FUSE Observer's Advisory Committee
Meeting Minutes No. 2
15 November 1999

The NASA Project Scientist for FUSE, George Sonneborn of the Goddard Space Flight Center, convened the second FUSE Observer's Advisory Committee (FOAC) meeting on Monday, 15 November 1999, at Johns Hopkins University (JHU) in Baltimore, Maryland. Attendees at the meeting included:

Committee members : Bregman (via telephone conference), Chu, Deharveng, Federman, Guinan, Harper, Hutchings, Koratkar, Raymond, Wannier

NASA : Hasan, Sonneborn

JHU : Andersson, Blair, Moos, Murphy, Oegerle, Weaver

Each boldface heading below represents an area of significant discussion at the second FOAC meeting. However, the order of the discussion below reflects logical groupings rather than the chronological order of the discussion at the meeting.

Mission Status and FUSE Science Highlights

George Sonneborn presented a high-level overview of the current status of FUSE. He emphasized the importance of getting out the word that FUSE is in orbit, is operating successfully, and is already producing outstanding scientific results. He showed several impressive examples of FUSE spectra taken through the large aperture (LWRS), including those of hot stars in the LMC and SMC, an AGN, and a White Dwarf. These results, and others, will be presented in a special FUSE poster session at the upcoming AAS Meeting. There are also plans for a Press Release of some examples of FUSE spectra in about a week or so, followed by a Press Conference at the AAS Meeting discussing some early FUSE scientific results.

The FOAC stressed the importance of providing information on FUSE's performance to the user community ASAP, and the Project agreed to post some relevant information on the FUSE web site in the very near future. The Project will also consider using various email exploders (e.g., that of the AAS, the NASA Office of Space Science, etc., in addition to the standard mailing list for the FUSE Newsletter) to alert people to the newly-posted information.

For a variety of reasons (detailed in FUSE Newsletter No. 9 ), the FUSE operations schedule is behind the pre-launch timeline by ~2 months. However, progress is being made, and the Project is planning to enter normal operations phase by mid-December, and this will include the routine scheduling of GI programs.

Although normal operations will begin in December, there will be significant limitations in FUSE's performance during the next several months. Alignment problems may restrict nominal operations mainly to the LWRS aperture and may sometimes (~15% of the time) result in an inability to obtain data in the SiC channels (i.e., at the shortest wavelengths). The spectral resolving power will probably be ~15,000 (~20 km/sec), which is lower than the value of ~25,000 that we eventually hope to achieve.

These limitations impact what programs can be attempted, and there was considerable discussion about the strategy for identifying suitable programs. Sonneborn reported that ~40% of the GI programs requested a spectral resolving power greater than 20,000, but several people pointed out that many observers would probably be willing to relax their constraints in exchange for executing their observations sooner rather than much later in Cycle 1. Owing to the excellent pointing stability of FUSE, observations in the LWRS aperture should have similar resolution to those taken through the MDRS aperture. Except for observations in crowded fields, or for spectral regions near strong airglow lines (e.g., H Ly-beta), there is not much advantage in using the MDRS aperture instead of the LWRS aperture. Using the LWRS aperture also means that there is a higher likelihood of achieving adequate alignment of all four channels (i.e., obtaining data in both the LiF and SiC channels). These facts, together with the risk of deferring observations for up to a year or more, suggest that many GIs will want to execute their observations now, or in the near future, rather than waiting an uncertain length of time until instrument operations and performance are optimized.

It was eventually decided that the FUSE Project would design a short and simple user survey that would help to identify those programs that are candidates for early execution. The survey will be distributed within the next week or so. Some background material will be included with the survey, so that GIs can understand the various tradeoffs that must be considered in answering the survey questions.

Observatory Operations and Instrument Performance

Bill Oegerle presented an overview of FUSE operations and performance since launch. The good news includes:

  1. Both detectors are operational and have very low background levels.
  2. The spectrograph has very low scattered light levels, which coupled with the low detector background, means that it will be relatively easy to establish the zero flux level.
  3. The instrument sensitivity is consistent with the pre-launch predictions.
  4. The guidance system operates within specifications with excellent pointing stability.
  5. The telescope PSF appears to be small, although detailed focussing has not yet been performed.
  6. The Focal Plane Assembly (FPA) mechanisms (containing the apertures) are operational and exceed design specifications.
  7. The data acquisition system works as designed.
  8. All mirror actuators are operational.
  9. All mission operations are being conducted successfully from JHU.
  10. There have been no spacecraft hardware failures.

Unfortunately, there were also some unexpected "features" of FUSE that have caused problems of varying severity, ranging from "annoyances" to "serious". They include:

  1. Channel alignment instability: FUSE has four separate telescopes that are supposed to be co-aligned for simultaneous use. However, the FUSE Project has discovered that the channel alignment does not seem to be stable when operating at high beta angles ("beta" is the angle between the anti-solar direction and the telescope boresight; "high" beta angle is ~90 degrees in this context), with misalignments of ~5-10 arcsecs being common. There is also some orbital variation in the channel alignment at the ~1 arcsec level. The exact cause of the alignment instability is not yet fully understood, and a "Tiger Team" has been formed to investigate the problem in detail. Gathering information on the channel alignment instability has been very time-consuming, and the instability itself seriously affects the ability to perform observations in the MDRS and HIRS apertures. The FUSE Project decided to spend a significant fraction of the observing time during November characterizing the problem and hopes to use that data to fix the problem or devise an acceptable workaround.
  2. Spectral drifts: The x-position ("x" is the dispersion direction) of the Lyman-beta airglow line moves by ~15 pixels during 1 orbit in the LiF1a and LiF2b channels. The motion is ~7-8 pixels in the SiC channels. The y-position of the airglow line moves by ~10 pixels in LiF2b and by ~5 pixels in the other channels. Preliminary analysis indicates that the x-drift can be reduced to ~5 pixels by adjusting the spectra during pipeline processing, but further work is needed. (A resolving power of 30,000 at 1000 A corresponds to a feature having a FWHM of ~5 x-pixels.)
  3. FUSE spectra sometimes have a dark "streak" in the optically active region, whose origin is unknown. In the regions of the spectrum affected by the streak, the photometric accuracy may be uncertain by up to ~50%. It is hoped that the streak will disappear after the telescopes are focussed in early December, although we do not know for sure that the streak is related to the focus.
  4. Single Event Upsets (SEUs) in the detector memory; these have been "major nuisances" because they sometimes require shutting down the detector HV, and the subsequent ramp-up is very slow (e.g., up to 1 day). The Project is developing new software and procedures that should significantly less the impacts of SEUs on observatory operations, and we hope to implement these changes during the next month.
  5. Occasionally we see "bursts" of large count rates (as many as a thousand times larger than the dark rate) ranging in duration from a few seconds to a few minutes. When using time-tagged mode, the corrupted data can easily be excised. When using histogram mode, one might have to throw out an entire observation. (Typically, there would be ~4 observations per visibility window.) For objects with fairly high count rates (e.g., hot stars), one would probably not even notice the effect of these bursts in the extracted spectrum. The cause of the bursts has not yet been determined.
  6. Until late September, a Fine Error Sensor (FES) software problem caused the FES to lock-up frequently and fail its Guide Star (GS) acquisition. This problem was completely eliminated with a new FES s/w load.
  7. The stray light levels in the FES are a couple of orders of magnitude larger than predicted. This affects FUSE's ability to perform GS acquisition during the daytime portion of the orbit. Nighttime GS acquisitions work just fine, and GSs acquired during the night can be tracked through the day. Thus, the stray light level has been a relatively minor nuisance.
  8. The second FES (the "b" side) is not yet in focus. This currently has essentially no impact as the other FES is used for all acquisitions. If a problem develops with FESA, then the FUSE Project will invest more effort to optimize the performance of FESB.
  9. The primary FUSE telemetry ground station at the University of Puerto Rico at Mayaguez (UPRM) has experienced intermittent failures that have seriously disrupted FUSE operations. A "Tiger Team" has been formed to investigate this problem because long-term stability of this ground station is very important for the success of the FUSE mission. Fortunately, the UPRM station has been performing reliably over the past month.

Data Processing

Ed Murphy presented a status report on the FUSE Pipeline, which is now up and running. Owing to the lack of various reference files (e.g., on-orbit flats, data on the astigmatism, etc.), the current pipeline processing does not include several steps that will eventually be implemented, after the relevant data are obtained.

New modules are being developed to enable the removal of bursts, to compensate for spectral drifts, and to remove stray light in the spectrograph (the latter being a minor effect for most programs). The spectral drift correction module may be implemented within the next month, but implementation of the others will take at least several more months.

A FUSE Data Handbook will be made available to GIs as soon as GI observations begin, which is expected to be sometime in December. The Handbook will describe in detail the content of FUSE data files and will also be the primary source of information on FUSE data analysis. It is expected that the information in the Data Handbook will be updated on relatively short timescales, at least during the first few months of GI observations, so a web-based format will be adopted. The FUSE Project also expects to make various other items available from the web that might help GIs during their data analysis (e.g., IDL routines provided on a as is basis, relevant internal reports, etc.).

GIs should be aware of the fact that FUSE observing programs can generate large amounts of data. A single time-tag exposure may produce up to ~49 MB of data, and a 20 exposure (~40 ksec of observing time, or ~20 orbits) time-tag observation may produce nearly 1 GB of data. A single orbit of histogram observations (which typically means 4 histogram exposures) will produce ~36 MB of data. GIs should also be aware that manipulating FUSE spectral images in computer memory requires large amounts of RAM (at least 256 MB), as the images are 16,384 x 1024 pixels. Thus, observers should buy lots of disk space and memory if they expect to analyze FUSE data on a routine basis.

FUSE observers will retrieve their data products from the Multi-Mission Archive at the Space Telescope Science Institute (MAST) in a manner similar to that used by HST observers. FUSE PIs must first register with the STScI before they can access any FUSE data, If they so desire, FUSE PIs can also designate that other people should also have access to their proprietary FUSE data (by providing a list to the STScI Archive Branch).

Miscellaneous Items

The FUSE Project expects to provide information on the scheduling of FUSE programs on its web site, so that observers will have some idea of when programs might execute. However, GIs must recognize that the dynamic nature of FUSE scheduling means that the actual schedule may depart radically from any particular posted schedule. The FUSE Project expects to notify GIs of the availability of data after the GI observations have executed.

The FUSE GI budget remains healthy (i.e., there have been no cuts since the last update), and Sonneborn expects that NASA will start distributing Cycle 1 money to GIs in January 2000.

The FUSE Project intends to organize a data workshop for FUSE GIs in February or March 2000, shortly after the NRA is released for Cycle 2. The timing of the workshop will be chosen so as to maximize the benefit for both Cycle 1 GIs who are analyzing their data and for potential Cycle 2 proposers.

The current schedule for Cycle 2 activities is:

  1. Release NRA : February 2000
  2. Proposals due : May 2000
  3. Proposal review : June 2000
  4. Selections announced : July 2000
  5. Phase 2 proposals due : September 2000
  6. Start Cycle 2 : December 2000

The next FOAC meeting will be held in the spring of 2000, but no firm date was set.

Go to the top of this webpage