A small targeted workshop on the subject of ``Complementary Optical Observations for AXAF and Other X-ray Missions" was held in San Diego on the morning of May 4, 1996, immediately following the meeting of the AAS High Energy Astrophysics Division. About 50 interested astronomers attended. The workshop, organized by N. Evans for the User Support Group of the ASC, was divided into two segments, the first on deep images for faint source identification, the second on multiwavelength campaigns. A summary of the presentations and some of the issues raised in discussions follows.
After a general introduction to the workshop, N. Evans (SAO) discussed source identification on deep AXAF images. If we use an estimated ACIS sensitivity and assume that the faint source population is dominated by AGNs, then providing optical identifications for 70% of the sources found in 100 ksec AXAF observations at high Galactic latitude would require optical images 3 magnitudes fainter than the Palomar Observatory Sky Survey (POSS II). By extrapolation from ROSAT experience, it is estimated that a large number of optical fields (per year) would have to be imaged to provide optical counterparts for high latitude AXAF observations. In the ensuing discussion, the question arose whether deep AXAF observations would be cluttered with red dwarfs; the consensus appears to be that this is unlikely at high Galactic latitude.
M. Watson (U. Leicester) reviewed the plans of the XMM Survey Science Center
regarding optical follow-up for the X-ray Multi-Mirror Mission, the second
cornerstone of ESA's Horizon 2000 program, scheduled for launch in August 1999.
XMM is expected to reach point sources as faint as 10 
erg cm 
s 
in
typical exposures of a few hours with the EPIC CCD cameras; at about 1000
sources deg ,
there might be 30,000 to 50,000 sources observed per year at
high Galactic latitudes. The XMM optical monitor should provide optical images
to better than 23rd magnitude for about 60% of the EPIC field.
L. Jones (GSFC) summarized experience obtained from deep surveys with ROSAT.
The fainter sources, with fluxes of about
10 erg cm s in the 0.5 - 2
keV band, require sensitivities to at least 25th magnitude in R for optical
identifications; this corresponds to about 1 hour exposures with a 4-meter
telescope. The burden is somewhat reduced by the use of modern, large-format
CCD cameras, such as the University of Hawaii 8kx8k mosaic camera, which covers a 30 arcmin field
in a single exposure. Discussion centered on the future role of improved multi-
object spectrographs and CCDs that yield some color as well as positional
information.
The AXAF Science Center has proposed to obtain CCD images, possibly three-color, of deep AXAF fields. J. Bechtold (U. Arizona and member of the AXAF Users' Committee) discussed issues raised by the Users' Committee in connection with this effort. In view of the Digital Sky Survey, the Sloan Survey, 2MASS, the VLA surveys, and others, is the ASC effort necessary, especially as most AXAF pointings are likely to be toward well-studied objects? Must all serendipitous sources be optically identified even if there are no specific scientific goals that demand it? There could be an important role for the ASC in arranging simultaneous optical observations in the case of transient and variable sources. Many questions remain about the technical details of the data that might be gathered -- choice of filters, field of view, magnitude limit, seeing -- as well as about the responsibility for assuring the quality of data. Many satellite observations will be obtained at solar elongation angles near 90 degrees (or less); it will be difficult to observe these targets simultaneously from dark sites on the ground. Similar concerns apply for targets observed at lunar bright time.
N. Reid (Caltech) provided an update on the progress of the second Palomar
Observatory Sky Survey (POSS II). It, together with the Second Epoch
Southern survey undertaken by the UK Schmidt telescope
(AAO and StScI), and the previous Schmidt surveys will provide coverage of
the whole sky
in at least two bands, to higher sensitivity
than the POSS I. Over 90% of the plates have been taken. The quality of the
plate material should permit astrometry to 0.25 arcsec and photometry to 0.2
mag. B. McLean continued with a report on the digitizing and
cataloging that is
in progress at STScI. The digitizing is proceeding at a rate of more than 1000
plates per year. The first generation scans, with 1.7 arcsec pixels, were used
to construct the first Guide Star Catalog, GSC-1, which is available by network
access or on CDROM. The second generation scans are using the new plate
materials, sampled with 1.0 arcsec pixels; they should be completed in 1998-99
and will be used to construct the GSC-II. It is expected that the GSC-II will
be completed in 1999 and will represent a vast improvement over the GSC-I:
completeness to V=18 rather than V=16 (at high Galactic latitude; the catalogs
are number density limited at low latitude); absolute positions
good to 
0.5
arcsec rather than 1.0 arcsec; magnitude errors
0.2 mag or better rather
than 0.4 mag.
In addition, GSC-II will have color and proper motion information
which is not available at all in GSC-I.
The Sloan Digital Sky Survey (SDSS) is a photometric and spectroscopic survey of
the northern Galactic cap and of a few southern regions. C. Stoughton
(Fermilab) summarized its status for the workshop. It is sponsored by the Sloan
Foundation, the NSF, and about eight participating institutions, and absorbs the
efforts of about 100 staff members. A 0.61-m automated monitor telescope will set
up a continuous, realtime photometric grid; the 2.5-m survey telescope, designed
to have a large field of view and a very flat focal plane, will perform both
imaging (in five bands, with 0.4 arcsec pixels and to 23rd magnitude in the red)
and spectroscopy (to 19th magnitude, with 1.5 Å resolution from 3900 to
9200 Å), using large array of CCDs as detectors. The
tentative schedule calls
for the telescopes to be instrumented and operating by the end of 1996, with
1997 devoted to tests and data to be gathered in 1998 to 2002. Reduction of the
expected 20 TBy of data must be highly automated.
The SDSS will release its data in two phases. Current plans are to
have a release of the first two years' of data in 2002.
The final catalogs, about 1 TBy in
size, will contain about 10 
imaged objects and 10 
spectra.
J. Vrtilek (CfA) started the second part of the workshop with a summary of the operational constraints on AXAF. Acquisition and guide stars will be available for the entire sky, and, as the only fundamental pointing constraint on AXAF is that the pointing direction must avoid the sun by at least 45 degrees, a large fraction of the sky should be available at most times. The high elliptical orbit of AXAF, with a period of 62 hours, of which about 50 hours should be above the adverse radiation environment and hence usable without interruption, is especially favorable for deep imaging and spectroscopy and for extended observations of rapidly-varying sources. To accommodate time-constrained observations, the scheduling system can place observations at fixed absolute times or within fixed windows, can group observations within a specified window, can fix the order of observations and the separations between them, and can place an observation in a specified phase range of a periodic window. There will be efficiency penalties if numerous time-constrained observations must be accommodated; the detailed tradeoff between number of constrained observations and observing efficiency is under study by the ASC.
B. Peterson (Ohio State) pointed out the fundamental changes in X-ray spectroscopy that result from high resolution of the kind that AXAF will offer. It will be readily possible to pick out spectral features and examine their variability in time, and that will allow X-ray reverberation mapping of active galactic nuclei -- i.e., the examination of the structure and velocity fields in the broad-line regions of AGN as determined by light travel-time effects in the response of the BLR to changes in the continuum. X-ray reverberation mapping should be able to map out the geometry and kinematics of the Fe K-alpha emitting region. An upcoming program that targets the Seyfert 1 galaxy NGC 7469 serves as an example of the magnitude and complexity of such a project: 30 days of nearly continuous monitoring with IUE and XTE, one full day of spectroscopy with HST, and several optical measurements per week at each of several observatories. Typically these individual monitoring projects involve over 100 people; a total of 232 individuals have participated in at least one of the four International AGN Watch (the largest monitoring consortium) projects that have reached publication. Substantial experience has led to the identification of a set of keys to success in such campaigns: secure broad participation; incorporate both scientific and technical expertise; focus on well-defined science objectives; define stand- alone science for each wavelength or observatory component; begin the program by securing the most critical component; establish a core of participating ground- based observers; and publish data quickly with minimal interpretation. Questions arose in discussion about the scheduling of multiple observatories and whether special procedures should be established; to this point the problem has been addressed by persistent work applied ad hoc in each case.
HD1405 is an RS CVn that is so highly variable that essentially continuous optical monitoring is required to understand certain other kinds of data, such as HST spectroscopy. The times of flaring are unpredictable, and flares may last from an hour to a day. A. Brown (U. Colorado) used this demanding source to illustrate what is needed for the conduct of effective multiwavelength observations: flexibility to respond to the detailed needs of observers; reliability in adhering to schedules once a commitment is made; cooperation and communication among observatories (which probably requires a ``gateman" -- a single point of contact - at each observatory); and a system of proposal deadlines that makes cooperation possible.
K. Peterson (STScI) presented suggestions for improving the chances of successful coordinated observations, based on experience with HST and other current observatories. It is helpful to choose targets that can be observed for long periods by one of the observatories (e.g., in or near the continuous viewing zones for HST or other low earth orbit observatories). It is important to furnish each observatory with full details of the observing requirements, to encourage the schedulers at various facilities to work together, and to answer questions quickly and participate actively in the coordination process. Finally, some concessions will probably have to be made and the observer should focus on the most necessary aspects of the observations and be prepared to compromise on others. The use of observatories in low earth orbit is challenging: there are issues of earth occultation and South Atlantic Anomaly avoidance, which can be especially awkward if two satellites in low earth orbit are required simultaneously. Nonetheless, such observations have been conducted successfully, for instance the combination of HST and XTE.
XTE, the X-ray Timing Explorer, is an example of a facility for which coordinated observations are a key activity. R. Corbet (GSFC) pointed out that XTE can point almost anywhere on the sky and is required to be able to perform target of opportunity observations within 7 hours after a triggering event. Long term XTE scheduling is done with a version of SPIKE, and short term with a locally-written program NEEDLE. Mission planning is a substantial task that required a major effort for developing software and now occupies a full-time person. In general, coordinated observing has been very successful: there have been a number of targets of opportunity, monitoring campaigns, and observations coordinated with HST, IUE, ASCA, and ground-based optical and radio observatories.
P. Puxley (Gemini Project) summarized the scheduling plans for the Gemini telescopes, which are to see first light in 1998 and 2000 for the Mauna Kea and Cerro Pachon sites, respectively. The two telescopes will be scheduled together in 6-month semesters, with at least 50% of the time devoted to queue scheduling. The national partners can apportion a part of their classical or queue time for quick-response observations. Applications will be in two stages: in the first, the scientific case for the observations will be presented and the basic choices of instrument and queue/classical scheduling made; in the second, the observing requirements will be specified in detail, which is especially important for queue observations. Classical scheduling will not ensure the best observing conditions; those observations that require such conditions must be queue scheduled. There is a potential conflict between queue and time-critical observations; those observations that are truly time-critical may have to be classically scheduled, and could be executed by the PI or in service mode.
Jan Vrtilek