Space & Astronomy/Space and Astronomy (October 1993).iso/mac/TEXT/STSCI/MARCH91.TXT

Text File | 1993-07-11 | 129KB | 2,465 lines

-- Message from the Director's Office

 " Slowly, Hubble Begins To Do Science."  This headline and Mitchell
Waldrop' s article in Science (1 February 1991) caught the mood at the
annual meeting of the American Astronomical Society meeting in Philadelphia
and within the HST community. After a prolonged commissioning, made more
difficult by the spherical aberration in the telescope optics, we have
begun to accelerate the formal calibration and early science program.

Significant progress has been made in the precision calibration of the
focal plane and the checkout of autonomous target acquisitions in the two
spectrographs. These were major hurdles, which have now been passed, and
operation of these apertured instruments is now more straightforward. As a
result, we anticipate a dramatic increase in the number of high quality,
ultraviolet spectra, like those obtained for Alpha Tau and 3C273 (see p.
3). We expect that science with the High Speed Photometer (HSP) and the
Fine Guidance Sensors (FGS) will soon be increasing as well.

To quote a letter from our distinguished colleague, Garth Illingworth:  "
For HST as for all space missions, which have substantial risks of limited
lifetime or the potential for (further) degraded performance, it is
important to maximize the scientific return at each point throughout the
mission."  We at STScI subscribe completely to this tenet and we have been
endeavoring to maximize science returns with current HST capabilities as
well as studying and advocating corrective measures to restore at the
earliest possible date the originally planned performance of the
Observatory.

As a first step, the first year program of science observations will have
been defined by the time of the next newsletter. We have received
reevaluations of the General Observer programs approved in 1989. The
reconvened Time Allocation Committee has reviewed and recommended a
reprioritized set of programs to the Director. While a few programs were
withdrawn as infeasible with the current optical performance, the great
majority of investigators  responded to the new circumstances by modifying
their programs to achieve many of their original goals.

The revised Guaranteed Time Observer (GTO) programs for the next four to
five years will be submitted to the STScI by the end of March, permitting
the publication of a revised exposure catalog in May. These revisions are
crucial to recasting the HST science program to better match the current
observatory performance. Knowledge of this performance has been
substantially improved by the early science assessment observations. Some
early GO and GTO programs have already begun, and current plans have the
time for GO and GTO science observations surpassing the time for
calibrations beginning in May.

Further in the future, the first shuttle servicing mission promises to
correct most of the solar array pointing and optical problems which affect
the spacecraft. NASA and ESA are negotiating the delivery of an improved
set of solar arrays which will not be susceptible to the sunrise/sunset
disturbances.

Also, recent improvements in our understanding of the OTA (Optical
Telescope Assembly) performance and extensive studies of corrective optical
systems have benefitted the design of the WF/PC II corrective optics and
the optics in COSTAR (previously called SmartStar; see December 1990
Newsletter) for correcting the FOS, FOC, and GHRS. These studies have also
benefitted the design of the next generation of science instruments. As
might be expected, achieving diffraction-limited corrections to the
severely aberrated wavefront requires great care in the manufacturing and
collimation of these optics. Nevertheless, these corrective systems appear
well within the state of the art.  Following detailed reviews of the WF/PC
II and COSTAR optical designs and development plans, NASA will define the
content of the manifested December 1993 servicing mission.

As this brief summary indicates, the HST community has made significant
progress in the last three months and the pace of the calibration and
scientific program continues to increase. Near-term and long-term prospects
for major scientific discoveries with HST appear much brighter and better
defined than they have since last summer. The entire community should share
in the hope and satisfaction these developments bring.

- Pete Stockman and Riccardo Giacconi

-- High Resolution Spectroscopy of the Chromosphere of Alpha Tau

The K5 giant star Alpha Tau was observed in late November 1990 with the
Goddard High Resolution Spectrometer (GHRS) as part of the GHRS Instrument
Definition Team's Science Assessment program. The scientific goals were to
resolve lines of C II and Si II in order to estimate turbulence, measure
flow velocities, and improve density estimates.

These quantities can then be used to constrain a realistic chromospheric
model for this star whose chromosphere is very different than that of the
sun. For example, the relative velocities between different ions give an
idea of velocity changes with depth in the chromosphere.

The accompanying figure shows approximately one-tenth of the data obtained,
the 2337-2343 
 region from a 20-minute GHRS G270M/SSA exposure, compared
to an optimal co-addition (by Phil Judge using the difference-filtering
technique developed by Tom Ayres) of six IUE spectra totaling 570 minutes
of exposure time. The improvement in signal to noise in the GHRS spectrum
is clear.

The strong lines are similar in the two spectra, although the detailed
profiles differ and the narrower lines are not quite resolved by IUE. The "
grass"  at the base of the strong lines is all noise in the IUE data but
contains many real emission features in the GHRS spectrum. Twenty-five new
emission lines are detected and at least fifteen identified, including
those from transitions of Fe I, Fe II, Ni II, and Co II. The fluorescent
nature of the Co II 2330 
 line has been confirmed, validating a
suggestion made a number of years ago by Wing and Carpenter. Initial
density estimates obtained from C II line ratios are consistent with
previous values.

The GHRS data reveal a chromospheric turbulence of roughly 24 km/second.
There is little or no asymmetry in the C II, Si II, and Co II line
profiles, but there is a downflow of approximately 4 km/second in C II
lines and not in the Co II lines. C II line profiles have narrower cores
and broader wings than pure Gaussian, possibly indicating a variation of
turbulent velocity with depth.

The Principal Investigators on this project are Ken Carpenter, Richard
Robinson, and Dennis Ebbets. Further details of this observation will
appear in upcoming issues of the Astrophysical Journal Letters and main
journal.

-  Ken Carpenter and Doug Duncan

-- FOS Observations of 3C273

On 14-16 January 1991, we observed 3C273, the famous nearby quasar, in
order to characterize the ultraviolet spectroscopic capabilities of the
Faint Object Spectrograph (FOS) for quasar absorption- and emission-line
research.

These science verification (SV) observations were made through five
apertures: three circular apertures of diameters 0.3 arcseconds, 0.5
arcseconds, and 1.0 arcsecond; the 4.3 arcsecond square aperture; and the
0.25 arcsecond by 2.0 arcsecond slit. For each aperture, observations were
made through all three of the high resolution gratings: 130H, 190H, and
270H.

3C273 was also monitored with ultraviolet cameras on IUE to determine the
absolute flux calibration. The data are being reduced in collaboration with
the FOS team. A report of these results will be available through STScI and
will include measurements of the throughput and resolution obtained with
each instrumental configuration.

-  John Bahcall, George Hartig, Buell Jannuzi, and Donald Schneider



-- Comet Levy: The First HST Target of Opportunity

In May 1990 David Levy, an amateur astronomer from Tucson, Arizona,
discovered a new comet. Although subsequent analysis indicated that Comet
Levy (1990c) was not " new"  in the technical sense - the comet's energy is
slightly more negative than that of a true Oort cloud comet -  it soon
became apparent that its orbit provided an excellent geometry for
Earth-based observations. This raised hopes that Comet Levy might be the
brightest since Comet Halley' s apparition during 1985-1986, and indeed by
mid-August it became clear that Levy would be an outstanding candidate for
an extensive observational campaign.

During late August, observations of Levy with IUE and from ground-based
telescopes demonstrated that the comet displayed significant temporal
variation with an apparent period of about 19 hours. Immediately questions
were raised concerning the source of the temporal activity.

Was the surface of Comet Levy dotted with pockets of highly volatile
material that ignited every time they were exposed to sunlight by the
rotation of the comet' s nucleus? Such a scenario would imply that Levy' s
nucleus was covered primarily with a non-volatile crust (only about 10% of
the surface of Comet Halley was active), and that the coma is produced by
relatively narrow " jets"  of gas and dust emanating from the volatile
pockets.

Alternatively, the surface of Comet Levy could be coated rather uniformly
with volatile material and the periodicity caused by very different cross
sections of a highly elongated nucleus being rotated into sunlight. In this
scenario, the production of gas and dust in the coma would presumably be
much more isotropic.

In any event, HST observations had the potential to provide important
insights into the nature of cometary nuclei. In late August, a team of GO
and GTO cometary observers submitted a Target of Opportunity proposal to
the STScI for Director's Discretionary time to study the inner coma
morphology of Comet Levy.

Since the HST observatory was still in a relatively immature stage with
regard to performing scientific investigations, and since the ability to
track moving targets was still being developed, the scope of the proposed
program was rather modest: only short " point and shoot"  images of the
comet with the Wide Field Camera (WFC) were requested. Exposure times were
restricted to less than four seconds in order to keep the trail of the
comet to less than one pixel. We requested two groups of images in order to
sample the coma at two different phases in its activity. Riccardo Giacconi
approved the program during the first week of September 1990, and thanks to
the efforts of many people within the HST ground system, the observations
were scheduled for 27 September 1990.

In the first HST image of Comet Levy, the nucleus landed within about three
arcseconds of its expected position, and the pointing for the first image
of the second group, taken about six hours later, was similarly good.
(Because of the passage of HST through the South Atlantic Anomaly, the two
groups could not be separated by exactly half the light curve period.) In
the 36 minutes it took to obtain the four images within each group,
however, the pointing drifted from what we had planned by up to 17
arcseconds. We hope to identify the source of this discrepancy eventually,
but it had no significant impact on the success of our scientific program.

At the time of the HST observations both the heliocentric and geocentric
distances of the comet were approximately      1 AU. The comet was imaged
with the WFC through both red and blue filters, which were selected to
isolate continuum emission peaking sharply at the nucleus. Each WFC pixel
is 0.1 arcseconds on a side, corresponding to approximately 78 kilometers
at the comet. (The comet' s nucleus is probably about 10-15 kilometers in
diameter.)

The longest exposures through the red filter have sufficient signal to
noise that image deconvolution can be employed to recover the full spatial
resolution of HST. The images show a fan-shaped inner coma in which the
sunward-facing hemisphere is clearly brighter than the tailward hemisphere,
consistent with volatile sublimation occurring primarily on the dayside of
the nucleus. Spatial brightness profiles perpendicular to the sun-comet
line are very symmetric about the nucleus and follow a 1/r brightness
distribution (where r is the projected distance to the nucleus) to within
100 kilometers of the nucleus.Thedeconvolved image shows no obvious
evidence for a dust jet, but generally such features are only revealed with
more sophisticated image processing techniques, which are currently being
pursued.

To investigate the temporal variability of Comet Levy quantitatively, we
have subtracted images taken during the first group of observations from
their counterparts in the second group. The resulting difference images
show a shell of dust propagating through the coma, which could be
responsible for the temporal variability in the observed light curve. The
velocity of the dust in the shell (more properly described as a
hemispherical arc, since the sunward side of the feature is much stronger
than the tailward side) is about 0.2 km/second.  The large angular extent
of the arc seems to favor isotropic emission over the entire sunward-
facing hemisphere of the nucleus, and we are in the process of determining
what limits can be placed on any jet activity from the nucleus.

-  Hal Weaver

-- GHRS Echelle Observations of Interstellar Gas Toward Xi Per

Xi Per (HD24912) is an O7.5 III star in the direction l=160.!4 and
b=-13.!1. With V=4.04 and E(B- V)=0.32, its distance is estimated from
spectroscopic parallax to be 700 parsecs. With a relatively large vsini of
216 km/second, Xi Per provides a relatively smooth continuum for studying
narrow interstellar absorption lines.

There are a number of different types of absorbing regions along the line
of sight to Xi Per. These include diffuse clouds, intercloud neutral and
ionized gas, and the H II region surrounding the star. In addition the
stellar wind of Xi Per may create a circumstellar bubble when it plows into
the surrounding H II region gas.

GHRS echelle mode observations of Xi Per were obtained in October 1990 and
January 1991. For the October observations, poor centering of the star in
the Small Science Aperture (SSA) produced spectra with lower than desired
signal to noise (about 20:1 was typically achieved). For the January
observations, spectra with signal to noise of up to 200:1 were obtained
with Echelle B and up to 30:1 with echelle A. The resulting spectra have a
resolution of ~4 km/second (FWHM) which is six times higher than the
resolution achieved by the Echelle mode of the International Ultraviolet
Explorer satellite.

In the combined data set, 72 interstellar absorption lines are detected.
Several Large Science Aperture spectra reveal that the blurry nature of the
HST image severely degrades the resolution of large aperture GHRS spectra.
The remainder of this article refers to SSA data.

The atomic species seen include C I,     C I*, C I**, C II, C II*, C IV, O
I, O I* (from the ISM and Earth atmosphere),      Mg I, Mg II, Si II, Si
IV, P II, S I, S II, S III, Cl I, Cr II, Fe II, Mn I, Ni II, Cu II, and
  Zn II. The important but very weak intersystem line of C II l2325 is
recorded with an equivalent width of approximately 1.6 m
 in a spectrum
having a signal to noise of 200:1. In addition, CO was measured. The
interstellar lines have complex multicomponent profiles. Sample spectra are
shown in the two figures.

The interstellar absorption line spectrum of Xi Per is complex. Optical
line data reveal strong components at heliocentric velocities from 6.2 to
15.8 km/second. The GHRS data reveal a rich component structure for neutral
gas phase absorption in the velocity range from 0 to 20 km/second. The
behavior of this absorption is seen to vary from species to species. For
example, the weak lines of O I, Fe II, Cr II, Mg II, P II have similar
absorption line shapes while those for Ni II and Ti II (from ground based
data) exhibit maximum absorption at somewhat larger velocities. The neutral
atoms   C I  and S I absorb more strongly at slightly smaller velocities.
These absorption profile differences will provide important information
about element abundances and physical conditions in the various diffuse
clouds toward Xi Per.

High velocity dispersion gas toward Xi Per is readily apparent in the
strongest UV lines of O I, C II, Si II, Mg II, and Fe II. In the case of O
I l1302 and C II l1335, the absorption in the core of the line is so strong
that Lorentzian damping wings have formed. Fitting line profiles to these
wings will permit an accurate estimate of the column densities of O I and C
II in the diffuse clouds toward Xi Per. The gas phase abundance of carbon
is crucial for interstellar grain theories.

The line profiles for the ionized gas (i.e. Al III, S III, Si IV and C IV )
are much broader than the profiles for the neutral gas phase species. The
ionized atoms also have average velocities blue shifted about 10 km/second
with respect to the diffuse cloud absorption features (see the Al III
profile in the second figure). These lines clearly are being formed in a
gas phase with very different kinematical properties than the neutral gas
features. An origin in the H II region surrounding Xi Per is likely,
although C IV may arise in a shocked interface between the wind of the star
and the surrounding gas.

The data reveal the great potential of the GHRS for high precision
diagnostic spectroscopy of the matter between the stars. The combination of
high spectral resolution, high signal to noise, and the very large number
of atomic and molecular species accessible at ultraviolet wavelengths means
the GHRS will provide interstellar observers with a powerful new tool for
studying physical conditions in interstellar space. Scientific results from
an analysis of the Xi Per data set will appear in an upcoming issue of
Astrophysical Journal Letters.

-  Blair D. Savage

-- A Detailed Look at Eta Carinae

Images of Eta Carinae obtained using the wide field mode of the WF/PC
reveal the detailed structure of this enigmatic object with new clarity.
The photo, a deconvolved combination of seven exposures with four different
exposure times, shows circumstellar material at a resolution approaching
0.1 arcsecond (roughly 441015 centimeters).

This material is thought to have been ejected from the star during an
outburst in the middle of the last century. The bright inner nebula (the "
homunculus" ) has an extremely well-defined edge and is clumpy down to the
resolution limit of the image. The outer material, including the ridge to
the southwest of the homunculus, has a knotty and filamentary structure,
consistent with the standard explanation that this is stellar ejecta
slamming into slower moving material which might be left over from a
previous episode of mass loss.

Perhaps the most remarkable aspect of the new data is the structure of the
extension to the northeast of the star. Here a limb-brightened
bow-shock-like structure is bisected by two linear features which point
straight back at the star. The morphology strongly suggests that this is an
active stellar jet from Eta Carinae which is driving a bow shock into the
material surrounding the star. To the south and east of the jet are a
number of uniformly spaced parallel linear features, presumably the result
of an instability or wave phenomenon in either the jet or the return flow
along the inside of the bow shock.

The structure near the star, in particular the annular depression and some
of the radially oriented structure to the southwest, is strongly influenced
by uncertainties in the deconvolution and the outer parts of point spread
function around this 6th magnitude star. Artifacts are also present along
the seams between the chips, and in the three chips where the PSF used for
deconvolution was inappropriate. These data provide an example of how, with
some effort, HST can be used effectively to do interesting morphological
work on bright objects.

 -  Jeff Hester

-- Determining the Distance to the LMC using the SN1987A Circumstellar Ring

The Faint Object Camera image of SN1987A and its circumstellar ring (see
December 1990 Newsletter) has now been deconvolved using the Lucy
algorithm. The resulting high resolution image clearly shows that the
brightness inside the ring is much lower than it is on the rim, meaning
that the emitting gas is truly confined to a ring. An inclination of ~43!
accounts for the apparent elliptical shape of the ring (Jakobsen et al.
1991, Astrophysical Journal Letters, in press).

The circumstellar ring around SN1987A is believed to have been ionized and
heated by the initial burst of UV radiation emitted in the first few hours
after the supernova explosion (Fransson et al. 1989, Astrophysical Journal,
336, 429). Therefore, the time delay between light received from the front
and back of the ring gives a measure of its absolute size.

Considering the tilted geometry as well as the recombination and cooling of
the ring gas, the intensity of an emission line from the ring is expected
to be zero for an initial period of time t0=(R/c)4(1-sini) where R is the
radius of the ring and i is its inclination. Then the intensity rises until
it reaches a maximum at tmax=(R/c)4(1+sini) and starts declining
afterwards.

Indeed, narrow emission lines of highly ionized species (N V, N IV, N III,
C III,      O III, He II) originating from the circumstellar ring were
first detected in the short wavelength IUE spectrum in late May 1987
(Wamsteker et al. 1987, IAU Circ. No. 4410), and their intensities have
been monitored regularly ever since. After reaching a maximum about fifteen
months after the explosion, the line intensities decreased for ten months,
then stabilized to a roughly constant value or a slower decline (Sonneborn
et al. 1990, IUE Toulouse Symposium).

Fitting the IUE data with a theoretical model indicates that the maximum
intensity occurred at 415125 days after the explosion (Panagia et al. 1991,
in preparation). For an adopted inclination of 43!, this translates into an
absolute diameter of (1.2710.08)41018 cm or 0.41510.025 pc.

By comparing this absolute diameter with the precise angular diameter
measured from the FOC image, 1.6610.03 arcseconds, we obtain a highly
accurate determination of the distance to the Large Magellanic Cloud, i.e.
51.513.1 kpc (Panagia et al. 1991). This represents an essential step in
the calibration of the cosmological distance scale.

-  Nino Panagia



-- What Does HST Do in its " Spare Time?"  The Non-Proprietary SnapShot
Survey

There are times when the HST is idle. Even with the sophisticated
scheduling algorithms employed by the STScI, there are periods that cannot
be filled by moving the telescope to one of the approved scientific targets
and executing the predetermined science program. The SnapShot Survey was
invented to fill in these gaps by obtaining important scientific and
engineering data.

In its current incarnation as a Director' s Discretionary Program through
Cycle 1, the SnapShot Survey is an imaging survey of bright quasars using
HST' s Planetary Camera (PC). Its purpose is to search for evidence of
gravitational lensing among known luminous quasars. The sharp core of HST'
s point spread function allows the detection of multiple images at
subarcsecond separations, which cannot be probed from the ground.
Theoretical lensing models predict that many gravitationally lensed quasars
have split images with these small separations. The intrinsically brightest
and relatively distant quasars are the best candidates for being lensed.

SnapShot observations involve short (currently two- or four-minute)
exposures, guided using only the spacecraft gyros, to avoid spending time
acquiring guide stars. This results in slightly trailed images. Targets are
distributed throughout the sky, so that only short slews are required to
move the telescope from any approved science target to a nearby SnapShot
target. SnapShot targets are scheduled only after all other programs have
been scheduled. The data are non-proprietary and can be obtained as part of
the archival research program (see p. 22).

To date, 38 SnapShots of 17 quasars have been taken. Unfortunately, in
about half the cases the targeted quasar was not in the field of view due
to errors in telescope pointing, indicating typical pointing errors of
order 20 arcseconds when on gyros. The length of the star trails obtained
in the gyro guiding mode are also several times longer than expected;
measured drift rates are characteristically five milliarcseconds/second.
The large gyro drift is believed due to the absence of corrections for
stellar aberration when HST pointing is under gyro control. The SnapShot
team (the authors plus N. Bahcall, O. Lahav, B. Yanny, and R. Romani) has
been monitoring and reporting on these and other engineering problems in
the Observatory' s performance to the STScI staff, who are working on
software solutions.

So far, of the six quasars that are definitely in the PC field of view,
none has shown evidence for gravitationally split images. From simulations
using the observed characteristics of the PC images (see photo), we have
found that secondary components would have been detected if present at
separations down to 0.1 arcseconds and brightness differences of 1-4
magnitudes, depending on the separation and the orientation of the
secondary component relative to the trailing direction. As more data are
gathered, this survey will both test and complement the results of similar
ground-based surveys having lower spatial resolution.

The Cycle 2 Call for Proposals solicits proposals for future SnapShot
programs. Additional information on the success of the current QSO program
and on the spacecraft performance will be published and available on STEIS
(Space Telescope Electronic Information System) prior to the deadline for
Cycle 2 proposals.

-  John Bahcall, Rodger Doxsey, Dan Maoz, and Donald Schneider



-- Saturn Movie Premieres at Philadelphia AAS Meeting

WF/PC observations of the great white spot on Saturn (see cover of December
1990 Newsletter) obtained in mid-November 1990 have been incorporated into
a video, shown at the Philadelphia AAS meeting, that covers one complete
rotation of the planet. The stormy, turbulent clouds extend throughout the
equatorial region of the planet, a relatively rare occurrence. The last
such storm was recorded in 1933.

 Images taken in the WF/PC red, blue, and green filters were combined to
produce approximately " true" color. The spatial resolution in the WF/PC
images corresponds to approximately 700 kilometers at the surface of
Saturn, the diameter of which is 120,000 kilometers.

The images used in the movie represent only about 15% of the data acquired
during the November observing session. The full body of data should reveal
new information about wind speeds in Saturn's atmosphere, the composition
and altitude of the clouds, and perhaps ultimately the cause of the great
storm.

-  Jim Westphal

-- The Art of Observing Planets with the Hubble Space Telescope

Many people have long regarded planetary (i.e., moving target) observations
as the most difficult and challenging aspect of the science mission of HST.
This brief article will not summarize the long history behind the
development of planetary capabilities with HST, but it should give a sense
of what it takes, today, to plan, schedule, and implement an observing
program involving a moving target.

The final plan for moving target capabilities for HST calls for the ability
to point and track any solar system object to within several tens of
milliarcseconds. This requires that the spacecraft know about parallax and
how that correction factor changes as HST orbits the earth. We also need to
be able to pre- program the motion of the field of view of the telescope on
the plane of the sky in such a way as to follow the complex motions of
planets or satellites or even surface features on these rotating bodies.

These are ambitious plans. Such capabilities are quite advanced when
compared to the usual abilities of ground-based telescopes. Although less
intricate than some of the Voyager commanding sequences, the software must
be designed, implemented, and executed with far less review and in less
time than is typical for flyby spacecraft. At least partly for this reason,
the planned planetary tracking capabilities of HST do not yet exist.

Then how does HST get images of Pluto, Saturn, Comet Levy, and Mars? The
answer is that we' ve worked hard at finding ways within current
capabilities to do some of the easier observations. Each of the moving
target observations done to date has probably required about an order of
magnitude more work on the part of STScI, CSC, and Goddard personnel to
ensure success compared to fixed target observations.

With a single exception, moving target observations have been limited to
imaging. The reasons should be obvious: the pointing tolerances are more
relaxed and the image itself indicates that the target was properly
acquired. A few cases will illustrate what happens behind the scenes before
the pictures are taken.

The first moving targets done were Saturn, Titan, and Comet Levy (1990c).
All of these objects were " ambushed,"  meaning that we point and hold the
telescope at the place where the target will be when the shutter opens,
letting the object come to us. Sound easy? Well, the position that we point
to depends on when we look, but when we look depends on where the target
is. Fortunately, this circular dependence can be broken by a good initial
guess for the targetJlocation. With an approximate position, the
observations can be scheduled; then from the approximate schedule, we
update the position. Usually this process converges after a single update
of the position.

This procedure takes care of pointing HST at the target. What about moving
the telescope with the target to avoid smearing the image? At the time of
the Comet Levy and Saturn observations, there was no such tracking method
so the exposure time was limited by the motion of the target on the sky.
For a fast moving target like a comet the exposure time is very short. For
Saturn, the exposures could be slightly longer. A typical rate for Saturn
is 0.001 arcseconds per second. Thus exposure times must be shorter than 43
seconds to prevent smearing the image by more than one PC pixel (0.043
arcseconds).

The pointing ambush mode was also used to make the Saturn movie
observations. The challenge in the Saturn movie was not in how the
observations were done, but in the speed with which we had to prepare for
the observations. Normally, an observing proposal is in hand well before
the time of observation, but in this case, the storm that erupted on Saturn
took everyone by surprise.

The initial request to observe the storm called for a limited number of
images in a few different filters at the time the bright cloud was expected
to be on the visible side of Saturn. This observing request came to STScI
shortly after the discovery of the cloud in September 1990, but given the
activities already scheduled on the telescope the best that could be done
was to plan observations four weeks later. This is not a terribly fast
turn-around time, but given all the constraints on the telescope it was
felt to be a reasonable compromise.

Preparing this program for execution in four weeks promised to be difficult
because of the many steps and large number of people required to put
together a valid observation. In retrospect, that work was easy compared to
what happened just after the first Saturn white cloud observations were
made.

In the finest sense of serendipitous science, the storm on Saturn turned
out to be so dynamic and complex that the minimal imaging sequences were
not enough to understand what was happening. Within 24 hours of completing
the first set of images, the observers were pleading for an in-depth look.
Life can get  interesting with no warning -  this second request happened
only a week before Saturn would be too close to the Sun to be observed with
HST and in the 100 days before it would emerge again the storm could be
gone. The decision was made to proceed and we had just over a day and a
half to put together the spacecraft commands and deliver them to Goddard.
Goddard would then have one day to finish the work and ship the commands up
to HST.

But wait, there' s more. If you ever have occasion to work with the WF/PC,
you will quickly find out how much data a single image contains. The
planned sequence of images, seven frames per HST orbit for two days,
generates so much data that it cannot be copied down to the ground
controllers before being overwritten by new images. Herculean efforts
resulted in special spacecraft instructions that would allow the readout
and storage of only one of the four CCDs that comprise a full frame. This
fix reduced the data flow to a level that could be collected on the ground.

There's still more. These observations took place during a Space Shuttle
mission, a military flight in fact. Whenever the Shuttle is in orbit, one
of the two TDRSS communication satellites is dedicated to the support of
the Shuttle mission. This means HST, which normally relies on both
satellites for transmission of the data to the ground, has to use whatever
time is left over. For the Saturn observations, we really needed both
satellites to keep up with the data flow. Fortunately, NASA was able to
shuffle things around in spite of the military shuttle mission and give HST
all the TDRSS  contact time needed to transmit the data.

The point and shoot procedure works well for WF/PC observations of most
bright solar system objects. What about the Faint Object Camera? This
instrument is just as sensitive if not more so than the WF/PC but is
limited by the photon counting rate. A decent image takes tens of minutes
to collect, regardless of the brightness of the object.

This was the challenge faced for the Pluto observations made with  the FOC
(see December 1990 Newsletter) since each image took fifteen minutes to
collect. Even with Pluto' s slow motion across the sky, the smear in the
image for stationary pointing would have been nearly 1 arcsecond, the same
size as the apparent separation of Pluto and its satellite, Charon!
Obviously, we had to track the motion of Pluto to obtain these pictures.

We tracked Pluto by instructing HST to execute a scan along the line
computed to best match the geocentric motion of Pluto. The onboard computer
would be responsible for converting this geocentric scan to the motion as
seen by HST from its position in orbit based on the distance to Pluto. Both
of these operations involved features of the telescope that had not yet
been tested in orbit but the decision was made to proceed anyway.

This method of using scans to track moving targets has worked so well that
we have now applied this technique to long-duration WF/PC imaging of Mars
in the UV where the planet is dark and the chip sensitivity is low. Still,
perhaps contrary to expectation, each new moving target observation seems
to get a little harder.

The WF/PC imaging for the Mars observations posed little additional
difficulty except when it became time to schedule them. The requirements
were to get three sets of images of Mars, spaced 120 degrees apart in
longitude within a single rotation of Mars. Sound easy? Well, the rotation
period of Mars is nearly the same as the Earth, and the South Atlantic
Anomaly (SAA) -  a region of high particle background - precludes HST
observations for roughly ten hours out of each day. Because this ten-hour
period drifts very slowly with respect to the visible longitudes on Mars,
there are regions of Mars that cannot be easily seen on any given day.

After many trials, we finally found times where we could squeeze in
observations just before the SAA crossing and then do the next set just
after clearing it. This experience made me aware of the harsh reality that
there are some planetary observations that will be impossible from HST no
matter how clever we are.

As if the scheduling difficulties of the Mars imaging were not enough, the
observing team wanted more. The other half of their program was to take
spectra of Mars, preferably scanning across the polar caps to get spectra
at different altitudes and locations. This was HST's first planetary
spectroscopic observation. The problem was that we were not confident about
the pointing at the sub-arcsecond level. To get around this problem we made
three passes across Mars with a dual beam aperture, offsetting the
telescope slightly between each north-south scan, figuring that one of the
scans would include what they wanted as long as the telescope could point
to within 8 arcseconds (which is not difficult). We also had to add the
motion of Mars into the desired north-south scan motion but fortunately
that was no more difficult than adding vectors.

The most recent challenge was to take long duration images of a satellite.
The object this time was Io, the inner most Galilean satellite. The goal
was to take fifteen-minute exposures while tracking the complex motion of
the satellite. In this case, a linear scan is a very crude approximation to
the motion of Io on the sky but for a fifteen-minute exposure the errors
can be held to less than 0.030 arcseconds.

For the Io images, the timing and the pointing and scan values for the
ambush were very important. Within a single ten-hour window centered on
eastern elongation, the apparent rate of motion changes by an order of
magnitude. The scan for the first image had to compensate for about 0.12
arcseconds of smear while for a later scan it was over 5 arcseconds of
motion. At the worst, the rate of motion changes by 4% per minute!

It is no wonder that design goals of HST include sophisticated nonlinear
tracking; most satellites are nearly inaccessible without the advanced
capabilities. In the meantime, however, we hope to find compromises that
allow planetary science programs to proceed unimpeded. Judging from the
results obtained to date, it is clear that HST will be a very capable and
useful tool for planetary studies, even if they are harder than average.

-  Marc W. Buie

-- Science Verification Continues

Most of the observing time on HST is currently devoted to the Science
Verification (SV) program for commissioning the science instruments. As of
1 March 1991, roughly 40% of this program will have been completed.

This is slightly behind the schedule jointly developed in November 1990 by
STScI and the HST Project Office at the Goddard Space Flight Center (GSFC).
The Commissioning Completion Plan calls for 46% completion of SV by March
1, with the fraction of time to be spent on SV running at 80% in December
1990, decreasing to 50% by May 1991, and to 10% by October 1991. During the
same period, the time spent on GTO and GO observations would increase from
10% in December, to 40% in May, and 80% in October.

The remaining fraction of time is allocated to engineering tests. At
present, the fraction of time spent on engineering observations is roughly
three times what was planned, hence we are slightly behind schedule for SV
and GTO/GO observations.

The observing efficiency is currently about 27%, considerably better than
was assumed in the commissioning plan (20%). However, the fraction of
observations that fail and need to be repeated (e.g., due to failure of
target acquisition) is currently about 25%, which is higher than the number
assumed in the plan (20%).

These observing efficiencies are for what is called " Spacecraft Time"  in
observing proposals, which starts with guide star acquisition and ends with
guide star release. This includes the time spent acquiring the guide stars
(fine lock, coarse track, or gyro control), acquiring the target, exposure
time, and other time used to set up or read out the instrument, but
excludes earth occultations, SAA passage, and other dead time.

Results from some of the science verification observations are reported
elsewhere in this newsletter.

-- Science Assessment and Early Release Observations

The Science Assessment Observations (SAO) and Early Release Observations
(ERO) were nearly completed by February 1991 (see  two previous Newsletters
for details of these programs). Only two HSP programs remain. The
information gained from these observations was distributed to General
Observers in preparation for their resubmission of Phase II information for
Cycle 1 observations and also to the Telescope Allocations Committee that
met at the end of February 1991.

The table on pages 11-12 provides an update on the completed SAO/ERO
observations, all of which are now non-proprietary. These data can be
requested as part of the archival research program (see p. 22).

-- GTO and GO Observations

The SV observations that were required before beginning most of the GTO and
GO observations have been completed for four of the instruments: the FOC,
FOS, GHRS, and WF/PC. Cycle 0 GTO observations have begun for these
instruments. In addition, a few Director' s Discretionary (DD) GO
observations and time critical Cycle 1 GO observations have also begun. See
the Tables on pages 11-12 for lists of completed, and currently planned
science observations.

Cycle 1 GO programs that were approved by the TAC in late February 1991
should begin showing up on the observing schedule in May. Those programs
with  no change or only minor changes will tend to show up first. However,
the fraction of time spent on GO observations will not become sizeable
until about July 1991 (i.e., the " beginning of Cycle 1" ).

Modifications to Cycle 1 GTO programs were due back to STScI by 28 February
1991. These observations will be worked into the schedule on about the same
time scale as the GO observations. The schedule of observations can be
found on STEIS (see p. 18).

-  Brad Whitmore

-- Observatory Status

While the Science Verification period and associated scientific program
have begun there remain a few activities of interest relating to overall
performance of the observatory.

Several problems in the ground system selection of reference stars and
operation of the Fixed-Head Star Trackers (FHSTs) have been fixed, with the
net result that the failure rate of FHST updates has dropped. There are
several other fixes in progress which should reduce the failure rate even
further.

HST suffered a failure of one of the attitude sensing gyroscopes on 3
December 1990. This failure was sensed by the onboard computers which
triggered a safe-mode entry. Following a short period of analysis this gyro
was shut down and one of the redundant gyros was activated and substituted
in the control logic. After a short recalibration, normal operations were
resumed on 8 December.

HST has six gyros and normally requires that four be operating, although it
is possible to operate with only three. Preliminary investigation by a
special NASA panel indicates that the failure was in the electronic rather
than mechanical portions of the gyro unit.

With this one exception the spacecraft support systems have continued to
perform as before. We have experienced no disruption to TDRSS
communications as a result of military operations.

The collimation of the telescope has been adjusted to compensate for an
initial erroneous setting of the secondary mirror, which has improved the
quality of the Fine Guidance Sensor (FGS) S-curves and the performance of
the FGS in fine lock; however, the guide stars continue to be restricted to
magnitudes brighter than 13. Analysis of FGS S-curves is in progress at
Hughes Danbury Optical Systems and STScI with the object of isolating
FGS-specific effects from the overall OTA (Optical Telescope Assembly)
effects, which in turn may allow the determination of an even better
position for the secondary mirror.

In December 1990 a new series of measurements was conducted to determine
the alignment of the FGSs. The updated alignments, implemented on 11
February 1991, improve the determination of positions of the Science
Instrument (SI) apertures and should increase considerably the reliability
with which the small aperture SIs acquire targets.

The jitter in HST pointing due to thermal transients in the Solar Arrays
continues to be a problem. As mentioned in the previous Newsletter the
first attempt to modify the control law to compensate for this was
unsuccessful and a second approach is currently being prepared. It is now
expected to be installed late this spring. We now know that both solar
panels contribute to the jitter. Replacement of the solar panels has been
proposed for the first servicing mission, and ESA is working on new panels
which will not exhibit this behavior.

-  Rodger Doxsey and Bob Milkey

-- HST Focus Update

Understanding the exact nature of the HST aberration is important both for
developing image restoration techniques and for building the replacement
science instruments. In the case of WF/PC II, the need is urgent with the
corrective optics being polished as this Newsletter goes to press.

Major contributions to evaluating the aberration have been made by the
Allen commission, JPL (the HST Aberration Recovery Program, or HARP), and
an Independent Optical Review Panel chaired by Dr. Duncan Moore of the
Institute of Optics in Rochester. The goal of these committees has been to
characterize the aberration by a variety of methods so that the corrective
optics can be prepared with exactly the right prescription to restore the
HST imaging performance.

This work is now complete, with a derived conic constant on the primary
mirror of -1.013510.0005. This accuracy is sufficient for the manufacturing
to proceed. Remarkably, the same result is obtained from three independent
means: measuring the field lens position error in the null lens that was
used in the construction of the primary mirror, examining existing ground
test interferograms of the null and primary mirror, and looking at the
imaging data obtained on-orbit.

The measurement from on-orbit observations was for several months an
uncomfortable 10% bigger than the ground test (" fossil" ) data. Several
teams of analysts reduced the data independently, including groups from
Hughes Danbury Optical Systems, JPL, and the STScI. The HST secondary
mirror was eliminated as a probable source of the discrepancy by
measurements on the essentially identical flight spare mirror at the
University of Arizona this past Christmas. The problem was finally traced
to the WF/PC optics which were found during ground testing to exhibit
spherical aberration of precisely the correct sign and magnitude to explain
the difference (although they met their specifications).

Thus all the various data and reduction methods point to the same result
and we are now reasonably confident that the aberration in the primary
mirror is fully understood. This enables us to proceed with confidence with
the fabrication of the second-generation instruments.

-  Chris Burrows

-- Status of the Goddard High Resolution Spectrograph

 The GHRS has achieved a number of important milestones over the last few
months. These include demonstrating a reliable capacity for Target
Acquisition, measurement of good sensitivity for several gratings, and
acquisition of outstanding scientific data for a number of early SAO
targets.

Experience with target acquisition shows that objects with good positions
are typically found within a few arcseconds of the initial HST pointing,
well within the capture range of the GHRS spiral search. The onboard spiral
search works well, centering targets automatically in the Large Science
Aperture (LSA). Centering accuracy with all acquisition mirrors except A1
has been very good; it is expected that further calibration using A1 will
increase its accuracy. The accuracy and reliability of placing targets into
the Small Science Aperture (SSA) is improving as well.

Stars of known ultraviolet flux have now been measured with all gratings
except G200M. The effects of loss of light due to the broad telescope point
spread function are close to expectations: a few tens of percent light loss
through the LSA, while the throughput of the SSA for a well-centered target
ranges from one- fifth to one-third of that of the LSA from 1200 
 to 3000

.

An exception is the short end of the wavelength range, where sensitivity is
significantly higher than pre- launch predictions. (Ground-based
calibrations at these wavelengths were known to be very uncertain.)

The most accurately calibrated grating so far is the low-resolution G140L.
Observations of two UV standard stars show excellent agreement. Significant
sensitivity extends below 1150 
, where the spectrograph is still one
third as sensitive as at 1200 
.

Combining the sensitivity measurements obtained so far with the GHRS and
FOS shows that the GHRS low-resolution (G140L, R=2000) grating has greater
sensitivity than the FOS in its highest resolution mode at wavelengths less
than 1500 
 (see next article).

Excellent scientific results have been obtained with the GHRS on targets in
the SAO program. These results include UV chromospheric line observations
of Alpha Tau (see p. 3), extensive echelle observations of the interstellar
medium line of sight to Xi Per (see p. 5), observations of a starburst knot
in NGC 1068, P-Cygni profiles in the massive star Melnick 42 in 30 Doradus,
and observations of the chemically peculiar star Chi Lupi. At press time,
extremely interesting observations of circumstellar material in the Beta
Pic system are underway.

-  Doug Duncan

-- Status of the Faint Object Spectrograph

Testing of the FOS has progressed well and several interesting scientific
investigations have been successfully performed as part of the SAO and ERO
programs. The FOS is performing as expected, with a few exceptions.

All mechanisms and calibration sources are operating nominally. The
detectors are both optimized to yield the expected resolution and are
operating well. The wavelength coverage of all dispersers has been
measured; only small shifts from the pre-launch spectral ranges are
present. An initial measurement of the spectral flat-field characteristics
has also been obtained for all dispersers; these show photocathode
granularity of less than 2%, although several " blemishes"  which require
corrections of order 10% are also present. The detector background rate is
higher than expected, with particle-induced events dominating. This
background may be reduced, for faint object observations, by judicious use
of the onboard burst noise rejection capability of the FOS.

The measured efficiency at wavelengths longer than about 1800 
 of both
the red and blue sides of the FOS matches the pre-flight estimates very
well after accounting for the losses at the entrance apertures due to the
degraded OTA PSF. However, the first blue side observations of standard
stars in mid-January 1991 showed that the far UV sensitivity begins to fall
below the predicted values at the shorter wavelengths. The sensitivity
drops to about half of that expected at Lyman alpha. The effect is seen
with both the G130H and G160L gratings on the blue side in data obtained
with three standard stars. The low efficiency is not due to OTA throughput
losses. There is no evidence that the FOS far UV throughput is degrading on
orbit; internal calibration lamp observations indicate that the efficiency
is stable.

Another problem discovered on orbit is cyclical image deflection due to the
geomagnetic field. This effect, ascribed to inadequate magnetic shielding
on the Digicon detectors, causes the images to wander with respect to the
diodes, with a period of roughly half that of the spacecraft orbit. The
images move by as much as 10.6 diodes (0.22 arcseconds) on the red
detector, and about one fourth of that amount on the blue detector.

The image motion will reduce spectrophotometric accuracy slightly, preclude
polarimetry on the red side, and impair the accuracy of some target
acquisitions. The resulting degradation of spectral resolution should be
recoverable if the spectra are read out at short intervals since we expect
to be able to model the image shifts with sufficient accuracy.

Efforts are underway to build software tools for this purpose and
additional observations to improve our understanding of the effect are now
scheduled for early March 1991. It appears possible to correct the problem
almost entirely by dithering the Digicon deflections in real time to
counteract the geomagnetic effect. This will require major software
changes, however, and will probably take about one year to implement.

The FOS has produced a number of high quality science spectra as part of
the SAO and ERO programs. These data show that the FOS can be used to probe
the nuclei of galaxies, to study the intergalactic medium and investigate
the astrophysics of active galactic nuclei by measuring the ultraviolet
through red spectra, of high and low redshift QSOs, and to obtain stellar
spectra even in crowded fields.

Virtually all of the science observations attempted were successful and
much experience has been accrued regarding effective target acquisition
strategy and efficient use of the spectrograph. SAO observations have been
used to confirm model calculations of the FOS efficiency and spectral
resolution as a function of aperture size and to test our strategy for
removing the geomagnetically induced smearing of the spectra.

A direct comparison of the count rates obtained with the FOS (G130H grating
and 4.3 arcsecond aperture) and the GHRS (G140L grating and LSA) from the
same standard spectrophotometric star shows that the GHRS is more sensitive
at wavelengths below 1500 
. Indeed, at Lyman alpha the GHRS is about five
times more efficient while covering about half the spectral range at nearly
twice the resolution.

In Cycle 1, GOs  may wish to take advantage of the better throughput by
switching from the FOS to the GHRS. Instrument scientists will work with
GOs to determine which spectrograph is best suited to the scientific aims
of the affected programs following the TAC review of the Cycle 1 program.

-  George Hartig

-- Fine Guidance Sensor Performance Update

The performance of the Fine Guidance Sensors is much better now that the
secondary mirror has been moved away from its extreme tilt. Extensive
testing in all three FGSs shows improvement in the fringe visibility
function, sometimes dramatically. We expect that a further de-tilt of the
secondary will bring about additional (though not so marked) improvements
in the FGS Transfer Function.

There are still clear, field dependent deformations of empirical Transfer
Functions relative to predicted Transfer Functions, however, which are not
understood. Hence making additional, rapid progress is not probable. These
deformations also imply that more is wrong in the OTA/FGS optical system
than primary mirror spherical aberration.

What does this mean for guidance? Now that the FGS Instrument Team is
performing the detailed analysis of Guide Star acquisitions, we are
confident that we can regain Fine Lock performance to V=14.0 magnitudes,
partly as a result of the improvements in the performance of the FGSs and
partly as a result of a deeper understanding of optimal FGS operation.

What does this mean for FGS science? Transfer Function astrometry, i.e.,
double stars, angular diameters, and color indices, should suffer
relatively little from uncertainties about the mirror. Such astrometry can
always be made relative, with little effort, to minimize other
complications. Also, much of the software intended for use on the
astrometric data reduction of Transfer Function curves has been debugged,
tested, and significantly expanded to deal with the engineering issues
discussed above. Hence, it is more robust conceptually and much better
performing functionally than we would have predicted at this stage of HST
operations.

For positional astrometry the situation is less promising. The
field-dependent aberrations will defeat all presently implemented efforts
to model the optical field angle distortions. Without milliarcsecond
success at the modeling effort, there will be no milliarcsecond positional
astrometry and perhaps only 5-10 milliarcsecond positions for proper motion
and parallax work.

The STScI Instrument Team has already taken steps to address these issues.
New techniques have been invented to specifically overcome these problems
and software to try them out is under development.

-  Larry Taff

-- Current Performance of the FOC

New observations and results have been obtained recently for several
aspects of Faint Object Camera performance. These include LED internal flat
fields, external flat fields, absolute detector sensitivity, and image
stability and plate scale.

Internal flat fields with LED illumination were obtained for both f/48 and
f/96 modes at wavelengths 4800 
, 5600  
, and 6600  
. From a
large-scale point of view, there are slight variations with respect to
similar flat fields obtained on the ground before launch, but no variation
has been observed between the different flat fields obtained in orbit.
Large-scale structure flat fields have been processed and installed in the
data processing pipeline.

External flat fields were obtained from observations of the Earth and the
Orion Nebula. Because of the very highly variable count rate of the Earth
flat fields, some problems with saturation were encountered. These
observations have been rescheduled in the case of f/96 mode for which
neutral density filters are available. The Orion Nebula was used as another
target, with observations made in both f/96 and f/48 modes.

An early estimate of the throughput was made from OV (Orbital Verification)
observations of two UV photometric standard stars. Over most of the
wavelength range the results are consistent to within a few percent with
the pre-launch expectation. However, below 2000 
 the sensitivity appears
to be about 20- 30% lower than the estimate published in the FOC handbook.

FOC images are stable on sub-pixel scales after warm-up. The long term
evolution shows no significant variation so far. The plate scale has been
determined to be 0.045110.0001 arcseconds/pixel for f/48 mode and
0.0221710.00001 arcseconds/pixel for f/96 mode. The case of the f/288 mode
is still under study.

In summary, the FOC is working perfectly well. We are now executing the SV
proposals.

-  Georges Meylan

-- WF/PC Update

The WF/PC was used in October 1990 for Target of Opportunity observations
of the Saturn White Spot, including a sequence of exposures which were
later used to make a short video (see p. 9). Pictures of the White Spot and
resulting cloud formations have been widely published (see December 1990
Newsletter).

Planetary camera exposures of Mars have also been obtained for Philip
James, (see cover) and calibration data have included earth flat fields,
internal flat fields, and exposures on the Omega Cen field.

The WF/PC was UV-flooded on 27-28 December 1990, using four orbits for the
WFC and fifteen orbits for the PC. Ultraviolet grism exposures on a
calibration star were used to measure cleanliness of the cold
field-flattening lenses in front of the CCDs before starting the UV flood.

The goal of the flood was to eliminate the effect known as Quantum
Efficiency Hysteresis (QEH) and to enhance the QE of the devices in the
region between 3000 
 and 4500 
, the so-called " coronene gap," where
response of the coronene phosphor is falling off and response of silicon is
rising with wavelength.

The WFC was flooded first, and showed excellent far-UV response after the
flood. During the UV flood of the PC, however, with PC camera head
temperatures at 20-30 C, the WFC became contaminated such that the UV
response in the 1300 
 to 1800 
 region was severely degraded.

It has been known for several years that far-UV observations with WF/PC
(using filters F122M, F157W, F194W and possibly F230W) would have to be
done immediately after UV flooding because a " low-temperature contaminant"
 builds up on the time scale of a day and seriously compromises the
response there. No far-UV science was planned during this first UV flood,
and the difficulties encountered may mean that far-UV observations with the
WF/PC will not be attempted. However, changes to the timing and temperature
profiles of various parts of the camera during the UV-flooding cycle may
prevent such a rapid return of the " low-temperature contaminant."

Unfortunately, HST entered safe-mode operation on 31 December 1990, raising
the temperature of the CCDs to about            -40 C, which is known to
produce further contamination, affecting all wavelengths. The contamination
was removed with a short episode of heating the CCDs above    0 C, so that
the effects of the UV flood would not be destroyed. The camera was then
returned to normal operation at -97 C.

Three weeks later, however, following wide-V exposures on the
Leiden-Berkeley deep survey area Lynx- 2, low-level contamination patterns
were noticed (" daisy patterns" ). The CCDs needed to be raised to about
-10 C before the contamination was removed. Even so, the effects of the UV
flood should still be intact. On 29 January 1991 the camera heads were
returned  to a new operating temperature of   -85 C in order to slow the
build-up of contaminants.

Calibration exposures on the Omega Cen field have been used to measure the
QE of the UV-flooded CCDs. Results in the 2000-5000 
 range are presently
inconclusive, but the response in the I-band is certainly deficient with
respect to pre-launch expectations, as reported at the November 1990 Users'
Workshop. The QE curves at I (F785LP) are low by about 40% with respect to
values published in the WF/PC Instrument Handbook. New QE curves will be
made available in time for Phase II proposals for Cycle 1.

-  Richard Griffiths

-- Deep WF/PC Image

A single 1800-second WFC exposure of a faint galaxy field in the wide V
filter (F555W)  was obtained in late January 1991 as a test of HST' s
capabilities on faint objects. A number of stars and galaxies can clearly
be seen in the four WFC CCD frames (see photo). In addition, this image
shows HST' s first unintentional detection of a moving target (the streak
on the right hand side). Some contamination may be evident on WF1 (upper
left), but was       removed after the camera was warmed to        -10 C
(see previous article).

Several performance figures should be of interest. Aperture photometric
measurements of one star in WF4 gives an estimated V magnitude of 21.52,
using the zeropoint supplied by the WF/PC IDT, measuring the star through a
0.3 arcsecond radius aperture, and applying a 1.5 magnitude aperture
correction (since only about 25% of the total light falls within such an
aperture).  This agrees very well with an estimated V=0.5(J+F) magnitude of
21.47 from photographic data supplied by Matthew Bershady (Yerkes
Observatory).  The random error of the counts within this aperture plus
uncertainty of the sky is 0.02 mag. Systematic errors can be expected to be
much larger, however, and the agreement could be completely fortuitous.

What is the limiting magnitude of this frame?  The star discussed above has
a peak intensity of roughly 275 DN above sky (on average, 5% of the total
flux of a star is in the central pixel) and the observed sigma in sky
values is 2.7 DN.  A three sigma peak stellar detection would thus be at
V=25.3.  With only a single frame available, such faint objects cannot be
distinguished from the numerous cosmic rays.  A/D saturation occurs at
V=18.7.

The observed sky level is 20 DN (corresponding to a sky flux of 0.084
electrons per second), and the ratio of photon noise from the sky to
readout noise (13 electrons rms) is roughly unity. In a half-orbit, 45-
minute exposure this ratio would increase to 1.2 and one could expect to
reach 0.3 magnitudes fainter, to 25.6.  One would certainly want to have
multiple exposures to reject cosmic rays.

Deconvolution techniques have been used to show that morphological
information and plausible light profiles can be extracted for galaxies down
to V~20 magnitudes in a single orbit. Furthermore, extended structure was
observed in a known faint (V=21.7 mag) radio galaxy with redshift z=0.465
(deconvolution by Rogier Windhorst at Arizona State University, using a
photometric version of CLEAN developed by Bill Keel).

These observations were taken as part of the SAO/ERO program. Extensive
analysis of this and similar FOC observations of faint stars and galaxies
is underway (including image restoration experiments). A full report is
available from the undersigned.

- Pat Seitzer and Ivan King

-- Progress on the WF/PC II

NASA Headquarters has approved a Science Team for WF/PC II, with membership
as follows :

John Trauger, PI (JPL)

Chris Burrows (STScI)

John Clarke (U. Michigan)

David Crisp (JPL)

Jay Gallagher (AURA)

Richard Griffiths (STScI)

Jeff Hester (IPAC)

John Hoessel (U. Wisconsin)

Jeremy Mould (Caltech)

Jim Westphal (Caltech)

This team will work to represent scientific interests within the WF/PC II
project and to assure the instrument is functioning and well calibrated.
These individuals may be contacted for further information.

Design work continues on the optics for WF/PC II in order to correct the
OTA spherical aberration. An independent review panel chaired by Charles
Townes will assess the new design on 12-13 March 1991 at JPL.

-  Richard Griffiths



-- Sample Spectroscopic Data Now Available

FOS and GHRS sample data are now available electronically through STEIS.
The spectroscopic data include a representative sample of observations made
since launch.  The files may be copied via anonymous ftp from the STEIS
subdirectory Instrument_News/FOS_GHRS_Data.

This directory contains seven FOS observations and twelve GHRS
observations, along with a README file.  Because of the large number of
files involved, you may wish to enter ftp with the -i qualifier to suppress
interactive prompting during multiple file transfers. The 161 individual
FOS files account for approximately 3.9 Kbytes of disk space, and the 335
GHRS files account for approximately 10.3 Kbytes of disk space.

These data may be examined and analyzed using the standard packages
available in IRAF/STSDAS.  If the final destination of your data is a VAX,
you must use the tool STSDAS.STLOCAL.SUN2VAX to correctly format the files
after ftp-ing.  Additional information is available through the SCARS
division at STScI.

- Bidushi Bhattacharya



-- Status of GO Cycles 1 and 2 Proposal Reviews

The Telescope Allocation Committee, comprised of a subset of the original
 Cycle 1 Subdiscipline Panels and TAC membership, met at the STScI during
25-27 February 1991 to reassess the initial GO program in the light of the
current HST capabilities.

The fundamental task of the Cycle 1 reassessment TAC was to examine all of
the proposals for continuing scientific viability and merit under the
changed circumstances. The TAC considered both the GO reassessment requests
and STScI technical feasibility reviews, according to the policies
summarized in the December 1990 Newsletter. One calendar year of estimated
available GO time was reallocated to programs from the originally approved
pool.

Only 10 of the 165 previously approved programs had been withdrawn entirely
prior to the review, but only another 10 requested substantially more than
a factor of two increase in spacecraft time, so that the oversubscription
was an unexpectedly modest 25%. That is, 1500 hours of high priority time
were requested versus the 1200 hours allocated previously and expected to
be available for GO Cycle 1.

The total supplemental and parallel requests were essentially the same as
previously allocated (480 and 380 hours, respectively). Most proposers
chose to reduce the scope of their programs rather than request
substantially more time to attempt the full original objectives.

Furthermore, only a few proposals had outstanding technical feasibility
questions prior to the TAC meeting. Hence, a larger fraction of the
programs than might have been anticipated could be retained, although all
of those which were had to have scientific goals approved by the TAC and by
the STScI Director in their revised forms.

The Director' s review of the TAC recommendations followed the meeting and
the PIs were notified shortly thereafter. Instructions for revised
budgetary submissions will follow soon. Further information about the
results of the Cycle 1 reassessment will be provided in the next issue of
the Newsletter.

The deadline for Cycle 2 proposals remains no earlier than 1 July 1991, and
the current progress of HST operations suggests that it may actually occur
on or near that date. The Cycle 2 peer review would then take place during
fall 1991, so that selected programs would be available for implementation
beginning in July 1992.

-  Nolan Walborn

-- New Exposure Catalog Due in May

 A new printed version of the Exposure Catalog will be compiled and
distributed in May 1991. It will also be posted on STEIS for computer
access. Proposers for Cycle 2 HST observations are urged to consult the
catalog to avoid duplication of observations already performed or slated
for Cycles 0 and 1.

The exposure catalog will incorporate the updated GTO programs due in
March, as well as the revised GO programs (following the February TAC
recommendations for Cycle 1) due in late April.

-  Abhijit Saha



-- How to Access Current HST Info

Information on the current status of HST, recently completed observations,
and planned observations can be found on STEIS. The December 1990 issue of
the STScI Newsletter gives details about how to use STEIS (the Space
Telescope Electronic Information System). Additional information and/or
help is available from the User Support Branch at STScI (301-338-4413,
userid USB, or 301-338-4551, userid REPPERT).

The reports posted on STEIS are derived from a variety of internal and
GSFC-generated sources, and are therefore rough cut and ephemeral. The
Institute will soon have revised reporting tools available which expand and
consolidate the information given in these reports. Comments and
suggestions are always welcome as we modify and improve our user
information systems. Specific existing information that does not appear in
the current reports can generally be provided by the User Support Branch
with a couple of days advance notice.

The reports listed below are ordered by earliest warning of an upcoming
observation:

Future Science Observations -  lists program ID number, title, and target
by SMS (Science Mission Specification) up to two months in advance of the
scheduled observation. It comes from an internal STScI report produced by
the Science Programs Division, and contains a list of currently planned
science observations organized in a manner similar to the Completed Science
Observations reports below. Please note that program schedules made more
than a few weeks in advance should be considered tentative. These reports
are installed in STEIS approximately weekly in the
Observer/Long_Range_Calendar directory and will eventually be replaced by a
longer term schedule.

Near-Term Plan -  lists all observations, including OV, SV, and engineering
programs, listed only by program ID number and title. Please note that
program schedules made more than a few weeks in advance should be
considered tentative. These reports are installed in STEIS under the
Observer/Near_Term_Plan directory once each week.

Weekly Timeline -  derived from an internal STScI report produced by the
Operations Division, this report contains a more detailed account of all
observations (science, calibration, and engineering) scheduled for the
designated week, including begin/end times, principal investigator, target,
instrument mode, filters, apertures, and so on. These reports are installed
in STEIS approximately once a week under the Observer/Weekly_Timeline
directory and show up about two weeks before observations.

HST Daily Reports -  excerpts from the GSFC reports written by Joe Ryan
(HSTP-G Operations Manager), each covering a 24-hour period beginning at
0800 Eastern time each day. Topics include activities accomplished in past
day, planned activities for upcoming day, and significant forthcoming
events. These reports are installed in STEIS as they come in (approximately
daily) under the STScI/HST_News directory.

Completed Science Observations -  derived from an internal STScI report
produced by the Science Programs Division,  this report contains
information on science observations indexed by proposal type and ID number.
It also gives the program title, a rough indication of the data quality,
target name, and the weekly SMS in which the observation was scheduled.
This report covers data taken by the science assessment and early release
observations programs, and GTO and GO programs. Not included are reports on
data from OV and SV programs, which can be found in the file called
weekly_summ_date. These reports are installed in STEIS approximately
          once a week under the Observer/Completed_Observations directory.

Additional information can be found in the electronic README files of the
appropriate STEIS directories.

-  Bruce Gillespie and Pete Reppert



-- Work on Restoration of HST Images Continues

Work on HST image restoration continues at STScI, and this effort is now
being coordinated with scientists outside the Institute through an Image
Restoration Working Group (IRWG) co-chaired by Bob Hanisch and Rick White
(STScI). The membership of the IRWG includes nine staff scientists at STScI
and fifteen collaborators at other institutions in the U.S., Canada, and
Europe (see table).

The IRWG has three general objectives: understanding the point spread
function, developing restoration algorithms properly adapted to HST, and
determining the best approaches to doing photometry. These are all
complicated problems, however, and much remains to be done. Progress to
date and the major outstanding issues are summarized below for each of
these objectives.

-- PSF Studies

A set of five consecutive images taken of R136 taken last fall has been
analyzed at great length. These images appear to show a change of telescope
focus over the period of an orbit. The possibility of a time variable PSF,
in addition to the known position dependencies of the PSF for the WF/PC,
adds another complication to doing high quality image restorations. Roberto
Gilmozzi (STScI) has studied these data carefully, and finds that much of
the apparent focus change is simply the effect of spacecraft jitter.
However, the possibility of short-term variations in the position of the
secondary is still open, and we hope to schedule observations to settle the
question. In any case, it will be necessary to know the spacecraft jitter
function in order to model the PSF properly.

The Telescope and Instruments Branch has begun work on computing an initial
catalog of point spread functions for the WF/PC and FOC using the TIM
software developed by Chris Burrows and Hashima Hasan (see December 1990
Newsletter). While these PSFs will not be rigorously correct for a
particular observation (since they will not include the jitter function and
will probably not correspond to the exact position of the secondary), they
provide a starting point for image reconstruction work.

The STSDAS software development group has begun transferring the
Burrows/Hasan Telescope and Instrument Modelling software (TIM) into the
IRAF/STSDAS environment. In addition to making TIM available on systems
other than VAX/VMS, this will allow us to compute PSFs much more quickly
using more powerful computers. We hope to have TIM available within STSDAS
in approximately three months.

One area in which additional software is needed is in the extraction of
empirical PSFs from observed data frames. This is currently a rather clumsy
process and we need to implement a facility similar to the DAOPHOT PSF
extraction program. Further work could be done in optimal PSF extractions,
where the known telescope optics (e.g., from TIM) are used to constrain a
best estimate of the empirical PSF.

--  Restoration Algorithms

Nick Weir (Caltech) has implemented a user-friendly front end to the
MEMSYS3 maximum entropy package. This is currently available on Unix
systems at STScI. Rick White has been investigating the Lucy-Richardson
algorithm, especially with regard to computing the restored image on a
higher density pixel grid than the original data. Mike Cobb (NRL) has
successfully demonstrated that WF/PC images can be deconvolved as separate,
overlapping patches, each with a separate PSF. This work was done on the
Connection Machine (a massively parallel architecture) and reported at the
January AAS meeting (BAAS, 22, 1281).

We recently obtained a start-up grant of time on the Cray Y/MP at the
Goddard Space Flight Center, and we have begun porting our image
reconstruction software to this system, beginning with IRAF/STSDAS. The
recently installed T1 communications link to Goddard (our new Internet
connection) should facilitate data transfer. We will explore the
possibilities of increasing the bandwidth of this link further if our work
on the Cray proves beneficial.

Successful image restoration requires that the data be properly conditioned
prior to the application of deconvolution algorithms. The data must be
flat-fielded, and cosmic rays and other bad pixels must be removed. The
Telescope and Instruments Branch, in cooperation with the IDTs, is now
working to improve the quality of flat field calibrations. Several
algorithms are currently available for cosmic ray rejection. Use of the
standard cosmicray task in the IRAF ccdred package has proved quite
successful in eliminating cosmic rays in WF/PC data, even for long
exposures. This algorithm uses a single data frame, however, and is far
from optimal. Recent software developed in STSDAS allows the use of two
frames taken at the same pointing, so-called CR-splits, to do
anti-coincidence testing for cosmic ray hits. As long as the telescope
pointing does not drift too badly, this method is very good at detecting
and removing cosmic rays. The technique will probably have to be improved
to accommodate CR-splits with a pointing drift by regridding one (or both)
images for the sake of the pixel by pixel comparisons, but correcting only
in the ungridded frames.

The STSDAS group has completed implementations of the Wiener filter,
Fourier quotient, and Lucy- Richardson deconvolution algorithms. Each
program can use either an observed or modelled PSF as input. The tasks
automatically normalize and center the PSF as appropriate for the algorithm
and operate on either one- or two-dimensional data. We continue working on
the Jansson-van Cittert algorithm, and plan to implement a maximum entropy
package that does not depend on the commercial MEMSYS software.

--  Photometry

The best approach to crowded field photometry in the presence of spherical
aberration is still unclear. An initial analysis of the WF/PC photometry by
Holtzman et al. (forthcoming Astrophysical Journal Letters issue on HST)
indicates that 6-7% photometry is about the best that can be done
currently. At least part of the problem, however, is not the aberration but
the current inaccuracies in the flat fields. We expect that the flat fields
will improve with time to the point where photometric results are limited
mostly by the aberration.

Crowded field photometry is complicated by the overlapping (and spatially
variant) point spread functions. The large PSF makes determination of the
sky level difficult. And if the PSF is not well known, the photometric
accuracy on faint stars will suffer particularly as their brightness
estimates are confused with the diffractions rings and tendrils in the
aberrated PSF. Given that the azimuthally averaged surface brightness of
the PSF halo ranges from 0.1-1% of the core brightness over radii from 1-3
arcseconds, and at a given radius the halo surface brightness varies by an
order of magnitude due to the tendrils and rings, it is clear that if the
PSF model is not correct the useful dynamic range near a bright star will
be limited to about a factor of ten.

  Possible focus changes and spacecraft jitter have a substantial effect on
photometry. An uncertainty in the position of the secondary mirror of as
little as 10 microns leads to a possible error in the scale size of the PSF
of 3%. At a radius of 1.5 arcseconds from the center of a stellar image,
this will cause a shift in the fine scale features of the PSF greater than
the size of a PC pixel. The effect on the photometry of fainter objects in
this region will clearly be substantial. Similarly, jitter at the level of
0.007 arcseconds (the nominal specification for fine lock) will be largely
insignificant even in the PC (only 16% of the pixel size), but the observed
jitter of 0.1 arcseconds will cause the light from a single PC pixel to be
distributed over five or more pixels. If this effect is not properly taken
into account in the PSF, the resulting photometry will be poor.

Recent work by Peter Stetson (DAO) indicates that with algorithmic
improvements to DAOPHOT, photometric errors of order 3-4% can be achieved
on Wide Field Camera images using an empirically determined, spatially
variable (quadratic only) PSF. Again, one can expect to do better using
good PSF models and multiple PSFs (i.e., treating the image as a set of
smaller subimages over which the PSF can be considered to be constant).

Roberto Gilmozzi (STScI) has studied the possibilities of doing photometry
using only the cores of the stellar images, an approach also investigated
by Holtzmann et al. The variability of the core/wing flux ratio in the PSF
as a function of field position makes the viability of this approach rather
uncertain. A similar method is being developed by Malamuth et al. 1991
(BAAS, 22, 1276).

We have not yet determined whether it is advantageous to do stellar
photometry on a restored image instead of the original data frame.
Experiments by Rick White and Bob Hanisch have shown that highly linear
image reconstructions can be computed, even using iterative restoration
algorithms such as the Lucy- Richardson technique. In such cases it may
then be better to do the photometry on the reconstructed image, or at least
to compute the initial object list on this image, than to work solely with
the raw data and PSF. This is one area where considerably more work is
required using carefully simulated images (so that the answer is known a
priori).

Observations of bright, extended objects such as the planets can be done
quite successfully with HST. Results to date have shown that even using a
spatially invariant PSF model, good image reconstructions can be computed
using several different techniques. These reconstructions typically have
artifacts at the 2- 3% level, however. Reducing these artifacts will
require better knowledge of the PSF and the application of restoration
algorithms which account for the spatial variance in the PSF (owing to the
large angular extent of the objects).

In summary, it would appear feasible to do photometry in moderately crowded
fields to better than 3% internal accuracy through (1) improvements in
WF/PC flat fields, (2) modifications to existing photometry programs and
developments of new algorithms, and (3) use of good model point spread
functions.

- Bob Hanisch

-- STSDAS Update

In conjunction with the image restoration work discussed in the previous
article, the STSDAS software development group has begun the implementation
of several image reconstruction programs.  To date, a program that can
compute either a Fourier quotient or Wiener filter has been completed, and
a second program implementing the Lucy-Richardson algorithm is now
available.  These will be distributed with the next general release of
STSDAS (late spring or early summer 1991).  Work has begun on the
Jansson-van Cittert algorithm and on the implementation of at least one
public domain maximum entropy package.

Work continues in many other areas.  The calibration pipeline and support
software for each of the Science Instruments is being updated as needed to
correct bugs and to compensate for actual in-orbit instrument performance.
A first implementation of a binary table's capability for the FITS I/O
package is now being tested. A new contouring task has been developed, and
world coordinates (right ascension, declination, or other image
coordinates) are now fully supported.

-  Bob Hanisch

-- Imaging with the Hubble Space Telescope - An Initial Study

Ed. note: This article was contributed to the Newsletter by GO Sue Simkin
(Mich. State U.), who presented her study in a poster at the AAS meeting in
Philadelphia.

-- Introduction

A series of models for different sub-arcsecond configurations of both
active galactic nuclei and gravitational lens images have been constructed.
These models are designed to examine the range of exposure times and image
recovery possible with the present imaging capabilities of the HST.

In principle many of the brighter imaging projects originally proposed can
still be accomplished by the use of appropriate image deconvolution
techniques. However, in practice the WF/PC spatial undersampling, time
constraints imposed by orbital characteristics and pointing stability, and
instrument saturation characteristics limit this recovery. In particular,
it is important to know just how accurately quantitative information can be
recovered using deconvolution techniques in the presence of realistic
noise.

To explore this problem, two general classes of models were built using a
combination of routines available in IRAF, STSDAS, and a separate,
VMS-based analysis package coded by the author. Their characteristics are
outlined below.

Point Spread Model Specifications -  Two PSFs, for the WF/PC F517N and
F547M filters, were constructed on a grid of 0.0108 arcsecond pixels using
the optical modelling software (TIM) developed by Chris Burrows and Hashima
Hasan at STScI (see previous Newsletter). All aberrations measured by mid-
July 1990 were incorporated in these PSFs, as were nominal configurations
for the OTA and PC secondary spiders and obstructions. The focus was set at
+0.76 microns from the paraxial position and spherical aberration of 0.5
waves at 550nm. The F547M PSF was modeled for a K star energy distribution
and the F517N was modeled for an A star.

Deconvolutions were done with an " erroneous"  F517N PSF focused  at +0.74
microns and spherical aberration of 0.48 waves at 550nm. The two F517N PSFs
differed by as much as 25% at points within a 0.5 arcsecond radius of the
central peak.

Models of Gravitational Lenses -  These consist of four images (JI-IV),
which are composites of one galaxy and four different lens sets. The galaxy
has brightness Mv=    -22.5, color B-V=1.4, and redshift z=0.5. It was
constructed from the sum of two de Vaucouleur profiles with core radii of
1.5 and 7.5 kpc and axial ratios 0.990 and 0.40, respectively, with 50% of
the total flux in each component.

The lensed images are from two models with lens eccentricity 0.1
constructed by     J. Hewitt (MIT). Each was scaled for both V=17 and V=20
QSOs at z=1.0. The angular separations of the images from the image center
are 0.66 and 0.72 arcseconds for the bisymmetric image, and 0.61, 0.68,
0.69, and 0.77 arcseconds for the slightly off-axis image. Intensity ratios
for the four lensed images are: 1:1:1:1 in the first case and 4:2:2:1 in
the second.

Models of AGNs -  These are based on an E galaxy of magnitude V=13.6 and
color B-V=1.4 (Mv=-21.4 at z=0.35), with a core radius of 2.8 kpc (5.6
arcseconds) and an axial ratio of 0.3 in position angle 120!.

Added to this is a set of three objects: a central ring of radius 0.64
arcseconds, two rings of radii 0.09 arcseconds offset from the center along
the galaxy major axis by one radius in opposite directions, and a central
point source. Two images were created by scaling the flux in these three
central objects to that observed from the [O III] l5007 line in the central
1.5 arcsecond region of Pic A (7.641014 ergs/cm2/sec). In one case, this
was distributed in the ratio1:1:0 (model AGI); in the second, the flux
ratios for the central objects were 0:1:3 (model AGII).

Observational Simulations -   The models were convolved with the
appropriate PSFs on a grid of 0.0108 arcseconds using two-dimensional Fast
Fourier Transform techniques. They were then " observed" with the PC by
binning them into 0.043 arcsecond pixels, scaling the detected photons by
the appropriate factor for the telescope plus detector plus filter
response, multiplying by an " integration"  time that kept the brightest
areas below detector saturation, and adding a preflash background. The "
observation times" were short, 7 seconds for the 17th magnitude QSO models
and 110 seconds for the 21st magnitude ones, 340 seconds for AGI and only
65 seconds for AGII.

Next, Poisson and readout noise were added on a pixel by pixel basis. To
test methods of reconstructing undersampled images, the AGII model was "
observed"  four times with four independent " noise" determinations, with
the grid position shifted in x-y space by different 0.0216 arcsecond steps
in each case (corresponding to zero point positions of [0,0]; [0,0.5];
[0.5,0.5]; and [0.5,0] in the [0.043] arcsecond pixel space).

-- Results

The lens models clearly indicate that for the image intensity ratios
considered, both the quasar images and the central object can be easily
measured after degradation by the aberrations, even with the undersampling
of the PC. The results suggest that much smaller angular separations (on
the order of 0.2 arcseconds) would also be discernible.

In addition, even though it is undersampled, the AGI model clearly shows
the central ring pair as well as the outer 1.3 arcsecond ring.
Deconvolution of this image using a maximum entropy (MEM) routine written
by T. Cornwell for images dominated by photon statistics yields a result
that is morphologically similar to the original model.

The AGII model, however, illustrates strikingly how a strong point source
can overwhelm a much lower surface brightness feature. Attempts to
deconvolve this with the same MEM routine that worked for AGI were
unsuccessful.

We then tried a more sophisticated approach. All four offset images were
combined to create an image that was adequately sampled. (Registration of
the images was done using a two-dimensional cross- correlation routine.) At
this sampling (0.0216 arcsecond pixel grid), the surface brightness of a
point source profile at 0.16-0.2 arcseconds from its center ranges from 0.1
to 1.3% of its central core. The surface brightness of the AGII model at
the same nuclear distance (in the rings) ranges between 0.09 and 0.1% of
its peak central brightness. Thus, after convolution, the contribution by
the central source to the total observed brightness at this point is as low
as 50% and may be as high as 93%.

As " observed,"  total detected photons lie in the range of 2700 per pixel
in the rings, leading to a possible range in signal to noise between 3.5
and 25. In spite of this low signal to noise, an iterative technique, which
first approximately subtracts the strong point sources from the raw data
and then applies a standard deconvolution routine, was successful in
resolving the inner ring structure. In this case, the image was deconvoled
with a " MEMSYS3"  based routine constructed by N. Weir and in experimental
use at STScI. This restores the morphology of the rings quite successfully
but completely obliterates any quantitative intensity information.

-- Conclusions

These results suggest that it is possible to obtain sufficient signal to
noise in 550- to 600-second exposures, even in undersampled PC images, to
restore HST images to their full 0.1 arcsecond resolution for point sources
brighter than V=22.5 magnitudes and features with surface brightness
brighter than 24.8 magnitudes per square arcsecond.

The key to this process is an accurate determination of the PSF matched in
wavelength response and position in the detector plane. This may be
possible with good calibration observations, but that is another (and more
difficult) project.

The models described above are available as FITS images via FTP for anyone
wishing to use them to explore new deconvolution and analysis techniques.

-  S. M. Simkin

-- Data Retrieval from the HST Archives

Although the normal proprietary period for HST data is one year, the STScI
Director has decided that all the Science Assessment and Early Release
Observations, taken during Science Verification in order to assess the
current capabilities of the telescope, will remain proprietary to the IDTs
for a period of just one month. Other observing programs carried out in the
Director' s discretionary time have also been allocated short proprietary
periods.

As a result of these decisions, there are already a number of
non-proprietary observations in the HST data archives. The Institute is
currently developing procedures to enable astronomers with a legitimate
scientific interest to obtain computer-readable copies of selected
non-proprietary observations. A catalog of the available data (including
target information, exposure parameters, archive data identification
numbers, and proprietary data release dates) will be posted on STEIS
together with the announcement of the request procedure as soon as the
Institute is ready to operate the service. General details pertaining to
the submission of Archival Research Proposals for specific data from the
HST archive will be distributed well before the Cycle 2 proposal deadline.

European and Canadian astronomers should wait for similar announcements
from the ST-ECF in Garching, Germany and the CADC in Victoria, Canada, who
both hold copies of the data files relating to these observations.

-  Ron Allen

-- Plans for STScI Ground System Improvement

Although HST flight operations are in progress, improvement of the HST
ground systems continues at full speed. Since March 1990, STScI has been
pursuing system enhancements based on a Long-Range Plan (LRP) for ground
system development.

The pre-launch plan has been affected significantly by HST's on-orbit
performance. In particular, resources intended for LRP projects after the
nominal seven-month commissioning period have had to be diverted as the HST
commissioning period has been extended. As a result, the LRP was revised in
December 1990 and will continue to be updated periodically.

The number of candidate LRP projects is large, as expected for a software
system containing close to two million lines of code. Prioritization of
candidate projects is an iterative process involving not only STScI staff
from several divisions but also the Space Telescope Institute Council, the
HST Users' Committee, and the Space Telescope Project at Goddard.

The major criteria for prioritization are:

% science capability improvements;

% HST observing efficiency improve        ments;

% HST operations staff productivity im        provements;

% replacement of obsolete hardware and         software.

Changes must have no impact on operations, so an extensive testing program
requiring several months is typically required before new code can be
installed. Some of the major projects now underway include:

Moving targets -  While observations of solar system objects have provided
some of the most spectacular HST images to date, the process of planning,
scheduling, and commanding these observations has required enormous manual
effort (see p. 24). Automation of this process for moving targets that can
be tracked by linear scans is nearly complete and will be installed in May
1991. Further improvements for more complex target trajectories will
follow.

Parallel science -  The original plans for parallel observing with HST were
complicated dramatically by the discovery during ground system testing that
readout collisions could cause the science instruments to stop reading out
data until reset by commands from the ground. As a result, the planning and
scheduling software and the instrument commanding rules have to be changed
to schedule readouts, not just exposures. These changes are in progress and
will be installed on the operational systems in November 1991.

Observing efficiency improvements -  Several projects to increase observing
efficiency are in progress. One of these, to have the Spike long-range
planning system minimize the impact of SAA avoidance, was completed last
summer. Another project, to modify the way exposures are split or ordered
to minimize wasted time, is proceeding based on an analysis of GO programs.

 Some of the pre-launch projects that required changes to the way
instruments are operated are presently lower priority than software fixes
to instrument problems discovered during commissioning. Taken together,
these efficiency improvement projects are expected to return to HST
observers some hundreds of hours of observing time per year that would
otherwise be lost.

WF/PC II -  The WF/PC II will be similar but not identical to the current
WF/PC, and will require changes to all of the STScI software systems.
Ground system testing of WF/PC II is expected to start in late 1992. In
order to be ready, changes are already in progress to add WF/PC II as a new
SI.

In addition to the science capability and efficiency projects described
above, other LRP projects address the replacement of the obsolete database
processor central to SOGS (Science Operations Ground System) and PEP
(Proposal Entry Processing) and the image and graphics display devices that
are no longer maintainable. Some of these devices are nearly ten years old
and are no longer manufactured or supported. Because the software systems
depend on the characteristics of these devices, major software changes are
required to replace them with current technology. One change that will be
welcomed by Observation Support System (OSS) users is that by the end of
this year, HST target acquisition will be done on IRAF/STSDAS workstations,
not on the soon-to-be-retired OSS Deanzas.

-  Mark Johnston

-- An Early GO Program is Started

In October 1990, Dr. Philip James (University of Toledo) and his
collaborators asked if their Cycle 1 GO program, " Synoptic Monitoring of
Seasonal Phenomena on Mars,"  could be undertaken early. This program is
designed to map the surface features of Mars quantitatively in the visible
and near-UV with the Planetary Camera, detect Martian ozone absorption with
the Faint Object Spectrograph, and also monitor seasonal variations.

After determining that there was an excellent chance that Dr. James' s
science objectives could be met with the current telescope performance, the
STScI Director approved starting these observations during the November
1990 Martian opposition; an occasion to observe Mars at opposition would
not recur for nearly two years, and no Cycle 0 program had planned
observations of the planet.

After frenzied planning and program modifications needed to enable the
observations during the HST commissioning period, science observations and
initial data analysis were begun successfully in December 1990 and have
continued through the spring. A black and white reproduction of an early
visible light image of Mars is on the cover of this Newsletter, courtesy of
Dr. James.

Dr. James and his team are distinctive in being among the first HST
observers who have not been directly involved with HST development,
operations, or commissioning activities. As we move intoCycle 1, a large
fraction of the HST science users will be astronomers from the general
community. Our experience with this program has given us an opportunity to
" tune up"  our user support systems and procedures. We are indebted to Dr.
James and his collaborators for working patiently with us, and for being
our first GO " customers."

-  Bruce Gillespie

-- Life as an Operations Astronomer

Trying to describe what it's like being an OSS OA ... Well, it' s chipping
your car out of the ice and inching your way to the Institute during a
winter storm so your counterpart can inch home and get a little sleep
before she has to come in again. And once you're there, it's anyone's guess
what will be waiting for you.

Could be that the data lines from Goddard have just died and you' ll have
to transfer operations down there (meaning another hour on the road). Could
be that the telescope safed and you' ll spend the shift writing procedures.
Could be that they finally got guide stars after trying for twenty hours,
and you' ll be scrambling to run " The Hunt for Pluto/Charon."  Could be
that the WF/PC suddenly developed a case of " hives"  and you' ll have
people crawling all over the station, wanting to see observations, wanting
to see engineering data, wanting special backup tapes, wanting to see our
scheduling information to set up real- time commanding, wanting ......

But all that' s the fun part. I' d better begin with some definitions and
descriptions, so you can understand what the fun' s all about.

OSS is the Observation Support System of the Science Operations Ground
System (that' s SOGS). We exist to command the telescope based on real-time
analysis of science data. We can tweak the pointing of the telescope by
several tenths of an arcsecond to insure that a target will be centered in
a very small aperture. We can locate a particular target that cannot be
precisely specified in advance from ground-based data, and command a small
angle maneuver (from arcseconds to tens of arcseconds) to center the target
in the aperture of the science instrument. We can also set instrument flags
to enable a pre-planned chain of commands, again based on analysis of data
taken shortly before the decision point.

In addition to real-time commanding, we also monitor observations as they
happen and do basic quality analysis on the science data. This is one of
the most important things we do, although it is usually not the most
exciting. We see all the data as soon as it comes off the spacecraft, so we
are in a position to spot problems early on. Because we get the data first,
raw though it may be, OSS is often used by the IDTs and observers for
preliminary evaluation and analysis of their data.

An OA is an Operations Astronomer. All OAs are research-level astronomers;
that is, we all have a Ph.D. or lots of publications (or both!) in an
astronomy-related field. My particular specialty is radiative transfer in
planetary atmospheres. We have 100% appointments in operations, although a
portion of our time can go to research if it is grant supported. In
practice, operations takes much more than 100% of a 40- hour week, and
research gets done (or often times doesn' t get done) on the " weekend."
(" Weekend"  is a euphemism for any two days in a row where you don' t have
to be on shift.)

The OA's primary responsibility is planning for all the activities coming
up, making sure that we' re ready to do what we have to do. In OSS, the OAs
work closely with an SOS (that' s Science Operations Specialist), all of
whom have a degree in a physical science and an intense interest in
astronomy. I cannot overstate the importance of the SOS in all that OSS
does: one of the more important jobs an OA has is running interference with
the observers, the IDTs, and the wandering onlookers, so the SOS can
actually get the work done. The SOS handles most of the actual execution of
commands, and is usually the person who knows what' s happening right now.
The OA is usually off in a corner, buried in the Mission Schedule, or
deeply engaged in conversation with the MOC SI Console Engineer and the
Shift Supervisor, trying to arrange a real-time contact.

Back to the original question: what' s it really like to be an OSS OA?
Well, a lot of what it' s like is related to the schedule we have to keep,
and the number of people we have available to staff it. OSS operates 24
hours a day, 7 days a week, 365.25 days a year. Three OAs and three SOSs
must be available at all times, including weekends and holidays. That means
you always come to work unless you' re sick unto death, and you hope the OA
following you can do the same. If someone can' t come in, someone else
loses their days off in order to cover the shift.

We currently are covering the shifts with four SOSs and five OAs. For the
SOSs, that means four people must cover three shifts a day. If we lose one
more person (e.g., to an accident or to a new job), the OAs will also be
covering three shifts a day with four people. In real terms that means just
two or three days off a month! We have one SOS in training (just started)
and one OA in training (almost ready). It takes six months just to learn
enough about the system(s) to cover a shift by yourself. It takes a full
year to really get to know your way around.

There are several reasons why it takes so long to train for OSS. One reason
is that OSS interfaces directly with almost every other part of operations.
(In case you haven' t noticed,  HST operations are complex.) We have to
learn what information we are supposed to get, where to get it from, how it
gets to us, what to do with it, how to figure out if it' s missing, and how
to ask for it if we need to.

Another reason is that OSS deals in real time; we have to be able to do it
right, first time, every time, and on time. That leaves little room for
error, and means that our training has to be both rigorous and detailed.

Finally, we have to understand all these systems well enough to figure out
ways to do the job when things go wrong. That means getting experience, and
experience takes time.

It all comes together on shift, when you never know what you may be called
upon to handle. I (and each of my co-OAs) have pulled proposals from the
brink of failure by putting together a real-time target acquisition
opportunity, from scratch, in minutes. (This is NOT the recommended way to
do things -  there are too many ways for it not to work.) And I have
watched a target acquisition uplink fail, and all the subsequent
observations go down the tubes, because I hadn' t thought to assume that
something critical (like getting guide stars) might take too long, and to
plan around that possibility. I don' t like to even mention the times that
I' ve had to nag an observer to make a decision, already, so we can send
the offsets...

All in all, being an OSS OA is very satisfying: I like coming in to find
the unexpected and the challenging every day. In the long run, shift work
does inevitably take its toll, and I wouldn't be surprised to burn out in
two to three years, sooner if we have to run short handed. But in the
meantime, I'll keep working enthusiastically at this job I'm so proud of,
and collectively with my fellow OAs, we'll continue to amass the expertise
that is so critical to HST operations.

-  Helen Hart

-- HST Talks Presented At Philadelphia AAS Meeting

Several talks and posters about HST, both its scientific results and
instrument performance, were presented at the January meeting of the
American Astronomical Society, held in Philadelphia.

A large and interested audience attended an oral session devoted to talks
from each of the instrument teams (FOC, WF/PC, FOS, and GHRS). This " last
set of preliminary HST talks"  was headed by Duccio Macchetto, who gave a
concise review of the effects of spherical aberration and used early
science results to demonstrate that each image revealed new information
about familiar objects.

Macchetto pointed out that no resolution has been lost, but that the image
core contains only about 15% of the total incoming light. The point spread
function shows this core surrounded by tendrils and diffraction lines.
Sensitivity is down by about two or three magnitudes from the pre-launch
expectation. The most difficult observations will be of crowded fields, but
with deconvolution, significant amounts of information are recoverable;
however, Macchetto concluded his talk with the caveat that software alone
will not solve all of the telescope' s problems.

Macchetto's talk featured a selection of early release data, much of which
was described in the December 1990 STScI Newsletter. The FOC image of Pluto
and its moon Charon allowed the separation to be measured extremely
accurately at 0.697 arcseconds. The image of R136 in 30 Doradus, which was
once proposed to be a 1000 solar mass star, showed it is in fact multiple
objects. The ring around SN1987A is clearly resolved in the FOC images,
something impossible to achieve from the ground. This has resulted in a new
determination of the distance to the LMC (see p. 7).

Finally, Macchetto showed the first images of strands in extragalactic jets
as seen by the FOC and the VLA. These filaments hold important clues for
theories of jet formation: taking the short lifetime of the emission and
the large size of the jets (several kiloparsecs) into account, a possible
explanation is that we are observing a limb brightening effect of electrons
flowing along a flux tube from the central object.

Jim Westphal spoke about the status of the WF/PC. A degree of UV
contamination was anticipated before launch, but visual contamination has
also been detected. The effect is most noticeable when the telescope enters
" safe mode"  (when the instruments are automatically turned off under
certain threatening conditions), resulting in the appearance of many
inverse pinhole images of the WF/PC pupil. When the instrument warms up,
the condensation, identified as organic material, disperses. A proposed
solution is to create a " soft safe"  mode which would leave the
instruments running.

Westphal concluded his talk with slides of Orion (see December 1990
Newsletter) and Eta Carinae (see p. 7).

 Richard Harms described the early FOS results as " quite heartening,"
although there might be some UV contamination, as indicated by data
obtained just before the meeting showing low sensitivity around Lyman
alpha.

New operating commands are handling a noisy reset line and the inadequate
magnetic shielding on the red detector. A viewgraph of quantum efficiency
versus wavelength showed the red detector to be down 30% from the same plot
with perfect imaging, a decrease in accordance with the predicted loss due
to spherical aberration.

Many features seen in the FOS spectra have not been commonly observed
previously. For example, one spectrum contained a Ne VIII line.

John Bahcall described the program for determining, for all five apertures,
how much additional time is needed to use the FOS. He also reviewed spectra
of 3C273 that had just arrived that very morning. Bahcall said, " This is
the first time we can look at a nearby quasar in the ultraviolet and find
out whether the rich quasar absorption systems which we see at large
redshifts actually show up at high resolution in small redshifts."  The
signal to noise ratio of the image of 3C273 was " absolutely spectacular."

The last speaker, Ken Carpenter, discussed the Science Verification results
from the GHRS. He summarized the effects of the spherical aberration on the
GHRS and stated that the team will be able to do most of the science they
intended using the small aperture with increased exposure time.

The preliminary GHRS data included Chi Lupi, Alpha Tau, the interstellar
medium around Xi Per (see p. 5), Beta Pic with its protoplanetary disc, and
three extragalactic images.

The first spectrum of an external target obtained with HST was of Chi Lupi,
a chemically peculiar star. The spectra included never before seen C I and
Ge II lines. The GHRS compares favorably with IUE, in several cases showing
clearly detected lines that are lost in the noise in IUE data.

A poster session devoted to HST featured many of the images mentioned
above. Sue Simkin announced that her point spread function model is
available for anyone interested in running simulations (see p. 21).

On the last day of the meeting, Larry Taff gave a short talk on the updated
guide star catalog.

As usual, STScI staff were on hand to answer questions and distribute
proposal information. The STScI video display featured the popular
Starfinder series produced by Maryland Public Television, animation
sequences by Dana Berry of the Institute' s Astronomy Visualization Lab,
and a fascinating 24-frame movie of the recent storm on Saturn (see p. 9).

-  Pete Reppert



-- Joint Discussion on HST Planned for IAU General Assembly in Buenos Aires

There will be a Joint Discussion session on " First Results from the Hubble
Space Telescope"  at the IAU General Assembly in Buenos Aires, Argentina.

This session will summarize the significant scientific results obtained
with HST during the first year of testing and operation, its current
performance capabilities, future plans for improving the performance
including second generation instruments, and additional information for
those wishing to propose for observing time.

 The preliminary date for the Joint Discussion is 25 July 1991. The members
of the Scientific Organizing Committee include A.A. Boyarchuk (P.-E. IAU
ex-officio), R. Giacconi, F.D. Macchetto, C.A. Norman (Chairman), with
other members to be confirmed.

Those interested in participating in the Joint Discussion may write to
Colin Norman, Chairman, SOC HST-JD, Space Telescope Science Institute, 3700
JJSan Martin Drive, Baltimore, MD 21218,  U.S.A. (301-338-4895, userid
NORMAN).

-  Nino Panagia

-- HST Users' Committee Considers COSTAR and the Advanced Scientific
Instruments for HST

The HST Users' Committee met on 30-31 January 1991 to consider a number of
issues pertaining to the ongoing and future performance of HST. Following
their deliberations, they summarized their position on COSTAR (the proposed
corrective optics for HST) and the Advance Scientific Instruments (WF/PC
II, STIS, and NICMOS) in a statement submitted to the Director and to NASA.
An excerpt of this statement gives the basic message:

 " The Committee responds positively to the report on COSTAR, supports the
concept fully, and feels that early progress on the optical and mechanical
design is going well.

" Certain basic ground rules should be followed, however. As noted by the
HST Strategy Panel, the value of COSTAR is predicated on its being
operational within a particular period, roughly 1994 through 1996. Delay
therefore reduces its value proportionately. Even worse, such delay could
also retard installation of WF/PC II and the new solar array. This simply
cannot be allowed to happen.

" To prevent any delay in this larger program, COSTAR should be managed
with a ruthless eye toward cost and schedule. At some point before the 1993
M&R mission, some hard choices may be necessary. To ease this process, it
would be wise now to develop guidelines for a minimally acceptable M&R
mission: can we go without COSTAR? without WF/PC II? without the solar
array? If not, how long should we wait? What is the maximum acceptable cost
for COSTAR? These guidelines would help us clarify the basic ground rules
under which the first M&R mission will be conducted. We have learned that
NASA Headquarters is planning to convene in March an important committee of
astronomers and others to consider these trade-offs, among other issues. We
strongly endorse this idea."

  Copies of the full statement are available from the Chariman of the HST
User's Committee, Arthur F. Davidsen:

Center for Astrophysical Sciences         Johns Hopkins University

Baltimore, MD 21218

(301-338-7370 or elan::afd or         [e-mail])

-  Pete Stockman

-- Sabbatical and Summer Visitors at STScI

In order to promote exchange of ideas and collaborations in HST-related
science, STScI has limited funds available to support visiting scientists
who wish to spend extended periods of time (three to twelve months),
typically on sabbatical leave from their home institutions or during the
summer, doing research at STScI.

In general, Sabbatical and Summer  Visitors have the status of STScI
employees and have access to the facilities available to staff members.

We are now accepting requests for support for sabbatical or summer visits
during the 1991/1992 academic year (September 1991 through August 1992).
Established scientists who are interested in this opportunity should send a
resume and a letter specifying the proposed period of time, and any other
relevant details to the Visiting Scientist Program c/o Tim Heckman, STScI,
3700 San Martin Drive, Baltimore, MD 21218, USA. Letters in hand by 1 May
1991 will receive full consideration.

-  Tim Heckman

-- STScI May Workshop

This year' s STScI May Workshop, hosted jointly by the STScI and the Hubble
Space Telescope Science Working Group (SWG), will be devoted to the first
year of HST operations. The workshop will be held at STScI and the
Bloomberg Center on the Homewood Campus of the Johns Hopkins University on
14-16 May 1991.

Invited presentations will describe the overall technical performance of
the observatory, including the scientific instruments, and many of the
exciting scientific results obtained during the first year of operation.
The primary goal of this workshop is to provide an up-to-date understanding
of the performance of HST and its scientific potential before the deadline
for submission of Cycle 2 proposals. Therefore, we encourage all current
and potential members of the HST scientific and technical community to
attend.

More information will be made available on STEIS or may be obtained by
contacting Barbara Eller, the conference coordinator (301-338-4836, userid
ELLER).

-  Pete Stockman

-- Hubble Fellowship Program

The 119 applications received for the second round of Hubble Fellowships
were considered by the Review Panel in late January 1991. At press time,
offers to successful candidates have been made, with replies due by
mid-February 1991.

The selection process should be completed by early March 1991, at which
time the names of this year' s new Hubble Fellows will be announced.

Dr. N. Katz, one of the first-round Hubble Fellows, has recently moved his
appointment from the Steward Observatory to the Massachusetts Institute of
Technology.

An Announcement of Opportunity for the third round of Hubble Fellowships
will be issued in early Summer 1991. The deadline for submitting
applications is anticipated to be mid-November 1991.

-  Nino Panagia

-- ESA Fellowships at STScI

Astronomers of European Space Agency (ESA) member countries are reminded of
the possibility of coming to STScI as ESA Fellows. Prospective fellowship
candidates should aim to work with a particular member or members of the
ESA staff at STScI, and for this reason applications must be accompanied by
a supporting letter from STScI.

Details of the interests of staff members at STScI can be obtained from Dr.
J. E. Pringle in the Academic Affairs Division (301-338-4477, userid
PRINGLE). Details of the fellowships and application procedures can be
obtained from the Education Office, ESA, 8-10 rue Mario Nikis, 75738 Paris
15, France. Competed application forms must be submitted through the
appropriate national authority and should reach ESA no later than March 31
for consideration in May, and no later than September 30 for consideration
in November.

-- Staff News

Blaise Canzian, previously at Caltech, joined STScI in October as a postdoc
working with Ron Allen. His scientific interests include the dynamics of
spiral galaxies, their interstellar media, molecular clouds, and star
formation.

David Soderblom, Chief of the Research Support Branch, began a sabbatical
in February at the Johns Hopkins University, which he will devote primarily
to ongoing research on the study of lithium depletion in young solar-type
stars and its implications for convection.

Andrzej Zdziarski, formerly an Assistant Astronomer in the Academic Affairs
Branch, left STScI in January to return to his native Poland. He will
continue his work on high energy astrophysical theory at the Copernicus
Astronomical Center in Warsaw.

Eric Chipman, formerly Chief of the Science Planning and Scheduling System
(SPSS) Branch, left the Institute to assume another CSC position, as
Project Manager/Chief Scientist of the Gamma Ray Observatory Science
Support Center at Goddard.

Doug McElroy  of CSC was appointed Chief of the SPSS Branch.

Tony Keyes from UCLA joined CSC as the Operations Astronomer in the SPSS
Branch.

-- Publication of HST Research

Publication of research papers based on HST data should carry the following
footnote:

Based on observations with the NASA/ESA Hubble Space Telescope, obtained at
the Space Telescope Science Institute, which is operated by the Association
of Universities for Research in Astronomy, Inc., under NASA contract
NAS5-26555.

If the research was supported by a grant from STScI, the publication should
also carry the following acknowledgment at the end of the text:

Support for this work was provided by NASA through grant number ____ from
the Space Telescope Science Institute, which is operated by the Associated
Universities for Research in Astronomy, Inc., under NASA contract
NAS5-26555.

For our records, please send one preprint of any research paper based on
HST data to:

Librarian

Space Telescope Science Institute

2700 San Martin Dr.

Baltimore, Maryland 21218.

Finally, please reference the relevant HST observing program identification
number(s) in your paper so that we can cross-index scientific papers with
the original observing proposals.

If you have questions regarding these instructions, please contact Bruce
Gillespie (301-338-4723, userid GILLESPIE) or Sarah Stevens-Rayburn
(301-338-4961, userid LIBRARY).

-  Bruce Gillespie and Sarah Stevens-Rayburn

-- Recent STScI Preprints

The following papers have recently appeared in the STScI Preprint Series.
Copies may be requested from Sharon Toolan (301-338-4898, userid TOOLAN) at
STScI. Please specify the preprint number when making a request.

485. " The Global Mass, Energy and Photoionization Balance of the Disk-Halo
Interaction,"  C.A. Norman.

486. " The Surface Brightness Test for the Expansion of the Universe. III.
Reduction of Data for the Several Brightest Galaxies in Clusters to
Standard Conditions and a First Indication that the Expansion is Real,"  A.
Sandage and J.-M. Perelmuter.

487. " 30 Doradus, Starburst Rosetta,"  N.R. Walborn.

488. " The Effect of Anisotropic Emission from Thick Accretion Disks on the
Luminosity Functions of AGN,"  C.M. Urry, P. Marziani, and M. Calvani.

489. " The Properties of External Accretion Disks,"  J.E. Pringle.

490. " HIPPARCOS Data Reduction and the Plate Overlap Technique,"  L.G.
Taff.

491. " Angular Momentum and Strippings in Tidal Interactions,"  T.A.
McGlynn and K.D. Borne.

492. " Time-Resolved CCD Photometry of an Ensemble of Stars in the Open
Cluster M67,"  R.L. Gilliland, T.M. Brown, D.K. Duncan, N.B. Suntzeff, G.W.
Lockwood, D.T. Thompson, R.E. Schild, W.A. Jeffrey, and B.E. Penprase.

493. " Star-Disc Interactions and Binary Star Formation,"  C.J. Clarke and
J.E. Pringle.

494. " The Role of Discs in the Formation of Binary and Multiple Star
Systems,"  C.J. Clarke and J.E. Pringle.

495. " NGC 891: A Summary of Observations,"  R.J. Allen and S. Sukumar.

496. " Southern IRAS Quasar with Extreme Fe II Emission,"  S. Lipari, F.D.
Macchetto, and D. Golombek.

497. " Si IV and C IV Resonance Lines as Indicators of Massive Stars in
Starburst Galaxies,"  C. Leitherer and H.J.G.L.M. Lamers.

498. " The Astrometric Calibration of the Hubble Space Telescope Fine
Guidance Sensors,"  L.G. Taff.

499. " Cooling Flows in Hierarchical Clustering Models,"  K.M. Ashman and
B.J. Carr.

500. " Echo Mapping of Broad Hb Emission in NGC 5548,"  K. Horne, W.F.
Welsh, and B.M. Peterson.

501. " The GUTs of AGN,"  C.M. Urry, L. Maraschi, and E.S. Phinney.

502. " Optical and IR Emission of a Sample of IRAS Galaxies,"  S. Lipari,
C. Bonatto, and M.G. Pastoriza.

503. " Main Sequence Angular Momentum Loss in Low-Mass Stars,"  D.R.
Soderblom.

504. " Collisional Pumping of H2O Masers in Star-Forming Regions,"  N.D.
Kylafis and C.A. Norman.

505. " Multifrequency VLA Observations of PKS 0745-191: the Archetypal
TCooling Flow'  Radio Source?"  S.A. Baum and C.P. O' Dea.

506. " He II Emission in Extragalactic H II Regions,"  D.R. Garnett, R--
Kennicutt, Y.-H. Chu, and E.D. Skillman.

507. " The Hydrodynamics of Relict Cosmological H II Regions and the
Formation of Objects at High Redshift,"  P. Madau and A. Meiksin.

508. " Models of Starburst Galaxies,"  C.A. Norman.

509. " The Complete Sample of 1 Jy BL Lac Objects I. Summary Properties,"
M. Stickel, P. Padovani, C.M. Urry, and H. Kuhr.

510. " The Effect of an External Disk on the Orbital Elements of a Central
Binary,"  P. Artymowicz, C.J. Clarke, S.H. Lubow, and J.E. Pringle.

511. " Fundamental Parameters of Brightest Cluster Galaxies,"  W.R. Oegerle
and J.G. Hoessel.

512. " The Starburst-AGN Connection,"  T.M. Heckman.

513. " Radius-Luminosity and Mass-Luminosity Relationships for Active
Galactic Nuclei,"  A.P. Koratkar and C.M. Gaskell.

514. " IUE Observations of NGC 5548, 1978-1988: The Sizes and Kinematics of
the Broad Line Region," A.P. Koratkar and C.M. Gaskell.

515. 1) " Rotation of Young Stars in the Orion Region,"  2) " Systematic
Effects in T Tauri Star Lithium Abundance Determinations,"  D.K. Duncan.

516. " The Warm Component of the ISM of Elliptical Galaxies,"  F. Macchetto
and W.B. Sparks.

517. " Dynamics of the Microwave-Decrement Cluster Abell 665,"  W.R.
Oegerle, M.J. Fitchett, J.M. Hill, and P. Hintzen.

518. " Binary Star Formation,"  J.E. Pringle.

519. " Variability and Structure of Accretion Disks in Cataclysmic
Variables,"  K. Horne.

520. " The Chromospheric Emission-Age Relation for Stars of the Lower Main
Sequence, and its Implications for the Star Formation Rate,"  D.R.
Soderblom, D.K. Duncan, and D.R.H. Johnson.

521. " Variable Soft X-Ray Excesses in AGNs from Non-thermal
Electron-Positron Pair Cascades,"  A.A. Zdziarski and P.S. Coppi.

522. " Observations of the Inner Core of the Southern Crab,"  D.
Burgarella, M. Clampin, and F. Paresce.

523. " Theoretical H II Region Models: The Effects of Stellar Atmosphere
Models,"  I.N. Evans.

524. " Supernova Remnant Evolution in an Interstellar Medium with
Evaporating Clouds,"  R.L. White and K.S. Long.

-- How to Contact STScI

Telephone: If an individual staff member's extension is not known, the
number for general use is 301-338-4700.

Telex: 6849101-STSCI

Fax: [phone]

Mail:     STScI

    3700 San Martin Drive

    Baltimore, MD 21218

    USA

E-mail: It is possible to reach most staff members at STScI on SPAN,
Bitnet, and Internet. Address formats are as follows:

SPAN:    SCIVAX::userid

            or 6559::userid

Bitnet:    [e-mail]

Internet:    [e-mail]

In most, but not all cases, the userid is the staff member's last name.
Alternatively, many userids are published in the Membership Directory of
the American Astronomical Society. If you have difficulty reaching someone,
please send the mail to the User Support Branch (userid USB), who will
forward it. The USB is the central point of contact for scientists who wish
to conduct research with HST.

-- Newsletter Notes

Comments on the STScI Newsletter should be sent to Meg Urry (301-338-4593,
userid CMU), who is serving as editor while Howard Bond is on sabbatical.
Any corrections, additions, or deletions to the mailing list should be sent
to Amy Connor in the User Support Branch (301-338-5015, userid CONNOR).

This Newsletter was put together by Bruce Gillespie, John Godfrey, Dave
Paradise,  Pete Reppert, Carl Schuetz, and Meg Urry using Aldus PageMaker
4.0 on an Apple Macintosh.

The STScI Newsletter is issued three to four times a year by the Space
Telescope Science Institute, which is operated by the Association of
Universities for Research in Astronomy, Inc., for the National Aeronautics
and Space Administration.