (meeting in December; HUG member Jon Miller is Chair of the SUG and takes note).
We support the HEASARC acquiring a copy of the AGILE archive. We did raise the question of funding for US AGILE Guest Investigators (there currently is none). While it is technically possible to propose for observations, waive proprietary data rights, and then apply to the ADP for analysis funding, this is an awkward situation. NASA should consider the establishment of a funded GO program.
The HUG is very enthusiastic about recent DS9 developments (one of which, the compatibility with Mac/OS X, was tested by several members immediately during Roger Brissenden's presentation, and found to be entirely satisfactory).
We were, again, very impressed with the E/PO program, and we would like to express our continuing strong support for this effort. We commend its leader, Jim Lochner, for his work.
We noted the various developments in the HEASOFT software, with one specific recommendation: that a Python interface to FITS be maintained. This could be as simple as maintaining and enhancing the existing interface, pyfits.py. Also, any such package should be compatible with the "numpy" numerical Python package, which seems to have become the standard for numerical python.
We noted the various developments in XSPEC. We have a number of specific recommendations and requests:
If possible, reduce the dependence of the spectral model software on the specific structure of the XSPEC environment, such that the model code can be used stand-alone for the construction of new spectral modeling code for general astrophysical purposes, not related to fitting with XSPEC, and alternative fitting procedures. A set of specific suggestions, as well as Keith Arnaud's initial response to these, is reproduced in the Appendix (an email from Julia Lee).
We strongly endorse Keith's plans for rewriting the XSPEC manual. Specifically, we think that the idea of having the documentation grow and expand by direct involvement of the user community, possibly through a Wiki, is very good. The HEASARC should investigate the practical feasibility of this idea, which would involve outside user access to NASA computers (and which should therefore perhaps materially be implemented differently).
We noted the 'Astrogravs' work (analysis of gravitational wave signals from BH-BH coalescence events) with interest. In this context: we support the HEASARC becoming the primary archive for any future gravitational wave data (and LIGO data, if it ever became public), in line with the position of the HEASARC as the primary archive for data related to NASA's 'Physics of the Cosmos' theme.
XSPEC should be checked (numerical accuracy) and possibly adapted for use over very large dynamic ranges in photon energy (such as involved, for instance, in simultaneous analysis of X-ray and GLAST data).
The development of a multi-core version of the XSPEC software receives strong support from the HUG.
We heard a presentation on Current and Future Data Access at the HEASARC (Tom McGlynn). Tom raised the question as to whether the standard Browse interface should be rewritten. The HUG reaction to this was cautious: we understand the reasons for having to rewrite it, but given the size of the effort, we would like to see a more developed proposal for the rewrite, and if and how it interacts or is driven by NVO development, with a timeline, before we comment in detail. However, given the timetable for the submission of the proposal to the Senior Review in Spring 2008, since Browse is at the heart of the HEASARC's existence, and it and similar pieces of the software will have to be maintained no matter how, continued funding of development for this software is justified. We agree with the HEASARC director that his work should be included explicitly in the proposal and it should be included as an item in the budget submitted to the Senior Review (either explicitly or implicitly). This recommendation effectively constitutes a cautious 'go ahead' endorsement.
Our comments on Nick's presentation on issues related to the current and future status of the HEASARC and the Senior Review, which opened the day, are below.
1. Change to a theme-based structure for NASA Astrophysics research
Jon Morse, the director of the Astrophysics Division in NASA’s Science Mission Directorate, has adopted three themes for space research in astrophysics: Exoplanets, 'Origins', and Physics of the Cosmos. The HEASARC will aim at becoming the primary archive site for the Physics of the Cosmos theme (and will be absorbing the microwave background archive, LAMBDA, for instance). The HUG is generally happy with the return of the theme-based structure; among other things, it will strengthen the position of the HEASARC which will no longer be based on a wavelength- or photon-energy derived criterion, but on an intellectually more compelling one. This will make its position more robust and flexible. We also endorse the associated overhaul of the org chart for the Astrophysics Science Division at GSFC, with the HEASARC assuming a central position (see presentation of Nick White).
Nick invited us to think about a possible alternative or parallel structure to the space astrophysics archives, based on Ideas, Issues, and Problems: 'What is the nature of Dark Energy?' 'Basic properties of Black Holes?', rather than the classical division by wavelength.
Our initial reaction is that this is an interesting idea; it would certainly establish links between disciplines, from which high energy astrophysics can only benefit (by drawing in researchers from other wavelength bands). However, we are unsure about how deep into the actual structure of the archive such a change should reach, and how it would be established. For one thing, it should always remain possible to use the archive as it is being used now: searchable by the classical descriptive categories such as instrument, object, position, wavelength, etc.
Derek Fox points out that such a 'conceptual' dimension to the archive could (or should ?) be developed organically by the community itself. He points to the example of the photography sharing website Flickr (http://www.flickr.com/), where the users establish links by posting tags. This 'tagging' could start (or trickle down) to the proposal level, where researchers can establish a characterization of their own invention (as opposed to a fixed menu of categories. Notice the very limited and restrictive nature of the lists of 'keywords' currently in use by journals).
The HUG members all agreed that trying to establish a 'conceptual' dimension to the archive by imposing a structure defined by committee or a single person who 'maintains' this structure is doomed to fail.
Another possible development would be the establishment of a 'My HEASARC' page, with user-defined links and environment, which could also be the starting point for setting up a user tagging system. We recommend that the HEASARC look into this possibility, and into the possible practical problems associated with having outside users modify NASA-maintained information.
2. The 2008 Senior Review
Nick invites the HUG members to help preparing the proposal for this review. He asks for specific examples of scientific successes based on use of the HEASARC, which HUG members promise they will think about. As a special challenge, we will try to think of a good example that emphasizes the software development work of the HEASARC (as opposed to a result just based on the archived data).
3. The HUG Compiled List of Strengths and Weaknesses of the HEASARC
Strengths (not in any specific order):
- The software is extensive and comprehensive; the archive is very homogeneous.
- The HEASARC has excellent responsiveness (both of the helpdesk type and of the development initiative type). The quality of service is very high.
- HEASARC indeed provides access to an entire discipline! (High Energy Astrophysics)
- The HEASARC very effectively ensures continuity: data from past missions remains accessible.
- Data retrieval is extremely efficient (retrieval times are measured in seconds or at most hours). The entire archive is effectively online.
- HEASARC's development of common software standards is extremely important (and has probably already led to significant savings).
- The E/PO program is extremely effective
- The support of peer reviews through the RPS system is very effective.
A summary statement might read as follows:
"The HEASARC is a model of what the Science Archives should look like: comprehensive, alive, and oriented towards the future".
- The HUG is concerned that the HEASARC cannot guarantee that the staff gets the 25% of their time to do research. This is vital to the health and effectiveness of HEASARC.
- 'Marketing' of sometimes very sophisticated features is not well developed. HEASARC should 'educate' its users more actively.
4. New Initiatives
Nick solicited items for new initiatives. A list of suggestions follows:
- A 'MyHEASARC' page
- An XSPEC Wiki
- Get the full IPN catalog from Kevin Hurley (current catalog only contains positional information on fully localized bursts, not on all bursts)
- It was suggested that the "fxbary" be extended so that it can correct TOA's for ground-based observeratories (optical and TeV gamma-ray, for example).
This was the last HUG meeting for Julia Lee and Nancy Levenson. We thank them for their service. This will also likely be the last meeting with Nick White as HEASARC director. The HUG expresses its gratitude to Nick for his dedicated service, and looks forward to working with his successor.
If needed, we will try to have a meeting around the time the proposal for the Senior Review has to be submitted (for example, two weeks prior to the submission date. NB: dates for the archives Senior Review: May 13/14/15, 2008, with the proposal due [WHEN? First Week April?] ), so we can comment on it. If that is not needed (or impossible to schedule), we will have our next meeting in about a year from now, October 2008.
HEASARC Home | Observatories | Archive | Calibration | Software | Tools | Students/Teachers/Public
Last modified: Friday, 02-Nov-2007 09:40:19 EDT