Äîêóìåíò âçÿò èç êýøà ïîèñêîâîé ìàøèíû. Àäðåñ îðèãèíàëüíîãî äîêóìåíòà : http://www.stsci.edu/science/preprints/prep1260/prep1260.ps Äàòà èçìåíåíèÿ: Mon Jun 29 15:39:54 1998 Äàòà èíäåêñèðîâàíèÿ: Sat Dec 22 14:16:25 2007 Êîäèðîâêà: Ïîèñêîâûå ñëîâà: cygnus loop

ULTRAVIOLET SKY SURVEYS
Instruments, findings, and prospects
NOAH BROSCH
Space Telescope Science Institute
3700 San Martin Drive
Baltimore MD 21218, U.S.A.
1
Abstract.
I review the development of UV and EUV astronomy, covering the spectral range from 5 to
300 nm, with emphasis on sky surveys for discrete sources. I discuss studies which resulted in
lists of sources observed by imaging and deliberately omit most spectroscopic studies. Technical
issues, such as detector and telescope developments, are treated separately from descriptions
of specific missions and their results, which contributed to the understanding of the UV sky.
The missions are compared in terms of their ``survey power'', a variable which combines sky
coverage and survey depth. I use the existing knowledge of UV sources to predict views of the
UV sky, which I then compare with those actually detected. Finally, UV missions which will
detect fainter sources and will fly in the near future are described, and a wish list for low­cost
ventures, which could advance considerably our knowledge of the UV sky is presented.
1. Introduction
Among all spectral bands, the ultraviolet has long been a neglected region, in which we hardly
have a good idea of how the sky looks like. This is despite the fact that in the UV there is a
distinct advantage of small payloads: first, the sky is very dark, thus detection of faint objects
does not compete against an enhanced background (O'Connell 1987); second, the telescope
construction techniques are very similar (at least longward of #70 nm) to those used for optical
astronomy. This apparent neglect is likely to change in the foreseeable future.
Another factor, which presumably acted against the initiation of new sky surveys in the UV,
was the argument that not much can be learned about the Universe from new UV sky surveys.
After all, it was said, we can infer about the appearance of the UV sky from a good data base in
the optical. I shall show later that this is only approximately correct. The availability of a mech­
anism which allows one to predict the appearance of the UV sky is a necessary ingredient in the
detection of outstanding sources, which may indicate the presence of new physical phenomena.
The short astronomical history of our knowledge of the UV sky can be divided into two
eras, the first from the dawn of UV astronomy until the flight of TD­1 and the second since the
1 On sabbatical leave from the Department of Astronomy and Astrophysics and the Wise Observatory, School
of Physics and Astronomy, Raymond and Beverly Sackler Faculty of Exact Sciences, Tel Aviv University, Tel Aviv
69978, Israel

2 NOAH BROSCH
availability of the TD­1 all­sky survey until today. Unfortunately, very little was accomplished in
terms of general sky surveys during this second era. We still await results from a major, modern
and sensitive UV all­sky survey, which hopefully will be performed in the early 2000's.
Likewise, the UV domain may be divided into the ``regular'' ultraviolet, that is, the segment
from just shortward of the spectral region observable from high ground­based observatories
(#320 nm) to just below the Lyman break at #90 nm, and the region from the Lyman break to
the fuzzy beginning of the X­ray domain. This short wavelength limit may be arbitrarily defined
as #6 nm (#200 eV). I shall call the first segment ``the UV'' and the second the ``extreme UV''
(EUV) bands. As shall be shown below, the observational techniques used in the EUV band are
more similar to those in X­ray astronomy, whereas the UV is more like the optical. Sometimes,
the UV region as defined here is separated into the near­UV and far­UV segments; the separation
is at #200 nm.
Although only few missions performed full sky surveys in the UV or EUV, a large number
of instruments scanned or imaged restricted sky areas. Those provided partial, or very partial,
information about the deeper UV sky. By thorough examination of their results one may form
an idea on what could be expected from a full sky survey to a similar depth. A review of UV
imaging experiments and their results, updated to 1990, was given by O'Connell (1991). A few
UV imaging experiments were described and compared by Rifatto et al. (1995), with the aim of
understanding galaxy UV emission.
In general, the sources of UV and EUV radiation can be either point or di#use. A summary list
of diverse sources, with the approximate order of importance in contribution, is shown below as
Table 1, which is derived from the similar table of Gondhalekar (1990). A note of caution should
be mentioned here: the relevance of the importance of various contributions to the background is
for broad­band measurements. In special circumstances, e.g., for narrow­band or spectroscopic
observations, line emission in the EUV domain may become the dominant contributor.
TABLE 1. Sources of UV and EUV
radiation
Source UV EUV
Direct starlight 1 1
Scattering o# dust 2 3
Hot & tepid ISM 6 2
H2 fluorescence 3 --
Galaxies 4 4
AGNs 5 5
I shall not discuss exotic mechanisms for the production of UV radiation (e.g., decays of
exotic particles), except in a single case where a UV observation by HUT (see below) puts the
most stringent limits on the mass of a heavy neutrino. The other sources shall be discussed
below, in the appropriate context.
It is clear that young stellar populations will emit copious amounts of UV radiation. To
demonstrate this, I show in Figure 1 how the peak of the spectral energy distribution changes
with age in case of a starburst galaxy model of star formation. For almost 150 Myrs the peak
stays in the UV for the three di#erent IMFs modeled here (Kroupa et al. 1993: heavy solid line,
Salpeter 1955: dotted line, and Miller & Scalo 1979: filled squares). All three models represent
instantaneous bursts, where 50% of the stellar ejecta is recycled into stars, all include nebular

UV SKY SURVEYS 3
emission, and in all 70% of the Lyman continuum is absorbed by the ISM gas. The models were
calculated with the codes of Fioc & Rocca­Volmerange (Leitherer et al. 1996). The steady UV
peak for #150 Myrs is not due to Lyman # emission; a comparison with models which do not
include nebular emission shows that this line can be important only during the few first Myrs
after the burst.
0.0 1.0 2.0 3.0
log (age[Myrs])
2.5
3.0
3.5
4.0
log
(Peak[A])
Wavelength of SED peak
Fioc+Rocca­Volmerange starbursts
Figure 1. Peak of spectral energy distribution of di#erent starburst galaxy models vs. age. The models di#er in
the IMF used to calculate the starburst. The dotted line represents the Salpeter IMF, the filled squares are for
the Miller & Scalo IMF, and the solid line represents models calculated with the Kroupa et al. (1993) IMF.
Before embarking on a description of various missions and their results, I discuss briefly the
units used in UV astronomy. A useful description of the brightness of astronomical sources is
by ``monochromatic magnitudes''. The monochromatic system is defined at wavelength # (using
the calibration of Vega by Hayes & Latham 1975) as:
m # = -2.5 log(f # ) - 21.175 (1)
where f # is the source flux density in erg/sec/cm 2 / š A at wavelength #. A variant of the magnitude
units is the AB system sometimes used in the UV. This is defined (Oke & Gunn 1983) as:
AB = -2.5 log(f # ) - 48.60 (2)
The constant was chosen to make AB=V for objects with flat spectra. The units of f # are
erg/sec/cm 2 /Hz. The conversion between the two systems of magnitudes is wavelength­dependent.
For #=200 nm, which is representative of the UV domain, the conversion is:
m # = AB - 2.26 (3)

4 NOAH BROSCH
Other units, useful for describing background brightness, are ``photon units'' (p.u. or c.u.=count
units). These simply count the photon flux in a spectral band, per cm 2 , per steradian, and per š A .
At 200 nm, one count unit equals 10 -11 erg/cm 2 /sec/ š A /steradian, or 10 -13 W/m 2 /nm/steradian,
or 32.9 mag/arcsec 2 . Of course, one can alway use the ``flux units'' (erg/sec/cm 2 /arcsec 2 / š A ) di­
rectly, or surface brightness units #I # (the units of I # are erg/sec/cm 2 /steradian/Hz). One unit
of surface brightness equals 5 10 7 c.u.'s.
Waller et al. (1995) introduced the S 10 concept, used in the visible part of the spectrum to
describe the di#use sky background, to the UV domain. S 10 measures the number of 10th mag
solar­type stars per square degree which would produce the same flux level as observed from the
di#use background. They quote for the region around 200 nm a conversion of
1 S 10 = 3.3 â 10 -21 erg/sec/cm 2 /arcsec 2 / š A = 3.33 â 10 -10 c.u. (4)
As the UV background does not originate from Solar­type stars, it is not clear what additional
insight the S 10 unit introduces, except when considering the influence of the zodiacal light.
This is because the evaluation of the zodiacal light contribution to the background relies on a
transformation from the visible band measurements by Levasseur­Regourd & Dumont (1980).
However, while the visible background is a manageable #200 S 10 (Roach & Gordon 1973), the
UV background is # 10 8 S 10 , as will become clear below.
2. The Early Years.
2.1. THE ''REGULAR'' UV
The beginning of UV astronomy lies in the mid­1950s, with rocket flights during which the skies
were scanned by the free (unstabilized) flight of the carrier vehicle. The first UV photometric data
were obtained by Byram et al. (1957), Kupperian & Milligan (1957), and Boggess & Dunkelman
(1959). In the early 1960s the first spectro­photometric observations were obtained using three­
axis stabilized rockets (Stecher & Milligan 1962; Morton & Spitzer 1966). However, rocket flights
o#er only a limited amount of observing time, of order 5 minutes or less, thus these surveys were
limited in depth and sky coverage.
The first instrument on a satellite which was dedicated to UV astronomy was a photome­
ter, on a US Navy spacecraft launched in 1964 (Smith 1967). One of NASA's first successful
``Observatories'' series of satellites was OAO­2, with UV cameras from the Smithsonian Ob­
servatory and a medium­dispersion spectrograph from GSFC, launched on 7 December 1968
(Davis et al. 1972, Code et al. 1970). The CELESCOPE imaging experiment, with its four 30
cm diameter telescopes, imaged 2 # wide fields with relatively low angular resolution. Its results,
consisting of photometric measurements of some 5,068 stars in four spectral bands, make up the
CELESCOPE catalogue (Davis et al. 1973).
Even manned space flights were harnessed into providing UV astronomical information; John
Glenn was given a hand­held 35 mm camera with an objective prism to record the UV spectra
of stars, but nobody checked whether the window of Freedom­7 was UV­transparent (it was not,
and no images were obtained; Boggess & Wilson 1987). Later, manned flights provided valuable
UV data from hand­held cameras on Gemini flights (Henize et al. 1975) and from an automated
Moon­based telescope (Carruthers 1973). A similar attempt at UV astrophysics from a manned
space platform was done on 19­24 December 1973, when the space observatory ORION­2 was
operated from the Soyuz­13 spacecraft (Gurzadyan et al. 1985). The ORION­2 telescope was a
22 cm Cassegrain with an objective prism, and the spectra of #900 stars were recorded on film.
Some information on the di#use UV background between 135 and 148 nm was published by
Hayakawa et al. (1969). The measurements were done during a rocket flight on 7 March 1967 and

UV SKY SURVEYS 5
the experiment, along with three­band photometry of six bright stars, as described by Yamashita
(1968). Approximately half the sky was mapped on 1 March 1970 in a UV band from 142.5 to
164 nm and with coarse angular resolution of about 10 # by a rocket­borne Geiger counter which
had an e#ective collecting area of 0.62 cm 2 . The experiment and its results were described by
Henry et al. (1977). The results were compared with a model of the UV sky calculated by Henry
(1977, see below).
The UV spectral range is blessed with very dark skies (O'Connell 1987), which permits
achieving a good S/N ratio on faint sources with relatively small optics, provided one has a
noiseless detector and keeps away from sources of background radiation. One well­known source
of background emission is the geocoronal Lyman # (Ly#), resonant scattering of Solar photons
o# hydrogen atoms in the halo around the Earth. A beautiful image showing the entire Earth
with its Ly# halo, as well as the tropical UV airglow bands and the auroral ovals around the
poles, has been obtained by the S201 experiment operating from the Moon during the Apollo
16 mission (Page et al. 1982, Fig. 4b).
The auroral emission is present essentially everywhere around the Earth, but its intensity
varies with geomagnetic latitude (Meier 1991). The Ly# emission, and the aurora and airglow,
are the two most troublesome backgrounds influencing orbiting UV experiments at low and
intermediate altitudes. The OI line emission at 130.2 and 135.6 nm band, and the N 2 Lyman­
Birge­Hopfield bands in the 140­180 nm region, are restricted to the upper atmosphere and a#ect
only observations done at low orbital altitudes, or at low incidence angles to the Earth limb.
The OI lines, in particular, form at altitudes of 250­300 km (Leinert et al. 1998). Special LEO
missions may, therefore, require special tailoring of mirror coatings and of filters to suppress this
background. Note that Ly# resonant scattering is produced also o# interplanetary hydrogen as
well as o# interstellar H 0 atoms passing through interplanetary space, thus no location in the
Solar System is really free of this emission.
2.2. THE EXTREME UV
In parallel with the development of the UV astronomy field, first steps were taken to study the
EUV sky. The EUV range is hampered by the opacity of the interstellar medium (ISM). From
91.2 nm shortward to about 10 nm the opacity is high, because of the photo­electric cross­section
of hydrogen, and to a lesser extent of neutral helium (below 50.4 nm) and singly ionized helium
(below 22.8 nm). The opacity of the ISM limits the detectable range to about 100 pc., with very
few exceptions. The objects expected to be detectable in the EUV were hot early­type stars, hot
white dwarfs, coronae of late­type stars, and bright non­thermal sources some of them possibly
extragalactic. Note that some of the EUV emission from early­type stars was much more intense
than expected from model atmospheres (e.g., # CMa; Vallerga et al. 1993, Wilkinson et al. 1996).
Also, not many very hot young stars were expected to be detectable because of the significant
intervening hydrogen column density to such objects.
The first studies in the EUV range were mostly, as in the UV domain, by rocket­flown
instruments (Henry et al. 1975 a, b, c), which measured a few very bright sources and established
calibrators. The earliest observations below Ly# were by Belyaev et al. (1971) with Geiger
counters on the Venera 5 and 6 spacecraft. These attempts were done with large field of view
detectors and concentrated on the detection of the EUV background. The EUV background
was also characterized (Kumar et al. 1974; Bowyer et al. 1977, 1981). The culmination of this
work was the EUV instrument flown on the Apollo­Soyuz mission in 1975, when four EUV point
sources were discovered (Lampton et al. 1976, Margon et al. 1976, Haisch et al. 1977, Margon et
al. 1978). At the completion of these, first, preliminary pilot surveys, NASA initiated the EUVE

6 NOAH BROSCH
survey, described in section 5.3.
The Voyager spacecraft explored the EUV sky with their Ultraviolet Spectrometers (UVS:
Sandel, Shemansky & Broadfoot 1979). For a number years, the UVS on the two Voyager
spacecraft were the most distant astronomical observatories in operation (Holberg 1990, 1991).
The operation of the UVS instruments is scheduled to end in 1998, in order to conserve power
for other Voyager instruments as the radioisotope power geenrators degrade. A look onto the
EUV sky, as side­benefit of an X­ray mission, was also o#ered by the ESA mission of EXOSAT.
This sky survey was performed from 0.6 to 40 nm and covered the short wavelength end of the
EUV range. The EXOSAT results were summarized by White (1991).
3. Technical issues
Before continuing with the description of UV missions in ``modern'' times, it is necessary to
discuss technical issues involved in UV observations. These have to do with both telescope
construction and detector development. Both aspects are driven not so much by scientific aspects
as by military usage of the UV, mainly for targeting missile launches and re­entry bodies.
3.1. OPTICS
In the telescope construction section we recognize that down to about 115 or 105 nm (and in some
cases even below that) ``regular'' telescope construction techniques apply. These imply parabolic­
hyperbolic mirror combinations (Cassegrain, Ritchey­Chr’etien, or even Schmidt configurations)
made of glasses, ceramic materials, or light­weight metals. The primary di#erence from ground­
based telescopes is that for space devices light­weightedness is a necessity. Therefore, space
mirrors will usually be light­weighted by machining the blanks to reduce the mass as much as
possible. Light­weighting to 40% of a solid blank is customary, although reports were published
of light­weighted mirrors to 20% or lower values.
The requirements for a reflecting coating for UV wavelengths above 115 nm are fulfilled by
aluminum overcoated by MgF 2 . Another 10 nm to shorter wavelengths can be gained if the
coating is LiF. Note that both materials are hygroscopic (LiF more than MgF 2 ) and the mirrors
have to be maintained in a controlled atmosphere until deployed in space. For UV observations
below 115 nm but above #50 nm the coating of choice has until recently been osmium or
iridium. These materials are inert at ground­level atmospheric conditions and o#er reasonable
reflectivities of #20%. Recent missions (second flight of HUT, second flight of ORFEUS, etc.)
used SiC mirror coatings. These allow high­e#ciency normal­incidence reflections for # #60 nm
(Keshi­Kuha et al. 1997). This mirror option was chosen also for the FUSE mission (Kennedy
et al. 1996).
One important issue with UV optics, as well as with EUV, is optics contamination. Not
only must dust particles be kept away, but also molecular contaminants. These are harder
to drive o#, because some have high molecular weights and will not evaporate fully during
outgassing procedures. In space, under the influence of high­energy particles and UV photons,
these substances change chemical form, may polymerize, and can form mono­layers over the
mirrors with very high optical depths in the UV (Noter et al. 1993). The significant reduction
of the UV throughput of HST's WFPC­1 by molecular contaminants, polymerized under the
di#use UV radiation from the Earth, was confirmed by MacKenty et al. (1995). It is possible
that a similar phenomenon was partially responsible for the sensitivity decrease of HUT in orbit
(for a discussion of the HUT sensitivity see Davidsen et al. 1992).
Observations in the EUV range cannot rely (with one exception) on normal­incidence optics.

UV SKY SURVEYS 7
For this spectral region the geometry of the telescope mirrors follows that used in the X­ray
domain, the Woltjer pairs, where the reflection is by grazing incidence. The one exception to
this rule is for narrow­band observations where the mirror acts as filter, by being coated with
multi­layers of di#erent metals. To improve the e#ciency, the telescope architecture normally
used in these cases is with the detector at the prime focus of a single parabolic mirror. This
technique was used for the ALEXIS telescopes and was planned for the first generation design
of the EUVITA telescopes for the Spectrum X­# (SRG) spacecraft (described by Courvoisier et
al. 1993; the option to orbit this configuration has since been abandoned).
In principle, multi­layer coatings can yield a total in­band reflectivity of about 40%, with
a bandwidth of about 10% of the peak band wavelength (Roussel­Dupr’e & Ameduri 1993). In
practice, it is extremely di#cult to produce multi­layer coatings on mirror substrates larger than
about 20 cm which are uniform, stable, and can survive the launch environmental stresses.
3.2. DETECTORS
The detector issue has been driven for many years by requirements of as high as possible quantum
e#ciency together with as good a resolution as possible. The first detectors used in UV and EUV
astronomy were photomultipliers with cathodes sensitive to the spectral bands to be observed,
or were just Geiger counters. TD­1 (see below) achieved spatial resolution by scanning the
photomultiplier aperture on the sky.
Most modern UV imaging detectors are intensifiers or image converters, which use a UV­
sensitive cathode as their main active ingredient. In most experiments the images were recorded
on film. This was processed on the ground after the observation and was usually scanned to
convert the information into digital pictures. Among the experiments which used this mode of
image recording we count SCAP­2000, FOCA, WF­UVCAM, FAUST (first flight only), and
UIT. These shall be discussed in detail below. At this point, it is worthwhile to remark that the
main limitations of this method of image recording are (i) the limited and non­linear dynamic
range, and (ii) possible film defects, that can be detected only after processing.
Carruthers (1973) developed a UV electronographic camera, patterned after visible­light
ECAMs. This camera extracted photo­electrons from a cathode by UV light and accelerated
them over #1000V potential di#erence. The electron trajectories were confined by the strong
magnetic field, parallel to the optical axis, of a permanent magnet wrapped around the cam­
era. The image was created by the energetic electrons onto special electronographic emulsion
placed at the focal plane. The readout was accomplished by micro­densitometry of the exposed
and developed emulsion. To achieve a better dynamic range, the same scene was imaged with
di#erent exposure times. Modern versions of Carruthers' cameras employ electron­bombarded
CCDs (EBCCDs) in place of the electronographic emulsion; each accelerated electron creates
many electron­hole pairs in the CCD, allowing high S/N detection of individual photons. The
obvious advantage of Carruther's design is that the quantum e#ciency of the opaque cathodes
he uses is much higher than that of the semi­transparent cathodes used in other designs. The
disadvantage is the need of an extended, strong, and uniform magnetic field along the axis of
the camera. This limits the size of the cameras built by his group to fairly small apertures.
It is also possible to use CCDs for UV observations, as done in the WFPC­1 and WFPC­2
of the HST. Even better UV performance is achieved by modern CCDs, with thinned, back­
illuminated chips which have fluorescing, anti­reflection coatings. A comparison of CCD and
other detectors, with specific application to low light level observations as encountered in as­
tronomical situations, was done by Vallerga & Lampton (1987). While the authors concluded
that the CCD and MCP (see below) devices are equivalent in terms of required number of in­

8 NOAH BROSCH
cident photons to reach a set S/N, the disadvantage inherent in the cryo­cooling required for
most CCDs is evident. In addition, MCP­based detectors with electronic readout o#er readily a
photon­counting option.
With the advent of multi­channel plates (MCPs) it became possible to build position­sensitive
detectors in the UV and EUV where photons could be ``confined'' to paths within individual
channels, while being accelerated and multiplied. With three­stack MCPs, gains of 10 6 to 10 7
can be achieved. Examples of such detectors are those of FAUST (in its second Shuttle flight),
and EUVE, where the readout is done by a position­sensitive anode (wedge­and­strip). The
design of TAUVEX (see below) is based on similar detectors. A review of readout methods for
photon­counting MCP­based detectors was presented by Lampton (1987).
At least three other forms of readout are available for MCP­based detectors. It is possible
to have a phosphor output activated by the electron cloud emerging from the last MCP of the
stack, which is optically­coupled to a regular, fast­readout CCD. Special fast phosphors, with fast
analyzing circuitry and hardwired, or transputer­based centroiding algorithms, yield sub­pixel
resolution of the center of each electron cloud. The resultant detector module is thus high­
resolution, fast, and photon­counting. Similar detectors are operating on the MSX UVISI and
on the future UV/optical monitor of XMM (see below). Recent results with this configuration
report resolution below 6 µm, resolving individual MCP pores, while supporting high count rates
(Vallerga et al. 1997).
A di#erent readout mode employs a resistive anode, with outputs at opposite locations at the
the anode edges and fast timing circuits to measure the propagation delay of the charge cloud to
each of the outputs. The centroid of the charge cloud is obtained by di#erencing the time delays
in the same direction. The new detector of FOCA (see below) operates in this mode. Finally,
another position­sensitive anode uses crossed pairs of helical delay lines behind the last MCP.
The readout is conducted by measuring time di#erences between the charge pulses received at
the two ends of a wire. Reported performance of helical delay line readouts appears to o#er very
high data rates (up to MHz counts) and very high resolution (about 20 µm, Siegmund et al.
1992). This, apparently, is the readout of choice for the GALEX UV sky survey mission.
The HST instrument Space Telescope Imaging Spectrometer (STIS), installed during the
refurbishing flight of 1997, includes one CCD detector for # > 200 nm and two Multi­Anode
Microchannel Array (MAMA) cameras for the 115­170 and 165­310 nm bands. The MAMAs
are essentially MCP intensifiers with CsI or Cs 2 Te cathodes, where the location of the electron
clouds is achieved by centroiding the readout of crossed micro­strip anodes.
A brief discussion of current detector technology for UV astronomy was presented by Ulmer
et al. (1995), and a more extensive discussion is by Joseph (1995).
Two kinds of promising detectors have not yet been used in UV astronomy. One is the
low­pressure multistep gaseous electron multiplier, which is used in high­energy physics for
detecting Ÿ
Cerenkov photons in ring­imaging detectors (e.g., Chechik & Breskin 1988). Such
detectors can reach very high UV quantum e#ciencies, up to 60% depending on the type of
gas filling the detector (A. Breskin, private communication), but have never been packaged for
space applications. The other type of detector is superconducting tunnel junctions packaged in
an array configuration (e.g., Perryman et al. 1994), which may prove to be the first energy­
sensitive imaging array for UV applications. This will be discussed below, near the end of this
paper.
New developments indicate that it may be possible to fabricate array devices similar to CCDs
which will have good UV sensitivitiy in the UV coupled with sharp response cuto#s near the
visible by using wide bandgap III­IV semiconductor materials, such as GaN and other nitrides
(e.g., Kung et al. 1996).

UV SKY SURVEYS 9
4. The TD­1 Era
The ``middle­ages'' of UV and EUV astronomy witnessed a number of high­atmosphere or space
missions, which shall be discussed below. In order to put these in proper perspective, I present
in Table 2 a summary of vital statistics for these missions. The spectral resolution is given in #
##
in the usual spectroscopic convention, where ## is the typical width of the spectral bands used
by each instrument. Only TD­1, IUE, and EUVE had spectroscopic capabilities. I include the
instantaneous field­of­view of each instrument in column 2. The table is arranged in chronological
order of missions and concentrates the three EUV surveys at its bottom. The entry for EUVE
includes the approximate spectral resolution of the imagers as well as that of the spectroscopic
telescopes.
TABLE 2. UV and EUV survey missions
Mission Inst. Apert. Spectral Spatial Spectral Launch End of Responsible
name FOV (deg.) cm. range (nm) resol'n resol'n year mission agency
TD­1 0.25 27.5 157­274 2' 6 1972 1974 ESRO
S201 20 7 125­160 3' NA 1972 1972 NASA (NRL)
FUVCAM 11­20 10 123­200 3' 3 1978 1991 NASA (NRL)
IUE 4 10 -3 45 110­320 NA 285 1978 1996 NASA, ESA, SERC
SCAP­2000 6 13 190­210 2' NA 1979 1990 CNES, CNRS, FNRS
GSFC CAM 11.4 31 140­262 1' 2 1979 1980 NASA (GSFC)
WF­UVC 66 20 125­280 5' 1 1983 1983 CNES (LAS)
GLAZAR 1.3 40 150­180 10''­40'' NA 1986 1991 Russia, Armenia
GUV 4 17 130­164 12' NA 1987 1987 Japan (ISAS)
UIT 0.67 38 125­290 3'' 2 1990 1995 NASA (GSFC)
FOCA 1.5, 2.3 39 190­210 10''­20'' NA 1991 -- CNES, CNRS, FNRS
FAUST 8 16.1 140­180 3'.5 NA 1992 1992 NASA, UCB, CNES
WFC 5 57.6 17­210 eV 2' 9, 120 eV 1990 -- SERC (Leicester)
EUVE 5 40 10­60 1' 1, 200 1992 -- NASA (Berkeley)
ALEXIS 33 10 13.3­18.8 15' 10 1993 -- NASA (LANL)
4.1. TD­1
The advent of modern UV astronomy can be marked as the first UV all­sky survey to a reasonable
depth and with a reasonable angular resolution, by the TD­1 satellite. As one of the first satellites
launched on 12 March 1972 by the European Space Research Agency (ESRO, later known as
ESA), the UV experiment was known as S2/68 (the second scientific satellite of year 1968). The
UV experiment was a collaboration of the British and Belgian scientists, and is described by
Boksenberg et al. (1973). It consisted of an f/3.5 telescope with a 27.5 cm diameter primary
mirror, feeding a spectrometer for the 130­255 nm region and a photometer with a single broad
band centered at 275 nm. The entrance aperture of the spectrometer was 11'.8â17' and the
photometer aperture was 1'.7â17'. The calibration of TD­1 was described also by Humphreys
et al. (1976).
Already during its operational period the TD­1 observations were used to search for outstand­
ing UV­bright objects (Carnochan et al. 1975). The results, in the form of an all­sky catalog

10 NOAH BROSCH
of UV sources, were published by Thompson et al. (1978). In order to produce the catalog, the
spectroscopic data were binned into three photometric channels, each 33 nm wide: 135­175 nm,
175­215 nm, and 215­255 nm. These, and the purely photometric 274 nm (31 nm wide) channel,
are the four TD­1 bands. The ESA­TD1 catalog contains 31,215 stars measured with S/N>10
in all four TD­1 bands. An unpublished version, with lower S/N restrictions, has 58,012 objects
(Landsman 1984).
The results of TD­1 were discussed by many and it appears that the TD­1 sensitivity limit
was not uniform across the sky. This was because of the variable number of scan repetitions of
the same source (higher at high ecliptic latitude), and also probably because of variations of
UV background, mainly in the 156.5 nm band due to geocoronal Ly# emission (Morgan et al.
1976). Gondhalekar et al. (1985) discussed the TD­1 results in the context of the galactic UV
interstellar radiation field emission. They mentioned that TD­1 is probably not linear for fluxes
fainter than 10 -12 erg/s/cm 2 / š A . Henry (1991) noted that the dark current of TD­1 varied with
time, another possible reason for non­linearity at low signal levels.
­150 ­100 ­50 0 50 100 150
­80
­60
­40
­20
0
20
40
60
80
Right ascension (degrees)
Declination
(degrees)
Figure 2. The TD­1 sky at 156.5 nm. The 31,215 sources of the published catalog were binned (in flux) into
``fat'' pixels to provide a fuzzy view of the entire sky. The intensity is shown logarithmically, to stetch the grey
scale, with brighter regions represented as darker pixels.
I show in Figures 2 and 3 the UV sky seen by TD­1. This was obtained by binning the flux
from the stars in the published catalog in boxes of 4 # â 4 # , and displaying the resultant average
surface brightness with a logarithmic scale. The view is in celestial coordinates; the Milky Way
is the darker (brighter) ``bullseye'' ring centered on both figures. The high flux area in the Milky
Way, visible on the upper right side of the image, is the Orion region. The dark region in the
diametrically opposite direction is produced by UV­bright stars in Scorpius­Sagittarius. Nearby
UV­bright stars produce isolated black fat pixels at random locations in the figures.
Based on the TD­1 results, Henry et al. (1988) produced an ``Atlas of the Ultraviolet Sky'',
which combines plots of the visible sky and of the corresponding 156.6 nm view. The faintest
objects shown in these plots are at 10 -12 erg cm -2 s -1 š A -1 , but among the objects fainter

UV SKY SURVEYS 11
­150 ­100 ­50 0 50 100 150
­80
­60
­40
­20
0
20
40
60
80
Right ascension (degrees)
Declination
(degrees)
Figure 3. The TD­1 sky at 274 nm. The format of the display is identical to that in Fig. 2.
than five times this value, some have had the UV brightness calculated from an optical­to­UV
transformation.
The all­sky coverage and the relative depth of the TD­1 S2/68 experiment have set its results
as a benchmark against which all other sky surveys are and will be measured. The TD­1 catalogs,
both the published S/N>10 version and the lower S/N unpublished version, are possible mines
of interesting objects. An example of the lasting value of the TD­1 survey is the report by
Landsman et al. (1996), who found two cases of white dwarf components in binaries with F­star
primaries. They selected sources with TD­1 156.5 nm UV excesses identified in the catalog as
late­type stars and followed them up with IUE. In two cases (56 Peg and HR3643), UV signatures
of white dwarf stars were detected.
Despite this long­lasting value, the TD­1 catalog is limited in the sky view it presents. The
TD­1 survey is as shallow in the UV as the HD catalog is in the optical; the sky knowledge it o#ers
through the total number of sources is equivalent to that of the visible sky as was known in the
early stages of astronomical photography, about 100 years ago (O'Connell 1992)! After TD­1, the
various UV and EUV e#orts can be characterized as either imagers or spectrometers. Among
the imagers, some were orbiters and others were on short­duration flights (rockets, balloons,
high­altitude planes). I shall list the various missions below, concentrating almost exclusively on
imagers.
4.2. ANS
This mission was launched on 30 August 1974 and was described by Van Duinen et al. (1975), by
Wesselius et al. (1982) and by de Boer (1982). The instrument consisted of a 22.5 cm diameter
telescope, which focuses the light through a 2'.5â2'.5 slit onto a spectrometer with fixed slits in
its focal surface. Each location is thus sampled in five spectral bands, defined by the sizes and
locations of these exit slits. The spectral coverage of ANS was from 150 nm to 325 nm and each

12 NOAH BROSCH
band was about 15­20 nm wide. The maximal sensitivity of ANS, determined by the instrument
design, was at 220 nm.
ANS did not perform a sky survey, but observed pre­selected targets. Wesselius et al. (1982)
reported that only 3,573 objects out of the more than 5,000 observed by ANS were actually
retained in the final catalog of the mission. These were objects with S/N> 4 in one band, or
S/N> 3 in at least three of the ANS bands. Measurements of 13 elliptical galaxies were reported
by de Boer (1982); in many cases only upper limits could be presented, because of the low
reflecting e#ciency of the grating, away from its blaze angle.
One important usage of the ANS data set was to derive UV­to­optical color indices from
observations of 182 main sequence stars and 56 giants and supergiants (Wesselius et al. 1980).
Note that the derivation of color runs only down to G0 stars; later­type stars were apparently
too faint for ANS to register. This is a characteristic of most UV surveys; the late­type stars are
under­represented, or absent altogether.
4.3. THE IUE OBSERVATORY
Undoubtely, the greatest success ever of any orbiting astronomical instrument lies with the IUE
observatory. Launched on 26 January 1978 for a nominal three years' mission, the observatory
operated for 18 years yielding more than 100,000 spectra of nearly 9600 diverse astronomical
targets. IUE operations closed down on 30 September 1996. The experiment consists of a 45
cm diameter Cassegrain telescope feeding two alternative spectrometers, one for the range 110­
190 nm and the other for 180­320 nm. Two di#erent dispersions were available, low (R=300)
and high (R=2 10 4 , with an echelle arrangement). The sensitivity of IUE, for low dispersion
operations, was aproximately m 150 =15. An early description of IUE can be found in Boggess et
al. (1978).
During the unexpectedly long operation period of IUE, its NASA operators at GSFC and
ESA personnel at VILSPA devised work­around methods to do with less than the minimal
number of gyros, and to operate despite unexpectedly high levels of straylight when a piece of
thermal blanket or reflecting tape fluttered in front of the telescope aperture.
Although the IUE data set does not represent a uniform survey of the sky, the large variety of
objects observed by it o#ers unique opportunities to derive ``average'' properties of populations.
This has been used by many (e.g., Fanelli et al. 1987) to derive UV­to­optical color indices for
various spectral types and luminosity classes. These are later used to derive transformations, to
create models of the UV sky (see below), or to determine the level of the di#use UV background.
Observations of galaxies are used to determine average UV spectra of irregular, spiral, and
elliptical galaxies, and of galactic bulges (Ellis et al. 1982, Burstein et al. 1988, Kinney et al.
1993, Storchi­Bergmann et al. 1994), important for the derivation of k­corrections.
The IUE data bank represents a valuable archival resource, even more so after the final
reprocessing of all the low­dispersion spectra into the final Uniform Low­Dispersion Archive
(ULDA) will be complete. To help the logical usage of the ULDA, atlases of UV spectra of
selected types of objects, based on IUE data were published (i.e., Longo & Capaccioli 1992).
Apart from the special­purpose atlases, note those dedicated to the classification of UV stars
by ESA and by NASA (Heck et al. 1984; Wu et al. 1991). The last distributed version of
ULDA (V4.0), including 54,247 spectra obtained until 31 December 1991, has been installed
at 27 regional or national centers. The final version of the archives contains 104,471 spectra
reprocessed with the most up­to­date calibration and includes echelle spectra binned to the
resolution of the low dispersion observations.

UV SKY SURVEYS 13
4.4. ASTRON SPACE STATION
The ASTRON space station was launched by the USSR on 23 March 1983. It was built on a
Venera­type platform and included a 3 # â 3 # (FWHM) proportional counter for observations in
the 2­25 keV X­ray regions and a 80 cm telescope for UV astronomy. ASTRON operated from
a highly elliptical orbit (#2000 km â # 200,000 km, four­day duration) until June 1989. Some
results on galaxy photometry in 2.8 nm wide bands, from 160 to 350 nm, were published by
Merkulova et al. (1990). Spectroscopy of SN1987A, and of flares on the red dwarf EV Lac, were
reported by Liubimkov (1990), by Burnasheva et al. (1989), and by Katsova & Livshits (1989).
An interesting and unique feature was the ability to perform fast UV spectro­photometry, with
a time resolution of 0.61 sec (Katsova & Livshits 1989).
The UV telescope on the ASTRON station did not operate in a survey mode, and the number
of objects observed by it was probably very restricted.
4.5. S201
The NRL experiment S201 was described by Page et al. (1982), where the revised list of sources
is given. A summary of results was given by Carruthers & Page (1984c). It consisted of an
electrographic Schmidt camera, which operated automatically on the Moon during the Apollo
16 mission in April 1972. The camera had a field of view of 20 # and an angular resolution of
about 3'. The limiting magnitude of the longest exposures, about 30 minutes long, was mUV =11
(typically mUV =10) in the spectral band 125­160 nm. The camera did not compensate for the
rotation of the Moon. This caused trailing of the long exposure images, by 0'.54 per minute.
The S201 experiment obtained the first UV image of the LMC (Page & Carruthers 1981) and
demonstrated the di#erence between its optical appearance, dominated by old, evolved stars,
and the UV where hot, young stars dominate. In total, ten fields each 20 # in diameter were
observed, thus the experiment covered #7% of the sky. The results of the S201 experiment were
discussed in a series of papers (Carruthers & Page 1983, 1984a, 1984b). A discussion of the UV
properties of nebulae in Cygnus was published by Carruthers & Page (1976).
4.6. GUV
The GUV experiment, flown on 21 February 1987 on sub­orbital flight with the S520­8 rocket,
was described by Onaka et al. (1989). It consisted of two 17 cm diameter Ritchey­Chr’etien
telescopes, which imaged fields 4 # in diameter. The telescopes were of f/3.2 F=36 cm design,
and the detectors used were CsI cathodes, tandem MCPs, and resistive anode readouts. The
cathode response, combined with the transmission of the BaF 2 windows, defined a spectral
band centered at 142 nm with a FWHM of 22 nm. The two sky field positions were o#set by
about 3 degrees and the final angular resolution, 16'x8'.2, was determined by the pixel size (2'.4),
the quality of the optics, and the stability of the platform in the pointing phase.
The GUV experiment produced rather shallow stellar photometric data on general sky regions
during the 133 second spin­stabilized ascent, in which 48 stars were detected, and deeper one­
band photometry of the Virgo cluster during the pointed phase, which lasted 181 seconds. The
limiting magnitude of the GUV experiment was approximately m 156 =14.6. Unfortunately, the
recovery of the S520­8 payload was not successful and the payload sank at sea, preventing the
possibility of post­flight recalibration.
The GUV observations were re­analyzed by Kodaira et al. (1990), and more than 40 galaxies
were detected in the Virgo cluster. The authors found correlations of UV emission with HI

14 NOAH BROSCH
flux and with FIR emission for spiral galaxies, and with X­ray and radio emission for elliptical
galaxies.
4.7. SCAP 2000
This is a very interesting collaboration between the Observatoire de G’en‘eve and the Laboratoire
d'Astrophysique Spatiale du CNRS of Marseille, running over more than two decades. The
collaboration (supported by CNES and CNRS in France, and by FNRS in Switzerland) produced
a stabilized balloon gondola (Huguenin & Magnan 1978) carrying a telescope tuned for imaging
observations in the UV. The experiment was described by Laget (1980), by Donas et al. (1981),
and by Milliard et al. (1983). It consists of a 13 cm Schmidt­Cassegrain reflector with a field of
view of 6 # and an e#ective collecting area of #95 cm 2 .
The detection was accomplished by an image converter­intensifier, which together with the
mirror coatings and absorption of the few mbar atmosphere at the balloon altitude defined
a bandpass centered at #200 nm and #15 nm wide, and was acheived mainly by multi­layer
coatings of the primary and secondary mirrors. The bandpass in identical to that of the newer
experiment FOCA (Milliard et al. 1991; see below). The final images of SCAP 2000 had a
resolution of about 1.5­2 arcmin and a limiting magnitude of about 13.5 for sources with a
spectrum similar to that of an A0 star.
SCAP­2000 surveyed about 15% of the sky, and some results, pertaining to galaxies and
their derived star formation rates, were published by Donas et al. (1987) and Buat et al. (1987,
1989).
4.8. WIDE­FIELD UV CAMERA
This instrument flew in December 1983, on the same SPACELAB­1 flight which carried FAUST
(see below) on its first orbital flight. The experiment was 1ES022 and was designed to image
a 66 # wide field with a resolution of 5 arcmin with an all­reflective camera. The WF­UVC was
a#ected by the high orbital background, just as was FAUST alongside it in the STS bay. Despite
this deficiency, due mainly to the high straylight at the Shuttle altitude in that specific orbit,
the best WF­UVC images reached m 193 =9.3. A description of the WF­UVC and its results was
given by Court’es et al. (1984). An interesting result was the detection of UV­bright stars along
the bridge connecting the Magellanic Clouds, an extension of the Shapley wing.
4.9. FOCA
The FOCA experiment flies a 39 cm diameter telescope on a balloon gondola to altitudes higher
than 40 km. There are two optical assemblies, one yielding a field of view of 1 # .5 and another
with 2 # .3, with image resolution of 10''­20''. The flights take place from a French launching
ground (Aire­sur­Adour) and end a few hours later in Italy.
The detectors were, as for SCAP 2000, image converter­intensifiers, and the image was
recorded on film. The measurements were done by PDS­ing the films, with proper calibration
of the intensities. The problem with both SCAP 2000 and FOCA were altitude changes of the
balloon, which modified the bandpass of observation because of the influence of the atmospheric
transparency. The FOCA experiment was described by Milliard et al. (1991). The bandpass
excludes contributions larger than 10% from the near UV to the visible for objects hotter than
the Sun, provided the balloon altitude is higher than 3­5 mbar and the zenith distance less than
55 # (Laget et al. 1991b).

UV SKY SURVEYS 15
FOCA surveyed some 70 square degrees of the sky. Some results on galactic metal­poor
globular clusters were reported by Laget et al. (1991a, 1991b). One interesting feature found by
the FOCA imaging is a small dark patch observed in visible against M13, which is considered
to be a foreground dark cloud, but which does not show up in the UV (Laget et al. 1991a).
FOCA results on UV images of nearby galaxies were reported by Vuillemin et al. (1991), and
analyzed (among others) by Court’es et al. (1993, NGC 4258), Buat et al. (1994, M33), Bersier
et al. (1994, M51), Reichen et al. (1994, M81) and Petit et al. (1996, M51). A study of UV
galaxies in the Coma cluster was published by Donas et al. (1995). Galaxy counts and color
distributions for objects in the magnitude range 15.0 to 18.5 were published by Milliard et al.
(1992); these served as basis for a prediction of UV galaxy counts by Armand & Milliard (1994),
which requires more late­type galaxies than predicted by optical data, or faster evolution of the
galaxies.
The LAS group have a parallel ground­based observational follow­up program, for system­
atic identification of their UV sources. Some of their UV sources are fainter than the limit of
POSS­I, requiring POSS­II data and dedicated observations at large telescopes, conducted (with
collaborators) at Palomar, Keck and WIYN telescopes. This confirms the identification as blue
galaxies, and their high projected density, which apparently reaches out to z=0.68 (Milliard
1996, private communication; Martin 1997). The 1 # .5 field centered on the Abell 2111 cluster
of galaxies contains #450 galaxies and #350 stars. The partial redshift survey mentioned above
indicates that most galaxies are in the fore­ and background, only #20% of them belonging
to the cluster. A similar study on the FOCA field centered on SA57 identified 45 sources with
mUV #18.5 with 40 galaxies, 3 QSOs and two stars (Treyer et al. 1997). The galaxies are mostly
later than Sb and their UV emission is interpreted as indicating enhanced star formation rates
in these objects.
During some test flights of FOCA in 1996 and 1997 the telescope was equipped with a 25 mm
diameter detector which has electronic readout (resistive anode). No results have been reported
yet from these flights. FOCA shall be upgraded with a 40 mm diameter detector, which matches
better its focal plane, and which shall have a crossed­delay­line anode readout for high angular
resolution observations.
4.10. FUVCAM
The NRL group headed by G. Carruthers flew a number of far­UV wide­field imagers (FU­
VCAMs) on rockets (Carruthers et al. 1978, 1980), which observed M31 and the North Ameri­
can Nebula. These flights used the Mark II FUVCAM, an electrographic Schmidt camera with
a field of view of 11 # and a (theoretical) resolution of 30''.
FUVCAMs flew on a number of rocket flights, and on a longer­duration space flight from 28
April to 6 May 1991, when it operated from the bay of the Space Shuttle Discovery (STS­39).
The experiment was constructed by the UV astronomy group of the Naval Research Laboratory
(G. Carruthers and collaborators). It consists of two electrographic cameras mounted side­by­
side and bore­sighted. Each camera covers a field of view of 10 # .5 square, with a final resolution
of #3 arcmin.
The images were recorded on film, which was later digitized. In order to overcome the limited
dynamic range problem, each field was imaged three times, with exposure times increasing by
about 3 to 10 times. The camera comprised two electronographic devices, one for the 105­160
nm band (123­160 nm with a CaF 2 filter), and the second for the 123­200 nm band (165­200 nm
with a SiO 2 filter). The experiment was described by Carruthers et al (1992) and its calibration
is described by Carruthers et al. (1994).

16 NOAH BROSCH
The results from the FUVCAM observations were published in a series of papers dealing with
individual fields: Monoceros (15 November 1982 rocket flight, Schmidt & Carruthers 1993a,
1994), Orion (6 December 1975 and 15 November 1982 rocket flights, Schmidt & Carruthers
1993b), and Sagittarius and Scorpio (Shuttle flight, Schmidt and Carruthers 1995). Not all
flights used exactly the same optical configuration, however the field of view was always very
wide.
Generally, the limiting magnitude of FUVCAM is mUV =9­10 mag and the observing band
was always shorter than 200 nm and longer than Ly#. The identification of UV sources was done
routinely by correlating against the SIMBAD data base. Thus, typically about 60% of the sources
were identified and between 24 and 40% of all sources could be attributed to blends of early­
type stars. Among the identified sources, about half are early­B stars. The raw, uncalibrated
FUVCAM images from the STS­39 flight are available on the web.
4.11. GSFC CAMERA (UIT PROTOTYPE)
A wide­field UV imager was flown by the Goddard Space Flight Center on a number of rocket
flights and with di#erent focal plane assemblies. The 31 cm diameter telescope had a field of
view of 11 # .4 and its spatial resolution was about 50'' FWHM. The detector was, in all cases, a
UV sensitive cathode with an MCP intensifier coupled to film.
Bohlin et al. (1982) described the instrument and its observations of the Orion nebula in
four spectral bands (140, 182, 224, and 262 nm) obtained during a flight on 11 December 1977.
Observations from a flight on 21 May 1979, when M51 was observed, are described by Bohlin et
al. (1990).
Smith & Cornett (1982) reported observations of the Virgo cluster with this telescope, where
the response band peaked at 242 nm and was about 110 nm wide. The sensitivity, for the Virgo
cluster exposure obtained during a flight on 22 May 1979, reached m(242 nm)=16.3.
The experiment also imaged the LMC (Smith et al. 1987) in two UV bands, 149.5 and 190 nm
with FWHM of 20 and 22 nm respectively. In the LMC, Smith et al. measured UV fluxes from
122 stellar associations, from which they derived a model for the progress of the star formation
process in this very nearby dwarf galaxy.
4.12. GLAZAR
A 40 cm telescope operated briefly on the Mir space station. This is the GLAZAR­2 telescope,
described by Tovmassian et al. (1991a), a direct follower of the GLAZAR­1 telescope (Tovmas­
sian et al. 1988). Its existence and results were described mainly in Soviet publications, thus it
was not well­known in the West. The limiting magnitude of GLAZAR­1 at 164 nm was originally
11 mag, but it steadily declined by about 2.5 mag from the launch and operation start in 1987
during the subsequent 2.5 years (Tovmassian et al. 1991b). This was probably a detector­related
problem (Tovmassian, private communication). A few scientific results were reported in this
describing paper.
The two GLAZAR telescopes have identical optical designs. They are 40 cm Ritchey­Chr’etien
designs, imaging a 1 # .3 field of view onto a 40 mm multi­channel plate intensifier with phosphor
output. The instantaneous image quality is #10'' and the images are recorded on film. The
spectral range of observation is determined by a filter close to the focal plane, with # c #164
nm and ## #25 nm. The filter acts as a vacuum barrier between the telescope tube (open to
space) and the detector­film transport mechanism. The film is exchanged with the telescope in
a parked position against a small airlock.

UV SKY SURVEYS 17
GLAZAR­1 was rigidly fixed to the MIR station; the low stability of this platform yielded
images of #40'' FWHM, thus the low sensitivity of the detections. GLAZAR­2 was mounted on
double gimbals and was equipped with coarse and fine star trackers, allowing access to wider sky
areas and better image quality through independent tracking of the telescope. This allowed it
to achieve #10'' resolution. The film was retrieved from the MIR by the supply spaceships, was
developed, digitized, and analyzed. Unfortunately, the film transport mechanism and the airlock
handle were damaged soon after launch and no results were obtained. Since then (#1990), the
GLAZAR­2 telescope has not been used.
Results from the GLAZAR flights, having to do with distribution and identification of blue
stars, were reported by Tovmassian et al. (1991c, 1992, 1993a, 1993b, 1994a, 1994b, 1996a). The
latter, in particular, emphasizes the decrease in sensitivity of the GLAZAR; only 217 stars were
detected in a field some 12 square degrees in size, with a limiting monochromatic magnitude of
m 164 <8.7 mag. The observations concentrated on OB associations. About 12 di#erent celestial
directions were observed and in each case an area of 10­20 degrees 2 was imaged. On all exposures
a total of 489 stars were measured.
4.13. FAUST
FAUST is the Fus’ee Astronomique pour l'Ultraviolet Spatiale, or the Far Ultraviolet Space
Telescope. It started life as a Wynne telescope (reversed optics: the secondary being larger
than the primary), whose purpose was to image the UV sky. The instrument was described by
Deharveng et al. (1979) and was originally intended for a rocket launch. This version of the
instrument was used for at least two successful flights, mainly used for calibrations.
The field of view of FAUST was # 8 # and the angular resolution was 1­2 arcmin. On
SPACELAB­1, during a flight on board the Space Shuttle in December 1983, FAUST did not
succeed to obtain significant data, although 47 exposures of 22 targets were obtained on film,
behind an image converter­intesifier. Unfortunately, the on­orbit background was very high and
almost no objects were recorded. From the few exposures which could be analyzed it is worth
mentioning an interesting image of the Cygnus Loop, published by Bixler et al. (1984).
For its second orbital flight FAUST was equipped with a new detector, which incorporated a
CsI cathode, a three­stage chevron arrangement of MCPs, and a novel wedge­and­strip anode.
The flight took place on board the Shuttle Atlantis in March 1992, and was described by Bowyer
et al. (1993). During this flight FAUST telemetered every detected photon. This allowed the
rejection of spurious photons, originating from the firing of the Shuttle attitude jets or from
high orbital background. FAUST observed 22 fields, among which were the North Galactic
Pole, the Virgo cluster of galaxies, and other galaxy clusters and regions of interest. FAUST's
observations yielded a catalog of 4698 UV sources (Bowyer et al. 1995), measured in a band
about 30 nm wide which was centered at 165 nm. The band was defined by the CsI cathode
and a CaF 2 window with multi­layer coatings. The detection was by an impartial automatic
algorithm, and the identification was through correlations with existing catalogs.
Selected results were published from the FAUST imagery obtained during the second flight.
Deharveng et al. (1994) analyzed the UV emission of galaxies. Haikala et al. (1995) imaged a
galactic cirrus cloud and showed the good correlation between the UV di#use emission and the
IRAS 100µm emission, from which they constrained the albedo a and the isotropy parameter
g of the dust particles. Sasseen et al. (1995) used the spatial power spectrum of FAUST im­
ages to search for an extragalactic UV background component. Sasseen & Deharveng (1996)
correlated the UV background detected by FAUST with the 100µm FIR emission measured by
COBE/DIRBE.

18 NOAH BROSCH
An interesting result was published by Court’es et al. (1995). It is a confirmation of the UV
extension of the Shapley wing of the SMC, first detected with the WF­UVCAM (Court’es et al.
1984). The FAUST observations are deeper (mUV <13.9) and have better angular resolution
than those of the WF­UVCAM. The observations confirm the earlier findings that the Shapley
wing contains a population of young stars formed at most 5 Myrs ago with an IMF which is
flatter than that of the stars in the SMC core (M up < 30M# ), i.e., later than mid­O type. The
FAUST observations of the SMC, combined with FIR, HI and H# data, form part of the thesis
of Okumura (1993).
A program to systematically investigate the FAUST data set takes place at Tel Aviv Uni­
versity. We re­detect sources with a di#erent automatic algorithm, based on local S/N ratio.
After attempting correlations with existing catalogs, we apply astrophysical criteria to identify
the sources. About 10­15% of the sources remain unidentified after this procedure. These are
identified with possible counterparts on the Palomar Sky Survey and the counterparts are ob­
served from the Wise Observatory. To date, we analyzed completely five FAUST fields, the North
Galactic Pole (Brosch et al. 1996a), three fields covering most of the Virgo cluster (Brosch et
al. 1997), and one image in the direction of Coma (Brosch et al. 1998). Analysis of other fields
(Ophiuchus and other southern fields) is very advanced.
The main results concern the distribution of UV stars, and in the case of the Virgo cluster,
also measurements of some 90 galaxies. In the Coma field we identified a large population of hot
evolved stars, which are probably connected with the open cluster Mel 111. In total, the FAUST
frames analyzed so far at Tel Aviv yielded #100 galaxies and a few hundred stars.
4.14. UIT
The Ultraviolet Imaging Telescope (UIT) was described by Stecher et al. (1992). It consists of a
38 cm diameter Ritchey­Chr’etien telescope with a 40 arcmin field of view (about 200 times larger
than the WFPC2 of the HST). The telescope was mounted on the Instrument Pointing System in
the Space Shuttle bay for the ASTRO­1 (December 1990) and ASTRO­2 flights (March 1995).
In addition to the tracking capabilities of the IPS used in the ASTRO flights, the UIT was
equipped with an articulated secondary mirror, which provided even finer tracking.
The focal plane consisted of two image converter­intensifiers, one for the far­UV and the
other for the mid­UV range. The image tubes were coupled to film on which the images were
registered. The film frames were digitized to a 2048â2048 pixel format, and to enhance the
dynamic range, multiple exposures of di#erent duration were taken through each filter. During
the ASTRO­1 flight more than 800 exposures were taken. The evaluation of the UIT results from
the ASTRO­1 flight indicates a resolution of about 3'' and a sensitivity su#cient to register stars
of UV monochromatic magnitude 19.5 (for a hot, unreddened source: O'Connell 1992). However,
for an indication of the actual sensitivity, see below.
First results from the UIT mission were published in a 1992 dedicated volume of the As­
trophys. J. Letters (vol. 395). The papers discuss the UV scattering properties of dust in NGC
7023 (Witt et al. 1992), observations of the Cygnus Loop (Cornett et al. 1992), of the Crab
Nebula (Hennessy et al. 1992), of globular clusters such as M79 (Hill et al. 1992a) and # Cen,
M3 and M13 (Landsman et al. 1992), of SN 1987A (Crots et al. 1992), the SN environment and
30 Doradus (Cheng et al. 1992), of the association NGC 206 in M31 (Hill et al. 1992b), of M81
(Hill et al. 1992c), of NGC 628 (Chen et al. 1992), of M33 (Landsman et al. 1992), of nearby
galaxies (O'Connell et al. 1992), and of NGC 1275 (Smith et al. 1992).
Most of the final results from both UIT flights have not yet been published, but some were
reported at meetings (notably, at AAS meetings). I mention in particular published studies of

UV SKY SURVEYS 19
nearby galaxies (M31: Hill et al. 1995a, and Magellanic Clouds: Hill et al. 1993, 1994, 1995b),
mostly from the ASTRO­1 flight. Pica et al. (1993) mention a catalog of #2,200 sources observed
near 250 nm derived from images of 66 fields obtained in the ASTRO­1 flight. Of these, about
300 did not have counterparts in published catalogues. The catalog (Smith et al. 1996) covers 16
square degrees of the sky and contains 2,244 objects culled from 48 pointings. The identification
was done through correlations with optical catalogs and, in fact, most sources are from the HST
Guide Star Catalog. The percentage of identified sources is 88%, mainly because of the good
spatial resolution of the images.
Unfortunately, the UIT catalog was based only on the near­UV (165­290 nm) observations
of the ASTRO­1 mission; during the ASTRO­2 mission this camera failed and no observations
were obtained. The catalog lists the sources obtained in the 48 selected fields, where the typical
exposure depth (though by no means the completion limit) was 17.2 mag.
From the ASTRO­2 flight, Ne# et al. (1995) reported that some 30 peculiar galaxies have been
observed in one or two UV bands (B1: 125­180 nm and/or B5: 140­180 nm). The galaxies include
interacting, starburst and/or active objects. Smith et al. (1995) reported at the same meeting
that five clusters of galaxies were observed during the ASTRO­1 flight, and additional clusters
were included among the targets observed during the ASTRO­2 flight. A recent interesting
paper (Hill et al. 1997) combines ASTRO­2 UIT observations at 152 nm and optical imagery to
determine the colors and extinctions of HII regions in M51. The authors find that the total­to­
selective extinction A(152)/E(B­V) in M51 increases with radius (or with decreasing metallicity).
In addition, they find that the H# flux is depleted in the inner regions of the galaxy; this they
interpret as increased Lyman continuum extinction. Parker et al. (1996) studied a few OB
associations in LMC using ASTRO­2 data at 152 nm.
O'Connell & Marcum (1996) remarked, from UIT images of galaxies, that a comparison of
visible and UV images of the same nearby galaxies indicates a trend from normal to abnormal,
galaxies turning into later morphological types with a higher incidence of irregular galaxies as
the wavelength of observation gets bluer. A similar claim was also made by Giavalisco et al.
(1996), based on simulated HST images using UIT images of nearby galaxies. Other papers
on galaxies, resulting from UIT imagery are by Waller et al. on M101 (1997), by Smith et al.
on NGC 3310 (1996b), on the SMC by Cornett et al. (1997), and a number of papers in the
proceedings of the Seventh Astrophysics Conference (Holt & Mundy 1996) by Waller et al. ,
Fanelli et al. , Ne# et al. , Hill et al. , Smith et al. and O'Connell.
The UV images of nearby galaxies obtained by UIT are the baseline templates showing how
galaxies appear in the UV, with which one can begin to understand the images of distant galaxies
taken by HST in optical bands; these correspond to rest­frame UV, proving the comparison valid
(Giavalisco et al. 1996).
4.15. UVISI
The UVISI instrument operated on­board the Mid­Course Space Experiment (MSX) satellite.
The mission was primarily military in character (BMDO), aiming at detecting and characterizing
sources of UV emission (or atmospheric opacity), which could a#ect the detection, identification,
and tracking of missiles and warheads. As side­benefits, these missions will produce full or partial
sky surveys in the UV. MSX was launched on 24 April 1996 and the operation of UVISI started
in 1997 and ended in early 1998.
The UVISI instrument consists of five spectrographic imagers and four imagers (Carbary et
al. 1994). The narrow­field UV imager is of interest to UV sky surveys. It images a 1 # .6â1 # .3 field
of view with a resolution of # 20'' in one of three spectral bands from 180 nm to 300 nm (180­300,

20 NOAH BROSCH
200­230, or 230­300 nm; J. Murthy, private communication). There is also a wide field imager,
with a FOV of 10 # â 13 # and a resolution of #3'. The detectors are CCDs coupled through
fiber­optic tapers to the phosphor outputs of image intensifiers. Based on the description of
UVISI (He#ernan et al. 1996), the narrow field UV imager with no filters is sensitive to sources
which produce 2 photons/cm 2 /sec; this should be equivalent to a limiting magnitude of 13.9
(monochromatic, at the band center, 240 nm). However, J. Murthy communicated an e#ective
sensitivity limit of mUV =20.0. The results of the UVISI observations have not yet been put in
the public domain.
4.16. HST
Although far from being a survey instrument, the HST has some UV capability with its WFPC­1
or WFPC­2 cameras, and with the FOC instrument. However, the long­wavelength rejection of
light by the WFPC filters is not fully satisfactory, with the notable exception of the F160BW
``Woods'' filter. Far UV observations su#er from red leaks, requiring complicated compensatory
measurements. The FOC has better rejection of optical light because of its cathode. Both cameras
have small fields of view: 2'.6 for WFPC­1, 2'.5 for WFPC­2 (total sky coverage 5.02 arcmin 2 ),
44'' for the pre­COSTAR FOC, and 28'' for the post­COSTAR FOC (7'' with full sampling
of the PSF and with #1.55% e#ciency). An example of the UV capability of the FOC is the
snapshot survey of nuclei of galaxies (Maoz et al. 1996), where circumnuclear star­forming rings
were identified in a few objects through 230 nm imaging.
The new instrument STIS has significant UV capabilities with the MAMA detectors. These
have a throughput higher by one order of magnitude or more than the WFPC­2 with UV filters
(F160BW or F170W), but their largest field of view is only 25''â25''. In particular, the far­UV
(FUV) MAMA with the Csi cathode and SrF 2 short band cuto# filter allows one to reach very
low background values while retaining reasonable throughput (#2.5% at peak). STIS with the
FUV MAMA and the SrF 2 cuto# filter have a throughput higher by # 27â than that of WFPC­
2 with the F160BW filter. This ``almost'' compensates for the field­of­view, which is smaller by
# 29â than that of the WFPC­2, disregarding the region vignetted when the F160BW filter is
used.
The next servicing mission for the HST, currently scheduled for December 2 1999, will see the
installation of the Advanced Camera for Surveys (ACS). This instrument is designed with three
separate channels, of which one has a UV capability and another is a solar­blind channel. The
High Resolution Camera shall cover a field of view of 27''â26'' with a plate scale of 0''.025/pixel
using a 1024â1024 CCD. The spectral range covered shall be from 200 to 1000 nm with a net
e#ciency of #15% in the UV part of the band. The Solar Blind Camera shall cover a FOV of
33''â30'' (1.6â that of the STIS FUV MAMA) with a plate scale of 0''.03/pixel using a spare
MAMA detector from STIS, with a net e#ciency of #4% at 140 nm.
The latest instrument (for the time being) selected for an HST upgrade is the Cosmic Origins
Spectrograph (COS). This shall replace COSTAR during the fourth HST servicing mission in
2002. It is an instrument optimized for high­throughput spectroscopy of point sources in the
115­205 nm band. The high throughput is achieved by minimizing the number of reflections
between the aperture and the detector (a single reflection, at the grating), allowing more than
one order of magnitude improvement in this aspect relative to STIS. The spectral resolution
can be high (R#20,000­24,000) or intermediate (R#2,500­3,500). The esimated sensitivity is
such that, in high resolution mode, a source with monochromatic magnitude 15.6 will produce
a spectrum with S/N=10 per resolution element in 10,000 sec.
Based on the ``survey power'' parameter # to be described below, UV imaging by the HST

UV SKY SURVEYS 21
has significant survey potential, despite the small sky area it can cover. In fact, while a STIS­
based survey would only be # 1.6â more e#ective than one with the WFPC­2, a survey with
the ACS/SBC would be # 45â better.
4.17. BACKGROUND MEASUREMENTS
Attempts to measure the di#use UV background in the TD­1 era are lumped together in one sec­
tion. They consist of observations by the UV photometer on the Apollo­Soyuz mission (Paresce
et al. 1980), the measurements of the D2B­Aura satellite (Maucherat­Joubert et al. 1980, Jou­
bert et al. 1983, Lequeux 1982), sky measurements from a rocket flight (Jakobsen et al. 1984),
and from the Dynamics Explorer satellite (Fix et al. 1989), and by the two Shuttle­borne UVX
instruments from JHU and Berkeley (Murthy et al. 1989; Hurwitz et al. 1989). In addition,
observations done for other purposes but used to derive the UV background were by Zvereva
et al. (1982), Weller (1983), Onaka (1990), and Jakobsen et al. (1984). The UVX observations
yielded one of the lower background values at 160 nm: 280±35 c.u. (Martin et al. 1991).
The FAUST instrument has been used to derive the UV background, with the emphasis on
an attempt to disentangle the Galactic from the extragalactic signal based on spatial power
spectra (Sasseen et al. 1995). Their best explanation for the signal detected is that it is due to
starlight scattered o# dust grains. This is apparently confirmed by the correlation between the
UV background measured by FAUST and the 100µm FIR emission measured by COBE/DIRBE
(Sasseen & Deharveng 1996).
The observations of UIT have also been used to analyze the UV background. Waller et al.
(1995) quote orbital night­time background levels of #3000 ph/sec/cm 2 /ster in the near­UV
band (#250 nm) and #5000 ph/sec/cm 2 /ster in the far­UV band (#150 nm). After correcting
for instrumental and orbit­dependent e#ects, and compensating for galactic di#use UV radia­
tion through a correlation of UV and FIR di#use emission, Waller et al. identified a possible
extragalactic component of 200±100 c.u. As will be discussed later, the UIT background mea­
surements and some deep Voyager UV spectra are now the strongest constraints for a cosmo­
logically interesting UV background. It appears that these constraints conflict extrapolated UV
observations of faint galaxies.
5. Modern EUV observations
5.1. ROSAT WFC
The EUV sky was explored almost simultaneously by two instruments. The ROSAT EUV Wide
Field Camera was an add­on instrument to the ROSAT X­ray all­sky survey satellite. Its first
results, and a description of the instrument, were published by Pounds and Wells (1991). The
camera consists of a nested three­mirror Woltjer­1 telescope with a MCP detector equipped with
a CsI cathode at the common focus of the mirrors. Selectable filters allow observations in four
spectral bands: 52­73 nm (17­24 eV), 14.9­22.1 nm (56­83 eV), 11.2­20 nm (62­111 eV), and 6­14
nm (90­210 eV).
The first EUV all­sky survey was performed by this instrument during 1990­1991 (Pye 1995).
The survey was done in two bands: S1 (6­14 nm) and S2 (11.2­20 nm). The e#ective spatial
resolution was 3 arcmin and the source location was good to within one arcmin. The initial
results were reported by Pounds et al. (1993) as the WFC Bright Source Catalog (BSC). The
reprocessed data, with more sources, were published by Pye et al. (1995) as the 2RE catalog.
The 2RE catalog contains 479 EUV sources, 120 more than the BSC, observed with a median
exposure of about 1600 sec. Of the entire 2RC catalog 52% of the sources are identified as active,

22 NOAH BROSCH
late­type (F, G, K, and M) stars; 29% are hot white dwarfs, and less than 2% are extragalactic
sources (AGNs). Thirty­four sources (about 7%) were still unidentified in 1995 (Pye et al. 1995).
The cumulative source distribution indicates that the late­type stars dominate the source counts
at faint count rates (<0.02 cps), surpassing the contribution from white dwarfs.
5.2. ALEXIS
The region bordering the EUV and the X­rays was explored by the ALEXIS spacecraft (Pried­
horsky 1991). ALEXIS stands for ``Array of Low Energy X­ray Imaging Sensors'' and consists
of an array of six wide­field small telescopes with multi­layer coated mirrors with prime­focus
imaging detectors, and sensitive to radiation at 66, 71, or 93 eV (18.8, 17.2, and 13.3 nm). The
passbands are defined by the mirror coatings and by filters positioned in front of the detectors.
Recent descriptions of ALEXIS are by Bloch (1995) and by Roussel­Dupr’e & Bloch (1996).
Unfortunately, during the Pegasus launch on 25 April 1993 the satellite was damaged, lost
one of its four solar panels, damaged the on­board magnetometer used for position­sensing, and
entered an uncontrolled spin. This reduced the available electrical power and questioned the
e#ectiveness of the survey. The spacecraft was eventually stabilized and e#ective observations
could be made from mid­July 1993. ALEXIS points its six telescopes perpendicular to the Sun­
Earth line and scans the sky with them every 50 seconds. The attitude of the satellite is known
to better than 0 # .5 and the information collected is received at a Los Alamos ground station.
The EUV telescopes of ALEXIS are arranged in three co­aligned pairs. Each has a 33 # field of
view and a resolution of 0 # .25 (limited by the spherical aberration of the mirrors). The first sky
maps of ALEXIS were produced on 4 November 1994. Since 1995, the daily maps are searched
for EUV transient sources, with a 12­24 hour response time. The transients are then serched
for in the error ellipse of ALEXIS by ground­based telescopes, and are followed up. Among the
more interesting EUV transients the more notable are the super­outburst of VW Hyi in June
1994, the ALEXIS J1114+43=1ES 1113+432 outburst in November­December 1994, and the
fast transient ALEXIS J1139­685. In addition, outbursts from U Gem, AM Her, and AR UMa
were also detected. Some results on transient EUV sources were reported by Roussel­Dupr’e et
al. (1996).
Using the pre­flight information, the ALEXIS team calculated that #10% of the brightest
EUVE sources (see below) should be detectable (Je# Bloch, private communication), and ap­
parently this is the situation. Most of the sources are WDs, and it is expected that the catalog
from the first three years of operation will contain #50 sources.
5.3. EUVE
The EUV sky was thoroughly investigated by the Extreme Ultraviolet Explorer (EUVE) space­
craft, launched on June 7, 1992. (Bowyer & Malina 1991). The instrument consists of three
``scanning'' telescopes, co­aligned and perpendicular to the spacecraft spin axis, which have a 5 #
field­of­view, and a deep survey spectrometer telescope looking along the spin axis and pointing
away from the Sun. The angular resolution is #1 arcmin.
EUVE mapped the sky in four spectral bands, from 7 to 70 nm (18 to 170 eV). The initial
results were published as ``The First EUVE Source Catalog'' (Bowyer et al. 1994), which contains
410 sources. A Second EUVE Source Catalog has also been published (Bowyer et al. 1996). For
the foreseeable future, the EUVE catalogs set the standard in the knowledge of the EUVE sky.
The second EUVE catalog includes additional observations to those from the all­sky survey.
In total, there are 734 sources, of which about 65% have optical, UV, X­ray, and/or radio

UV SKY SURVEYS 23
counterparts. It is noteworthy that 211 of these sources were never before observed in the EUV
range. The majority of the identified sources (55%) appears to be late­type stars (G, K, and M),
which presumably have active chromospheres.
The observation that large numbers of late­type stars are detected as EUV sources is sup­
ported by the findings of the ``right angle'' survey, which has the deepest sensitivity. These
observations are typically 10 4 ­10 5 sec, whereas the typical observations used for the derivation
of the catalog are #500 sec, thus the right angle survey can detect much weaker sources than
in the general survey. The late­type stars make up almost half the sources in the right angle
survey, whereas they are only less than a third of the sources in the general sky survey. The
late­type stars are also the absolute majority among the EUV sources detected in the ``ecliptic
deep survey'', another sky region with very high exposure (#20,000 sec/pixel over a 2 # â180 #
area along the ecliptic). Among all the sources, #8% are identified with extragalactic objects.
Bowyer et al. (1996) compared the second EUVE catalog with other EUV surveys. They find
that 263 (of the 479) 2RE sources appear in the new catalog, and note that the undetected
sources are either variable or belong to the ``unidentified'' category of faint sources that may be
spurious.
A new catalog, reaching sources fainter by #60% than the second EUVE catalog, has been
produced by Lampton et al. (1997). The catalog is based on coincident sources between the
EUVE 10 nm list and the ROSAT all­sky survey sources detected in the broadband event
window (0.1­2 keV). It contains 534 EUV sources, of which 166 were not previously discovered.
Of these, 105 have been identified and 77% of them are late­type stars. White dwarfs and early­
type stars make up only #14% of the sources, and there are no extragalactic objects at all among
them.
One important outcome from the EUVE mission is the first Spectral Atlas in the EUV range
(Craig et al. 1997). The Atlas contains EUV spectra of 95 stars ranging from bright B stars to
M dwarfs, white dwarfs, and cataclysmic binaries.
5.4. UVS, HUT AND ORFEUS
One should note that some information about the EUV properties of stars and galaxies was
obtained with the Voyager Ultraviolet Spectrometers (UVS), and with the Hopkins Ultraviolet
Spectrometer (HUT) and ORFEUS (Orbiting and Retrievable Far and Extreme Ultraviolet
Spectrometer) flights. While the UVS instruments use non­imaging optics to record the EUV
spectra of bright objects, HUT and ORFEUS used imaging telescopes to study spectra in the
FUV range.
The generic UVS instrument installed on both Voyager spacecraft was described by Broadfoot
et al. (1977). It consists of a mechanically collimated objective grating spectrometer covering
the range 50­170 nm with 1 nm resolution and was included in the Voyager missions primarily
to study the composition and structure of the atmospheres of giant planets and their satellites.
Due to the long cruise periods between planetary encounters, the Voyager UVS instruments
were uniquely able to make deep FUV observations of stars (e.g., Chavez et al. 1995), galaxies
(Alloin et al. 1995), and the interstellar medium (Lallement 1993).
HUT is a 0.9­meter telescope which operated from the Space Shuttle bay while mounted on
the instrument pointer of ASTRO, alongside the UIT and WUPPE telescopes. A description
of the instrument and its calibration was published by Davidsen et al. (1992). Briefly, HUT
performed spectroscopy in the 91.2­185 nm band with a resolution of #0.3 nm, and in the
second order, in the 41.5­91.2 nm band with #0.15 nm resolution. The e#ective area of HUT
ranged up to 10 cm 2 for the ASTRO­1 flight, and up to 25 cm 2 for the second flight. Note that

24 NOAH BROSCH
HUT su#ered a steady decrease of sensitivity at short wavelengths throughout the flight; a 25%
reduction at 80 nm from flight start to close to the end of the ASTRO­2 flight. The results from
the two flights of HUT were reported in some 60 papers, ranging from the Moon's UV emission
(Henry et al. 1995) to a search for the signature of decaying heavy neutrinos (Davidsen et al.
1991).
ORFEUS flew twice on the ASTRO­SPAS platform with a 1.0­meter telescope and spectro­
graphs covering the range 40­115 nm (R=3000) and 90­125 nm (R=10,000). The same ASTRO­
SPAS platform carried the Interstellar Medium Absorption Profile Spectrograph (IMAPS; Prince­
ton University). In this instrument, a mechanical collimator restricts the FOV to #1 # and feeds
an echelle spectrograph with R#120,000 in the range 95 to 115 nm. This is intended for the study
of ISM absorption profiles in the spectra of bright stars. Results from the IMAPS experiment
were reported by Jenkins et al. (1996).
OREFUS had #5 cm 2 e#ective area in the 40­90 nm band and #9 cm 2 in the 90­120 nm.
Its main results, from the first flight, include the discovery of the S and P elements in the
photospheres of two white dwarfs (Vennes et al. 1996), various studies of coronal gas (e.g.,
Hurwitz et al. 1995; Hurwitz & Bowyer 1996) and an analysis of the DO white dwarf HD149449B
(Napiwotzki et al. 1995). An interesting result was also the analysis of the O3If star HD93129A
(Taresch et al. 1997); this object has a ZAMS mass of 120 M# . The ORFEUS results from the
second flight are being prepared for publication in a special Astrophys. J. Letters issue.
6. Modeling the UV sky
In parallel with observations by rocket and satellite instruments, attempts to simulate the UV
sky were made, e.g., by Henry (1977). He used a transformation from optical to UV based
on Apollo 17 measurements of bright, early­type stars and data from other experiments for
cooler stars at 148.2 nm. Data for other spectral regions were based on model atmospheres from
Kurucz et al. (1974). The transformations used are described in Henry et al. (1975) and are
essentially linear relationships between the photon flux and the (B­V) color of a star for all UV
bands longward of Ly#. Henry (1977) calculated that to approximate the total UV starlight it
is not necessary to observe very faint stars; most of the UV light originates in relatively bright
(apparent) stars because of the influence of the interstellar extinction.
Gondhalekar & Wilson (1975) used a simple model for the distribution of stars (plane­parallel
distributions of di#erent scale heights for various types of O, B and A stars) to calculate the
interstellar radiation field between 91.2 nm and 274 nm. Gondhalekar (1990) used this model to
calculate the integrated UV emission by starlight in the Galaxy. The UV observations of TD­1
were used along with model atmospheres to derive stellar properties in the UV. These were
combined with parameters of interstellar dust and the spectrum of the di#use UV background
was derived. This was then compared succesfully with measurements by Kurt & Sunyaev (1968),
Hayakawa et al. (1969) and Henry (1972).
Two UV sky models were published in the last decade, by Brosch (1991) and by Cohen
(1994), later adopted for the FAUST bandpass by Cohen et al. (1995). The Cohen (1994) model
consists of a complicated model of the galaxy, with a disk, a halo, the bulge, spiral arms and two
local spurs, as well as the Gould belt. The model of Brosch (1991) is a much simpler adaptation
of the Bahcall & Soneira (1980) disk­bulge galaxy model, to which the Gould belt was added,
along with a thick disk of white dwarfs, where the distribution followed the scale height of Boyle
(1989).
The two models use di#erent approaches in modeling the UV sky. While Cohen (1994) creates
absolute UV magnitudes for every class of objects used by the model, Brosch (1991) calculates

UV SKY SURVEYS 25
a transformation from visible to UV using observed properties of IUE standard stars. The
approach used by Cohen employs model atmospheres from Kurucz (1991) and integration across
the filter bandpasses to yield the required [UV­V] color indices. Later unpublished developments
of the Brosch (1991) transformations use Kurucz (1992) stellar atmosphere models to extend
the optical­to­UV transformation to later stellar types.
Cohen (1994) tested successfully his model against the TD­1 results and against those of
the S201 Apollo 16 Moon telescope measurements. In a later paper (Cohen et al. 1995) the
predictions of the model were tested against the detailed spectral distribution of stars in the
North Galactic Pole region identified by Brosch et al. (1996) in the FAUST observation.
Similar encouraging results were obtained by Brosch (1991) in comparisons with the S201
measurements. The visible­to­UV transformation, derived by Brosch (1991) from IUE spectra,
was later used by Bilenko (1995) to verify a method of determining the three­dimensional dis­
tribution of UV extinguishing ISM clouds. The method uses the transformation to predict the
apparent UV magnitude of a star, based on its known optical properties as listed in the Hippar­
cos Input Catalog (HIC). The expected UV magnitude is compared with that measured by TD­1
and the total UV extinction is determined. Knowledge of the distance modulus (from the HIC
data) for an entire stellar population in a given sky region, allows one to locate the extinguishing
dust clouds in 3D space.
In Figure 4 I show how well can one predict the UV magnitude of a star, when the only known
parameters are the V and B magnitudes. The comparison is done for stars in the direction of
Virgo, appearing on three images obtained by the FAUST experiment. The optical information
of magnitude and color is obtained from the Tycho data set.
4.0 6.0 8.0 10.0 12.0 14.0 16.0
FAUST measured UV magnitudes
4.0
6.0
8.0
10.0
12.0
14.0
16.0
Predicted
UV
magnitudes
Virgo region observed by FAUST
Predicted vs. measured UV magnitudes
Figure 4. Predicted vs. observed UV magnitudes at 165 nm for stars in the three FAUST fields on Virgo. The
transformation relies on IUE observed spectra of stars as well as on Kurucz (1992) model atmospheres.

26 NOAH BROSCH
The di#use UV sky background has been modeled by Murthy & Henry (1995). In this paper
there are references to previous models accounting for the UV background. Note though that
not all published simulations attempt to explain the faint UV sources. Extragalactic objects,
such as large or dwarf galaxies, AGNs, and QSOs, are not included in any simulation. Nor has
there ever been a satisfactory accounting for the patchiness of the interstellar dust in the models,
with the possible exception of a recent account by Witt et al. (1997), when analyzing the UV
observations of FAUST.
7. Comparison of Survey Missions
The various missions surveying the UV sky can be compared in terms of a ``power'' parameter #
introduced by Terebizh (1986) and used by Lipovetsky (1992) in a comparison of optical surveys:
#
=# 4# 10 0.6#(m L-10) (5)
where# is the sky area covered by the survey (4# for TD­1) and mL is the limiting (monochro­
matic) magnitude of the survey. An all­sky survey to a limiting magnitude mL #8.5 (such as
done by TD­1), has the same ``survey power'' as a single HST WFPC­2 image exposed to show
21st mag UV objects. The di#erent UV and EUV missions discussed here are compared in Table
3 in terms of their survey power.
A parameter similar to #, to compare missions whose goal was to estimate the di#use UV
background, was introduced by Henry (1982):
S = 100
[A# ##] (6)
where A is the collecting aperture of the
experiment,# is the solid angle of its field of view, and
## is the spectral bandpass of the observation. Henry compared a number of experiments and
concluded that the Voyager UVS had the largest S­parameter.
In order to evaluate the UIT performance as a survey instrument for point sources, I assume
that the ASTRO­1 flight imaged 66 fields and ASTRO­2 flight another 100 fields, and that
the depth of the UIT survey was mL =20. For the di#use background case, I take the spectral
bandwidth to be 100 nm. Among the ``future missions'' categories I assume that GIMI will
ultimately conduct a full sky survey. In comparing TAUVEX and the UV/Optical monitor of
XMM, I assume the same mission duration and mode of operation for SRG and XMM. For HST,
I assume that 1000 independent WFPC­2 pointings with the F160BW filter, 1000 with STIS
FUV MAMA and SrF 2 and 1000 pointings with the ACS/SBC (see below) could be assigned
under a parallel imaging program. The sensitivity of the ACS/SBC is assumed to be slightly
higher than that of the STIS FUV MAMA, because of ther educed number of reflections.
For MSX/UVISI the sensitivity is that of the high resolution imager. A full analysis of this
data set is not yet available. For the purpose of comparison, I assumed that UVISI performed 400
independent pointings, yielding a sky coverage of 400 degrees 2 with the high resolution images
and perhaps 4,000 degrees 2 with the wide field imager. The di#erence in the expected perfor­
mance of the XMM/OUVM and that of the SRG/TAUVEX lies in the multiple­telescope design
of TAUVEX, which allows pure­UV simultaneous observations in three bands, compounded by
its larger field of view. The large value of # for GALEX, despite the similar mL to that of
TAUVEX, results from the all­sky coverage. For GALEX, the comparison is done with the pub­
lished expectations for the all­sky survey phase, and the limiting magnitude for 200 nm has been
transformed from the AB system to 200 nm monochromatic magnitudes.

UV SKY SURVEYS 27
TABLE 3. UV and EUV survey missions
Mission
Year# (ster) mL # ## (nm) Nsources Notes
TD­1 1968­73 4# 8.8 0.19 150­280 31,215 1
S201 1972 0.96 11 0.30 125­160 6,266
WF­UVCAM 1983 1.02 9.3 0.03 193 ?
SCAP­2000 1985 1.88 13.5 18.9 200 241 2
GUV 1987 5 10 -3 14.5 0.2 156 52 Pointed phase
GSFC CAM 1987+ 0.03 16.3 14.4 242 #200 Virgo observation
FOCA 1990+ 0.02 19 377 200 #4,000 Estimated
UIT­1 1990 3.8 10 -4 17 0.48 #270 2,244 UIT catalog
GLAZAR 1990 4.4 10 -3 8.7 6 10 -4 164 489
FUVCAM 1991 0.09 10 7.5 10 -3 133, 178 1,252 3
FAUST 1992 0.33 13.5 3.3 165 4,698
UIT 1+2 1990, 95 1.3 10 -3 19 26 152­270 6,000 ? 4
HST WFPC 1990+ 4.3 10 -4 21.0 134.8 120­300 50,000 ? 5
HST STIS 1997+ 1.5 10 -5 23.8 222.7 120­200 10,000 ?
HST ACS/SBC 1999+ 2.4 10 -5 24.0 480 120­170 10,000 ?
MSX UVISI 1997­98 #0.01 18.0 50 180­300 ? 6
ARGOS GIMI 1998+ 4# 13.6 136 155 2.5 10 5 7
SRG TAUVEX 1999+ 0.06 19 1200 135­270 10 6 8
XMM OUVM 1999+ #6 10 -3 19 #100 185­600 10 5 ?
GALEX 2002+ 4# 19.4 4.4 10 6 130­300 2 10 7 9
WFC 1992 4# ­ ­ 10, 16 479
ALEXIS 1994+ 4# ­ ­ 13­19 50 ?
EUVE 1992 4# ­ ­ 7­70 734 10
Notes to Table 3:
1: The unpublished extended version has 58,012 sources.
2: 92 stars (Laget 1980) and 149 galaxies (Donas et al. 1987).
3: Only the Sag and Sco fields (Shuttle flights) included.
4: Assumes 66 pointings for ASTRO 1 and 100 for ASTRO 2.
5: Assumes 1000 observations with HST for each of the following combinations: WFPC­2+F160BW, STIS FUV
MAMA+SrF2 , and ACS/SBC.
6: Numbers are calculated for the high resolution imager; the wide field imager has a much smaller #. 7: Assumes
2â stars per magnitude w.r.t. TD­1.
8: Assumes 5000 independent pointings to end­of­life.
9: mL transformed from AB; data refers to the all­sky imaging survey (AIS) phase.
10: Number of sources in the 2nd EUVE catalog.
Because of the ``unfair'' comparison based on #, and because not all surveys cover the entire
sky, it may be more useful to look at another estimator, the density of sources detected (or which
are expected to be detected) by a certain experiment, shown in Figure 5. Expected results from
a hypothetical HST survey with the ACS/SBC have not been plotted in the figure, but they are
expected to be significantly deeper than those of GALEX DIS.

28 NOAH BROSCH
1970 1980 1990 2000 2010
Year
2.0
4.0
6.0
8.0
Logarithmic
source
density
(N/ster)
Progress in UV astronomy
Past, present, and future missions
GALEX DIS
GALEX AIS
TAUVEX
UIT-1
FOCA
GIMI
SCAP-2000
TD-1
Figure 5. Source density in the UV for various sky surveys. All­sky and partial surveys (past or future) are
included.
8. The Resultant Sky Picture
Any idea we may form of the appearance of the UV sky is necessarily limited by the existence
of a single all­sky survey in the UV (TD--1) and by the deeper surveys in limited regions of the
sky. Comparing the number of UV sources charted by TD--1 with similar mapping done in the
optical we are forced to conclude that the status of UV astronomy is now at par, in the number
of sources, with the Carte du Ciel catalogs from the end of the last century. The importance of
knowing the properties of a large number of sources is the astronomical analog of the biological
``diversity of species'' argument; the more species are identified, the higher is the chance of
discovering a new phenomenon. By skipping the stage of deep and extensive mapping of the sky
in the UV, the astronomers may be missing the discovery of some new types of sources.
I compared above the status of the UV sky knowledge in the post­TD­1 era to that which
prevailed in optical astronomy at the turn of the 20th century. In the same vein, our knowledge
of the EUV sky in the post­EUVE era can be compared to a naked­eye look at the sky on a
misty night; very few stars, only the brighter or nearer ones, can be seen. The present EUV
catalogs are at par, in the number of sources, with the catalogs of Hipparcos (127 BC), Tycho
Brahe (1601), and Hevelius (1660). If a full­sky survey could be conducted to the depth of the
right angle survey of EUVE, it would presumably record some 50,000 sources, about 100 times
fainter than those of the ''Second EUVE Source Catalog'' from the full­sky survey.
The picture resulting from the all­sky survey of the TD­1 mission, from the EUVE and
ROSAT WFC surveys, and from the limited surveys performed previously is that of a UV sky
dotted with many stellar and extragalactic sources, where the background is produced by natural
emissions (Ly#, other emission lines), and/or by UV light scattered o# dust clouds. In sharp

UV SKY SURVEYS 29
contrast, the opacity of the ISM is such that in the EUV range, only the more intense and/or
nearby sources are detectable.
In this section I discuss the accumulated knowledge on di#erent types of celestial objects and
on the di#use background, both for the UV and for the EUV ranges. The various sources have
been listed in Table 1. The main conclusions of this section are, for the UV, that (a) most of the
brighter UV sources are early­type stars, where the spectral type contributing mostly depends
on the UV band of the observation, (b) at faint 200 nm magnitudes most sources are galaxies,
(c) the UV background is mostly starlight from within the Milky Way, scattered by interstellar
dust, (d) the small fraction of extragalactic UV background can be explained by integrated
light from galaxies. For the EUV, the conclusions are that (a) most of the sources are late­type,
coronally­active stars, and (b) there are only very few extragalactic sources, presumably because
of the opacity of the ISM and the intrinsic source opacity for EUV photons.
8.1. WHAT IS KNOWN ABOUT STARS ?
The various experiments construct a picture in which most of the stars detected by TD­1,
FAUST, SCAP and FOCA are relatively early­type B, A and F. However, most of the stars
included in the UIT catalog are probably late­type (G and later). Specifically for TD­1 (Thomp­
son et al. 1978) the spectral classes B, A, and F make up 95.9% of all the stars in the published
version of the catalog. Carnochan & Wilson (1983) found many TD­1 stars with more intense
UV emission than B8 stars (m 156.5 ­m 274 #--1.30). They identified these objects as unreddened
subdwarfs, with a scale height in the Milky Way similar to that of the central stars of planetary
nebulae.
I show in Tables 4, 5, 6, and 7 a comparison of the TD­1 spectral distribution with that from
the FAUST observations of the North Galactic Pole, Virgo, and Coma regions. The column
labeled HES indicates the number of hot evolved stars (horizontal branch, subdwarfs, white
dwarfs) in each magnitude bin. Note that while the FAUST statistics refer to the sources detected
at 165 nm, the TD­1 values are for sources fulfilling the selection criteria for catalog inclusion,
in particular a S/N#10 at 156.5 nm.
TABLE 4. UV sources in the TD­1 catalog
UV mag O0--O9 B0--B9 A0--A9 F0--F9 G0­K9 M0--M9 Other Total
# 6.9 34 2132 2998 1886 1033 19 2 8104
7.0--7.9 22 2411 4397 2388 41 0 10 9269
8.0--8.9 24 3185 6260 611 13 1 37 10131
>9.0 17 1757 1928 12 5 0 115 3834
Total 97 9485 15583 4897 995 20 138 31338
A comparison of the tables shows that at high |b| there are few B stars, thus the A and F
types dominate the source counts in the FAUST data. The Coma region shows an unexpectedly
high number of hot evolved stars; these are probably related to the high galactic latitude open
cluster Mel 111 (Brosch et al. 1998).
Except for the FAUST fields studied at Tel Aviv University (Brosch et al. 1995, 1998), most
surveys used exclusively correlations with existing catalogs to identify sources. In the fields of
NGP­NB and Virgo, where the reduction and identification processes are complete, we find

30 NOAH BROSCH
TABLE 5. UV sources in the North Galactic Pole region (l#281, b#84)
UV mag B0--B9 A0--A9 F0--F9 G0­K9 HES AGN/galaxies Total
# 6.9 0 3 0 0 1 0 4
7.0--7.9 0 1 0 0 1 0 2
8.0--8.9 0 2 1 0 1 0 4
9.0--9.9 1 3 3 1 0 0 8
10.0--10.9 0 4 5 0 3 2 14
11.0--11.9 1 5 5 1 2 3 17
12.0--12.9 1 2 7 3 2 2 17
13.0--13.9 1 3 5 3 0 1 13
Total 4 23 26 8 10 8 79
TABLE 6. UV sources in the Virgo region (l#279, b#77)
mUV B0--B9 A0--A9 F0--F9 G0--G9 HES AGN/Galaxies No ID No Sp Total
5.0--5.9 0 2 0 0 0 0 0 0 2
6.0--6.9 0 5 0 0 1 0 0 0 6
7.0--7.9 1 3 0 0 0 0 0 0 4
8.0--8.9 0 4 0 1 1 0 0 0 6
9.0--9.9 2 5 2 0 0 0 1 0 10
10.0--10.9 3 2 2 0 2 6 2 0 17
11.0--11.9 3 6 14 0 2 16 2 0 43
12.0--12.9 4 8 8 1 0 31 3 2 57
13.0--13.9 2 5 6 0 1 20 6 0 40
14.0--14.9 0 2 0 0 0 3 0 0 5
#15.0 0 1 0 0 0 0 0 0 1
Total 15 43 32 2 7 76 14 2 191
almost equal fractions of A­F stars (70 and 75%). The only disturbing fact is that we identify
no white dwarfs in the three Virgo fields, whereas about 7 are expected; it is possible that these
hide among the small fraction of unidentified sources, or that they were identified as ``normal''
stars. The Coma field includes the open cluster Mel 111; for this reason it is not typical of other
high­b fields.
In the EUV range it is possible to compare the di#erent findings of the survey instruments
described in the relevant section, with the exception of ALEXIS, whose sources have not yet
been collated into a table. I show the ROSAT WFC and EUV lists in Table 8. It is clear that
the large majority of EUV sources mapped by the di#erent experiments are late­type stars; the
number of extragalactic objects is extremely limited.
8.2. WHAT IS KNOWN ABOUT GALAXIES ?
The information about galaxies is very sparse and we still lack a large sample of a few 1000's
galaxies, from which to perform good statistical studies.

UV SKY SURVEYS 31
TABLE 7. UV sources in the Coma region (l#60, b#88)
mUV B0--B9 A0--A9 F0--F9 G0--G9 K0--M9 HES Total
<7 0 1 0 0 0 0 1
7.0­7.9 0 1 0 0 0 0 1
8.0­8.9 0 1 0 0 0 1 2
9.0­9.9 1 1 0 0 0 2 4
10.0­10.9 0 4 3 1 0 3 11
11.0­11.9 0 5 7 1 0 2 15
12.0­12.9 0 5 2 1 0 3 11
13.0­13.9 0 0 0 0 1 0 1
Total 1 18 12 3 1 11 46
TABLE 8. Nature of identified EUV sources
Source WFC WFC EUVE EUVE EUVE EUVE
nature BSC 2RE BSL 1st 2nd ROSAT
AGN, QSO 7 18 10 9 6 0
XRB, CVs 20 10 14 16 15 5
O­A­B stars 8 8 13 32 18 43
F­G­K­M stars 181 251 172 184 161 411
WDs (PNN) 119 140 117 109 98 27
Total 335 421 326 350 298 486
Significant information on selected objects, mainly on stellar populations and the nature of
the ISM, was obtained from IUE spectra (O'Connell 1992). Similar observational data, combining
UV with optical and near­IR spectrophotometry through matched apertures, was recently used
to derive template spectral energy distributions (SEDs) for various types of galaxies (Storchi­
Bergmann et al. 1994; McQuade et al. 1995). A combination of IUE observations and data from
other UV imaging missions was used to extract ``total'' UV information on galaxies (Longo et
al. 1991; Rifatto et al. 1995a, 1995b).
There is hope to derive the star­forming histories of galaxies through a combination of data
from the UV to the near­IR, in the manner of the Storchi­Bergmann et al. (1994) templates.
However, we have recently shown (Almoznino & Brosch 1997a, 1997b) that such decompositions
are not unique, at least for a sample of blue compact dwarf galaxies (BCDs). In order to ac­
count for recent star­formation bursts, it is necessary to include information about the ionizing
continuum; as the ISM in the Milky Way and in the target object prevents direct observation
of the ionizing continuum, this can be derived from H# observations under simplifying assump­
tions. UV data collected by IUE or UIT are usually at # >140 nm. This region contains mainly
radiation from A­type stars and requires extrapolation of the stellar population to earlier types,
to account for Lyman continuum photons.
To understand large populations of galaxies, in terms of stellar populations and star formation
histories, it is impractical to rely only on the detailed modeling of spectral features in the optical
region. One should combine information from many spectral bands, covering a spectral region

32 NOAH BROSCH
as wide as possible.
In the absence of very deep UV surveys in more than a single spectral band, such as those ex­
pected to result from the UIT exposures, our information about a significant number of galaxies
originates from the SCAP­2000 (Donas et al. 1987) and FOCA (Milliard et al. 1992) measure­
ments. These consist of integrated photometry at 200 nm of a few hundred galaxies. In the 200
nm band and in the brightness range 16.5­18.5 galaxies apparently dominate the source counts.
The corresponding blue magnitudes of these galaxies are B=18--20 and their typical color index
is [200 - V ] # -1.5. Comparing this index with the template spectra of Kinney et al. (1996),
the FOCA galaxies fit the SB2 template, i.e., a slightly reddened starburst galaxy.
The galaxy counts from the FOCA flights originate from the analysis of three high latitude
fields covering about 4.5 square degrees, which include galaxy or globular clusters (M3, Abell
2111, and SA 57, which contains a few Coma cluster galaxies). Milliard et al. (1992) indicate
negligible contamination of their galaxy counts by either UV stars or cluster galaxies. In order
to reproduce the observed galaxy density distribution, Armand and Milliard (1994) find that
a simple transformation to the UV of the known visible galaxy counts is not su#cient. They
require a larger contribution by star­forming galaxies in the recent past, such as late­type spirals
or blue dwarf galaxies.
In the field of A2111, Milliard et al. (1996, private communication) find that the UV galaxies
are mostly in the foreground or the background of the cluster, ranging as far as a redshift of 0.68.
Note that the Lyman break in the rest frame of a distant galaxy will enter the FOCA window
only at z#1.3. Spectra of some of these optically faint galaxies reveal narrow emission lines,
justifying a link with star­forming galaxies. A larger fraction of starforming galaxies at high z
was claimed also from HST data, i.e., the HDF or the MDS. However, note that Giavalisco et al.
(1996) claim that not all HST faint galaxies are starbursts or peculiar galaxies; being observed
at UV wavelengths in their rest frame their morphological appearance is later than would be
determined from optical studies.
Using the ``field'' galaxy luminosity function in the UV, derived by Deharveng et al. (1994)
from the smooth linking of the projected density of UV galaxies from the FAUST and FOCA
counts, it appears that:
logN(m) = 0.625 âm 200 - 9.5 (7)
There is a faster increase in the number density of galaxies at faint UV magnitudes than it is
for stars.
The UIT results, derived from the analysis of the 48 independent pointings from the ASTRO­
1 mission, indicate that the majority of UV sources recorded by this instrument are stars (Smith
et al. 1996). This is at odds with the analysis of Milliard et al. (1992). One possible explanation
could be the longer wavelength response of UIT; most of the analyzed images were taken through
the A1 filter. This filter has a significant color term because of its considerable width, thus its
e#ective central wavelength is 230 nm for early­type stars and about 280 nm for G stars. In
contrast, the FOCA bandpass is much narrower and better defined. The di#erence in the amount
of recorded stars could then be the result of UIT seeing more late­type stars, to which FOCA
would be ``blind''.
To check this possibility, we calculated models of stellar densities patterned after those of
Brosch (1991), with the di#erence that the present models were calculated for the UIT A1 band.
Specifically, we derived an optical­to­UV transformation based on IUE spectra convolved with
the A1 filter response, shown in Fig. 6, calculated the cumulative stellar densities in the direction
of UIT targets, and compared the expected and the measured source counts and properties, in
the deepest UIT exposures (all with the A1 filter and with 600 sec exposure or longer). We found
that we can reproduce the total counts, the distribution of star counts by magnitude bins, and

UV SKY SURVEYS 33
the [UIT--V] color distribution with a model cuto# at V#18. Results for the UIT field centered
at (06:22 ; --13:03) are shown in Figures 7 and 8, and the error bars on the experimental points
originate from an assumption of Poisson statistics in the number counts. It follows that UIT
could indeed have detected mostly stars, and that the Galaxy model in Brosch (1991), with
its subsequent modifications and transformation to the UIT A1 band, reproduces faithfully the
observed stellar distribution.
­0.5 0.0 0.5 1.0 1.5
B­V
­5.0
0.0
5.0
10.0
UV
­
B
UIT A1
LUM V
Figure 6. Derived transformation between observed B­V of a star and its UV--B color for main sequence stars.
The UV magnitude is derived by convolving the UV spectrum (from the IUE observations) with the transmission
of the A1 filter of UIT.
One must be wary of the UV--V colors quoted in the UIT catalog; the V magnitude is derived
in many cases from the ``quick­V'' magnitude in the HST Guide Star Catalog, which may be
significantly o# (0.15 mag=1# for objects near the plate center calibrating sequences, going as
high as 0.30 mag=1# for objects far from the sequence: Russell et al. 1990).
Another comparison of UIT and FOCA performance is possible using the studies of M51 by
both instruments (Bersier et al. 1994, Petit et al. 1996 for FOCA; Hill et al. 1996 for UIT). I
measured the integrated flux of 22 HII regions in M51 on the calibrated FOCA image (kindly
provided by D. Bersier), out of the 28 identified in the UIT B1 image of Hill et al. (1995). The
comparison is shown in Figure 9. In general, the m 200 and the m 152 magnitudes correlate well.
There is no significant calibration o#set between UIT and FOCA, at least in this image, and
the HII regions do not show strong reddening e#ects between 152 nm and 200 nm. This nice
correspondence demonstrates fully the advantage of a balloon­borne UV survey telescope, at a
fraction of the cost of UIT or of other orbital telescopes (see below).
One missing ingredient to a fuller model of the sky consists of a proper representation of
galaxies and AGNs/QSOs. The latter, in particular, are UV­bright objects. Their detection in

34 NOAH BROSCH
8.0 10.0 12.0 14.0 16.0 18.0
UV magnitude (UIT­A1)
­5
5
15
25
35
45
55
65
75
Stars
in
UIT
field/magnitude
06:22­13:03 (l=221, b=­12)
UV sky model vs. UIT star counts
Figure 7. Predicted vs. observed UV star counts at the UIT A1 band, for a field at (06 h 22 m : ­13 # 03'). In general,
and up to mUV #16.5, the model predictions are within one standard deviation of the real star counts. For fainter
stars there is progressive incompleteness.
deep UV surveys should be relatively easy. Moreover, their spectral energy distribution is very
di#erent from that of stars, thus using UV and optical colors it is possible to discriminate them
from the foreground stars. The claim by the FOCA group of a large contribution of UV galaxies
is supported by theoretical arguments requiring a fast­evolving population of perhaps dwarf
galaxies for z=0.2­1.0, in order to explain the faint source counts in other spectral domains
(Ellis 1997). A similar possibility, that #60 of the almost star­like blue objects in the Hubble
Deep Field are distant extragalactic sources, has been proposed by M’endez et al. (1996). One
possibility could be that these sources are distant blue dwarf galaxies or starbursting galaxies,
which are smaller than their present­day counterparts. This is apparently confirmed by the Keck
spectra of some of these objects, though the inferred star formation activity in these objects
appears much higher than in local starbursts (e.g., Guzman et al. 1997).
8.3. WHAT IS KNOWN ABOUT THE ISM ?
Studies by UIT and FAUST emphasize the relative importance of the dust in understanding the
UV emission. Bilenko (1995) analyzed the TD­1 catalog and a version of the Hipparcos Input
Catalog transformed to the TD­1 band with information derived from IUE spectra of di#erent
types of stars. They showed that the UV exinction is very patchy, with very di#erent values of
extinction per kpc on scales smaller than 10 # .
Tovmassian et al. (1996b) used the results of GLAZAR observations of a 12 degrees 2 area
in Crux, to establish that the distribution of ISM dust is very patchy, with most of the space
between stellar associations being relatively free of dust. They did not attempt to delimit better

UV SKY SURVEYS 35
­5.0 0.0 5.0 10.0
(UIT­A1)­B
­10
10
30
50
70
90
110
N(stars)
in
UIT
field/color
bin
06:22­13:03 (l=221, b=­12)
Figure 8. Predicted (solid line) vs. observed (points with error bars) UV star color distribution, for the UIT
field at (06 h 22 m ; ­13 # 03'). As in Fig. 7, most points follow the theoretical color distribution and are within one
standard deviation of it. I have no explanation for the outlier point at UV--B=1.5.
the location of dust clouds in 3D space, because of the rather sparse sky coverage of the GLAZAR
stars.
Much of our present understanding of the structure of the local ISM comes from studies in
the EUV range. In particular, the various EUVE catalogs (Bowyer et al. 1994, 1996; Lampton
et al. 1996) confirm the previously known features of the local ISM (a ``tunnel'' to CMa with
very low HI column density to 200 pc. and close to the Galactic plane first identified by Gry
et al. (1985) from the COPERNICUS data, a cavity connected with the Gum Nebula in Vela,
a shorter 100 pc. tunnel to 36 Lynx, and the very clear region in the direction of the Lockman
hole). It is clear that the present EUV catalogs are not deep enough to reveal finer, or deeper
details.
The new UIT study of M51 shows that A #
E(B-V )
changes with decreasing metallicity (or
galactocentric distance). This is reminiscent of the finding by Kiszkurno­Koziej & Lequeux
(1987), that the extinction law in the Milky Way changes with distance away from the galactic
plane. The authors find that the H# flux is depleted in the inner regions of M51; this they
interpret as increased Lyman continuum extinction in the inner parts of M51.
8.4. WHAT IS KNOWN ABOUT THE UV AND EUV BACKGROUNDS ?
The accurate measurement of the UV sky background (UVB), with the expectation that it could
put meaningful cosmological limits, has been the goal of many rocket, orbital, and deep space
experiments. Many observational results were summarized in reviews by Henry (1982), Bowyer
(1990), Bowyer (1991), and Henry (1991). It is worth noting the many pitfalls of accurate UV

36 NOAH BROSCH
13.0 14.0 15.0 16.0 17.0
UIT (156 nm) mag.
14.0
15.0
16.0
17.0
18.0
FOCA
(200
nm)
mag.
M51 HII regions
Comparison of two UV magnitudes
Figure 9. Comparison of UIT and FOCA UV photometry of the same regions of M51.
background measurements: sensitivity to instrumental dark current, atmospheric emissions at
low orbital and rocket altitudes, etc. Leinert et al. (1998) discuss observations and interpretation
relevant to the UV and EUV backgrounds in a general discussion of the sky background in many
wavelengths.
Bowyer (1991) separates the various origins of the di#use UVB into ``galactic'' and ``high
latitude''. The latter forms an #uniform pedestal, onto which the galactic component is added
in various amounts depending on the direction of observation. The ``galactic'' component can
be #one order of magnitude more intense than the ``high latitude'' one. Most of this intensity
is probably from light scattered o# dust particles in the ISM, and the rest is from the gaseous
component of the ISM (HII two photon emission and H 2 fluorescense in molecular clouds).
The majority of the ``high latitude'' component is probably also galactic, originating from light
scattering o# dust clouds at high galactic latitude.
An analysis of the di#use UV emission (Murthy & Henry 1995) indicates that the extra­
galactic component of the UV background can be at most 100­400 photon units. Whenever the
column density of HI is larger than 2 10 20 cm -2 , the main contributor to the UVB originates
from dust­scattered starlight. The low levels of the extragalactic component are produced prob­
ably by the integrated light of galaxies (Armand et al. 1994), or the integrated light of the
Milky Way scattered o# dust grains in the Galactic halo (Hurwitz et al. 1991). Even the Lyman
# clouds in intergalactic space may contribute, through their recombination radiation (Henry
1991).
The confirmation that most UVB originates from dust­scattered light was obtained by
Sasseen et al. (1995) and Sasseen & Deharveng (1996). Sasseen et al. showed that the spa­
tial power spectrum of the di#use UVB on FAUST images is consistent in spectral index and

UV SKY SURVEYS 37
amplitude ratio with the IRAS 100 µm signal in the same sky areas, while Sasseen & Dehar­
veng found that the UV background correlates with the FIR measurements of COBE/DIRBE.
Galactic cirrus clouds apparently scatter back significant FUV radiation from the Milky Way;
Harkala et al. (1995) found such UV emission with FAUST from the direction of a dark cloud
detected at 100 µm by IRAS.
The low levels of UV background, away from orbital and galactic contaminants, have been
confirmed by an analysis of 489 UIT images from the ASTRO­1 flight (Waller et al. 1995). After
correcting for orbital background and zodiacal light, and after accounting for scattered Galactic
light by ISM cirrus clouds (through the IRAS 100 micron emission), the authors extrapolated
the UV­to­FIR correlation to negligible FIR emission to find that the extragalactic (cosmic)
UV background must be 200±100 count units. Note one e#ect in all Shuttle­based experiments,
which was not accounted for in the UIT UVB analysis; the influence of background light produced
by the Shuttle attitude jets firing. During the FAUST flight this was observed as an enhanced
background in the photon stream, which was # 4â higher than the level from all other sources
(Sasseen 1996, private communication). As this is di#use light, the influence is clearly dependent
on the solid angle of observation and on the size of the aperture. For UIT, this should have
resulted in a level #2.5% that of FAUST. Because of this, the minimal UVB levels estimated by
Waller et al. (1995) must be considered upper limits.
0 2000 4000 6000 8000
Wavelength
28.0
26.0
24.0
22.0
20.0
Sky
brightness
(mag./square
arcsec)
Sky background
Ground vs. orbit
Space
Ground
Figure 10. UV and optical sky background from orbit vs. ground observatories. This is a revised version of a
similar figure by O'Oconnell (1987), where the values related to the optical domain have been retained.
The shorter wavelength background has been observed by e.g., Holberg (1986) with the
Voyager UVS, reaching as far down as 50 nm. The UV range shortward of Ly# but longward of
the Lyman break was studied from an integrated data set of all the Voyager UVS observations
(Murthy et al. 1996). They identified 272 UVS observations of which #50% have intensities

38 NOAH BROSCH
under 100 count units. In particular, note the observations on the North Galactic Pole region,
which give consistently a zero background level (upper limit 100 c.u., 1#­to be revised). Murthy
et al. concluded, after accounting for the scattered light in the Ly# and Ly# line wings and
for the radiation­induced background from the radio­isotope thermal generator on­board the
Voyager spacecrafts, that the minimal sky background between 91.2 nm and 115 nm is 0±100
c.u. (1#). This is, by far, the lowest value of any extragalactic UVB component, and the formal
1# upper limit is the value plotted in Fig. 10. Its implication is that in the nearby Universe, to
z#0.2, there is only a very small contribution of photons with # >912 š A . It is also clear that
galaxies at any redshift are rather opaque to Lyman continuum photons; these could contribute
in the Voyager band from any (reasonable) redshift.
It is possible to evaluate the level of the ``true'' extragalactic background which would be
detected by instruments with moderate angular resolution, from the FOCA galaxy counts at
200 nm (Milliard et al. 1992). Their actual galaxy counts, for 15.0# m 200 #18.5, give (when
extrapolated to m 200 =20.0) a contribution of #100 c.u.'s already from the UV­detected galaxies.
Note that such faint objects would hardly be detectable in the UIT exposures. Armand et al.
(1994) used the FOCA counts at 200 nm, with a limiting magnitude of 18.5, to calculate the
contribution to the UV background from galaxies. They estimated that 40--130 c.u. can be due
to galaxies. A similar analysis by Martin (1997), using FOCA galaxies, indicates that at least
25% of the UV background can be due to galaxies.
Treyer (1997, private communication) calculated the expected UV background due to galax­
ies. This was done by integrating the present­day 200 nm luminosity function, obtained from
the FOCA observations, to high redshifts. Under the assumption of ``no evolution'', she finds
that the di#use 200 nm surface brightness should be # 10 -9 erg s -1 cm -2 š A -1 ster -1 , which is
equivalent to a surface magnitude # 200 #28 mag arcsec -2 , or #100 c.u. This must be considered
a lower limit to the background produced by galaxies, as evolutionary considerations, such as
having more star formation at high redshifts (e.g., Madau et al. 1996), would increase the value
of the EUB.
Any reasonable spectral energy distribution one would invoke for the FOCA UV galaxies
should have a SED rising into the UV, thus even higher flux levels are expected in the 100
nm region than at 200 nm where FOCA observes. The Voyager results of Murthy et al. are in
strong conflict with the assumption that the di#erential count numbers of UV galaxies extend
unmodified to mUV =24, and in mild conflict (2#) with the assumption that they extend to
mUV =20. In the latter case, the entire extragalactic UVB could be due to unresolved galaxies,
leaving no room for hydrogen­recombination photons from the Ly# clouds, photons scattered o#
extragalactic dust clouds, etc. The result of Treyer presented above indicates a di#culty with
the longer wavelength UVB as well; the contribution by galaxies accounts for almost the entire
UVB measured by UIT !
A very deep EUVE spectroscopic observation of a large region on the ecliptic has recently
been reported (Jelinsky et al. 1995). It indicates that the only emission lines observed are He
I and He II (58.4, 53.7, and 30.4 nm), which originate from scatted Sunlight by the geocoronal
and/or interplanetary medium. The spectrum of the EUV background (Fig. 2 in Jelinsky et
al.) shows a tantalizing continuum, which decreases from 800 c.u. at 16 nm to 100 c.u. at 35
nm. This is probably grating scattering of the 58.4 and 30.4 nm helium lines (Jelinsky, private
communication). It is expected that the knowledge of the EUV background will improve, as more
data from the full ecliptic survey (3 10 6 sec exposure) will be processed. The lowest value for
the EUV background originates from a ROSAT observation (Barber et al. 1996). The detection
of the shadow cast by the nearby galaxy NGC 55 on the di#use 250 eV (49.6 nm) background
sets a level of the EUV background coming from extragalactic sources at 29.4±7.2 keV cm -2

UV SKY SURVEYS 39
sec -1 keV -1 (#9.5±2.3 c.u. at 49.6 nm).
9. Future missions
The future development of UV Astronomy, at least in the context of a full sky survey, has
been influenced strongly by the decisions of two major committees in the US. The first is the
``Decade Survey of Astronomy and Astrophysics for the 1990s'', a committee formed in 1989 by
the National Research Council in the US and chaired by John Bahcall to set priorities in an
age of diminishing funds. The workings of the Decade Survey committee have been described
by Bahcall (1991) and the report itself was published in 1991.
The Bahcall committee attempted to prioritize by ``democratic consultation'' throughout
the astronomical community. It is interesting that among the large and moderate programs
recommended by the Bahcall committee there is only one program in the UV domain (FUSE, see
below), and the emphasis is on infrared and optical programs. The sociological trend continued
well into the small programs, which the committee did not specify beyond a few examples; none
of these relate to UV astronomy.
The second very influencial committee was charted by the Association of Universities for
Research in Astronomy (AURA) with support from NASA, and is called ``HST and Beyond''.
The committee was formed in 1993, was chaired by Alan Dressler, and produced its report in
1995 (Dressler 1996). The Dressler committee identified two goals for the Space Astronomy of
the early­21st century: the detailed study of the birth and evolution of normal galaxies, and the
detection of Earth­like planets around other stars and the search for life on them. The goals
identified by the Dressler committee were to be realized by the Next Generation Space Telescope
(NGST), by a space interferometer, and by extending the operation of the HST beyond 2005. In
particular, note that HST was to be the major instrument for UV astronomy beyond that date.
The decisions of the Bahcall and Dressler committees have been adopted by NASA and by
the NRC, which control the major funding in the US. In particular, NASA headquarters has
been restructured into four major ``themes'', of which the two most related to space astronomy
are ``Structure and Evolution of the Universe'', and ``Search for Origins''. It is possible that these
decisions have also had far­flung influences beyond the borders of the US, and that other Space
Agencies are prioritizing their programs in a hidden race against the US Space Program. This
could have had the e#ect of blocking UV astronomy, at least in the direction of conducting a
new all­sky survey, by the major players in space astronomy.
In the immediate future, only one space mission (GIMI, see below) has the potential of
yielding results relevant to the derivation of a full UV sky survey. GIMI will reach only the
relatively brighter objects, an extension of a few magnitudes below the TD­1 limit. With the
exception of nearby galaxies, no extragalactic objects are expected to be measured by it. In
addition, there is the FUSE mission, which will hopefully begin to operate within 1998.
Given these rather pessimistic constraints, it is gratifying that NASA selected in late­1997
the GALEX mission (see below), which was mentioned above in the context of a comparison
of surveys. GALEX will conduct an all­sky UV survey, comparable to that of the Palomar Sky
Survey in the optical domain. The results from GALEX will become available in the second half
of the next decade.
Before describing the future UV missions, I review a few examples of missions which did not
succeed to launch.

40 NOAH BROSCH
9.1. POSSIBLE MISSIONS
A number of missions to map the sky to depths comparable with the Palomar Sky Survey
(PSS) in the optical have been proposed, but none bar one was accepted by the suitable funding
agency. An early attempt was to create a geo­synchronous updated UIT (Largefield Ultraviolet
Explorer=LUX), proposed in 1986. Another proposed attempt was to have a SMEX for a UV
imaging and spectroscopy all­sky survey (PAX, JUNO), then a MIDEX for a similar purpose
(MUSIC). These proposals, although very good, did not survive the NASA selection. A third
attempt, GALEX (from the same PI as PAX, JUNO, and MUSIC), was finally selected in 1997
for a 2001 launch.
A mission to study the di#use UV background (HUBE) was retained by NASA as backup
for a selected mission. A Canadian group proposed to study a mission similar to JUNO for a
National Canadian small satellite. Some of its members were among the initiators of STARLAB,
a one­meter class Space Schmidt proposed for UV surveys, which was scrapped in the late­80s.
It is possible that the Canadian option will be to fly on a Shuttle­carried long duration platform.
A conceptual design for a very ambitious survey telescope, dedicated to UV astronomy,
was presented by the Byurakan Observatory (Armenia) and by the Astronomical Institute of
Potsdam, Germany (Tovmassian et al. 1991d). The Astrophysical Schmidt Orbital Telescope
(ASCHOT) would be able to image a 2 # .5 field of view with 2'' resolution, using an all­reflecting
Schmidt design with an 80 cm aperture. The calculated performance indicated a detection limit
with S/N=10 at m 150 =24.0 in a 30 min exposure. However, there are formidable aspects in
dealing with a 9000â9000 pixel array. The primary mirror (diameter 120 cm) and the reflective
corrector plate (diameter 80 cm) of ASCHOT have been fabricated by Carl Zeiss Jena prior
to the re­unification of Germany. The realization of the ASCHOT project remains a possibility
worth exploring.
In Russia Boyarchuck and collaborators are building Spectrum UV (SUV), an HST­like
mission intended for deep spectroscopy in the UV of selected targets. Note that this mission, third
in line among the SPECTRUM spacecraft, will not be a survey instrument but will be dedicated
to studies of known celestial objects. The SUV mission will be almost exclusively dedicated to
intermediate resolution spectroscopy, but its focal plane camera has imaging possibilities similar
to those of the HST ACS (B. Shustov, 1997 private communication). Also, given the delays
in launching the first Spectrum mission (Spectrum X­#), it is doubtful whether SUV will be
realized before the year 2000.
9.2. REAL MISSIONS
The one space experiment constructed exclusively for UV astronomy is FUSE (the Far Ultraviolet
Spectroscopic Explorer mission), manifested for a late­1998 launch. This satellite is desiged to
provide high spectral resolution observations ( #
##
=30,000) in the band 90.5 to 119.5 nm. FUSE
is designed for a three year life in low Earth orbit and has significant e#ective area peaking at
#100 cm 2 near 105 nm. This allows high resolution spectroscopy with S/N=20 of objects with
monochromatic magnitude 12 in 100,000 seconds of observation, which is equivalent to about
3.3 days, given the expected operating e#ciency. Most of the observing time will be dedicated
to a number of key program, such as the study of deuterium abundance in the interstellar and
intergalactic space.
The GIMI imaging instrument is integrated into its carrier spacecraft (ARGUS, from the
USAF). It is primarily military in character aiming at detecting and characterizing sources
of UV emission (or atmospheric opacity), which could a#ect the detection, identification, and

UV SKY SURVEYS 41
tracking of missiles and warheads. ARGUS is expected to be launched in late­1998 on a Delta
II vehicle.
The GIMI instrument on ARGOS has as one of its declared goals the production of a full
sky survey in three UV bands. The most recent description of GIMI is by Carruthers & Seeley
(1996). GIMI consists of two bore­sighted EBCCD cameras, each imaging a 10 # .5 â 10 # .5 field
of view with a resolution of #3'.9. The paper gives also the sensitivities expected (Fig. 9). The
GIMI imagers cover three spectral regions with two detector­telescope assemblies. Camera 1
has a KCl cathode sensitive to the 75­110 nm band and with a nominal field­of­view (FOV) of
10 # .5 â 10 # .5. Camera 2 has a split FOV: an area of 7 # â 10 # .5 is covered by a KBr cathode
sensitive to the 131­160 nm band, while an adjacent area of 3 # .5 â 10 # .5 is imaged on a CsI
cathode sensitive to the 131­200 nm region. The most sensitive arrangement is Cam. 2 with the
CsI cathode, whereas Cam. 1, with its KCl cathode, has a sensitivity about 1/20 lower. In its
sky survey mode, GIMI will stare for at least 100 sec at each source; from this, its magnitude
limit is expected to be #13.6 in the CsI band.
The TAUVEX (Tel Aviv University UV Explorer) payload (Brosch et al. 1994) represents
the most advanced attempt to design, build and operate a flexible instrument for observations in
the entire UV band. The experiment is built for Tel Aviv University by El­Op, Electro­Optical
Industries, Ltd., with the largest part of the funding from the Government of Israel through
the Ministry of Science and Arts and the Israel Space Agency. TAUVEX images the same 0 # .9
FOV with three co­aligned telescopes and with an image quality of about 10''. It was originally
conceived for a small satellite of the OFEQ or SMEX class, but is now part of the scientific
complement of the SRG spacecraft scheduled to launch in late­1999 or in 2000.
The detectors used in TAUVEX are tailor­made by DEP­Delft Instruments of Roden, the
Netherlands. They consist of CsTe cathodes deposited on CaF 2 windows. The photo­electrons
are amplified by a three­stack chevron arrangement of MCPs and the detection is by a wedge­
and­strip anode. The light passes through filters defining six spectral bands, from a very wide
one which is essentially a ``blue cuto# '' at #200 nm, to three intermediate (#300 nm wide) bands
spanning the working range 135­290 nm, to two ``narrow'' bands at selected regions of interest.
A view of TAUVEX as it would look in space is shown in Fig. 11, where the TAUVEX thermal
model (now at Lavotchkin Industries in Moscow) is installed in a solar simulator chamber in
preparation of a thermal vaccum test. The apertures of the three telescopes are visible on the
right part of the figure and the telescopes themselves are wrapped in their thermal blankets.
TAUVEX o#ers significant redundancy at component and system levels in comparison with
other missions. The projected performance is detection of objects 19 mag (monochromatic) and
brighter, with S/N>10 and in three #40 nm wide bands, after a four hour pointing. At high
galactic latitudes each such pointing is expected to result in the detection of some tens of QSOs
and AGNs (mainly low­z objects) and some hundreds of galaxies and stars. It is expected that a
three­year operation of TAUVEX on SRG will yield a deep survey of the UV sky to m(UV)=19
mag in at least three UV bands, which will cover #5% of the (high galactic latitude) sky. This is
based on the size of the field of view and the expectation that the total number of independent
pointings after three years in orbit will add up to #3,000.
On the same SRG platform, co­aligned with TAUVEX and the rest of the higher energy
instruments, will operate also the F­UVITA instrument. It consists of a pair of 20 cm telescopes
imaging a 1 # .2 field with 10'' resolution in spectral bands between 91 and 99 nm. Each band is
#9 nm wide. The most up­to­date description of F­UVITA can be found in its homepage at the
Paul­Scherrer­Institute in Switzerland, where its sensitivity is advertised as adequate to detect
V=15 B0 stars.
Some UV capability has been constructed into the Optical/UV Monitor (OUVM) of XMM

42 NOAH BROSCH
Figure 11. The thermal demonstrator model of TAUVEX is identical in external appearance to the flight model.
It is shown here during installation in a space environmental chamber, prior to qualification under simulated Solar
radiation. The three telescope apertures are visible in the front of the instrument.
(Mason et al. 1996). The OUVM is a 30 cm modified Ritchey­Chr’etien telescope which images a
17' FOV with #1'' resolution. The light is analyzed with filters and with a grism. The UV­optical
detector operates in the 160­600 nm spectral range and consists of a fast readout 256â256 pixel
CCD, coupled by fiber boule to the phosphor output of an image intensifier. The CCD readout is
analyzed by a transputer­based processor and each photon event is centered to #0.2 pixel (0.5''
for normal operations). The di#erence from TAUVEX lies in the XMM OUVM being a single
telescope time­sharing between the optical and UV ranges, and in the higher spatial resolution,
realized at the expense of the smaller FOV.
In fall of 1997 NASA announced the selection of a Small Explorer Experiment (SMEX)
dedicated to a new all­sky survey in the UV. This is the Galaxy Evolution Explorer (GALEX: C.
Martin P.I.; Bianchi & Martin 1997) whose prime goal is the survey of galaxy evolution e#ects for
the range 0#z#2. GALEX is a collaboration of CALTECH, University of California­Berkeley, the
Johns Hopkins University, Columbia University, and the Laboratoire d'Astrophysique Spatiale
du CNRS of Marseille, France.
GALEX is a single wide­field imaging telescope with a 50 cm diameter primary mirror and
a 26 cm diameter secondary. Its focal plane is 1 # .2 wide and is equipped with two photon­
counting position­sensitive detectors. The angular resolution is 3''­5'' at 80% encircled energy.
The light is shared between the two detectors by a dichroic beam splitter and a folding flat mirror.
The detectors cover the spectral region 135­300 nm, selected by detector window transmission,
cathode response, and beamsplitter coatings. The nominal performance of the detectors, which
are cathodes with stacked MCPs and crossed, helical delay lines as anodes, allow for 4096 pixels
across the field­of­view. The two GALEX bands, as presently defined are 135­180 nm (CsI and

UV SKY SURVEYS 43
CaF 2 ), and 180­300 nm (CsTe and fused silica).
The GALEX mission is very ambitious, as it has both imaging as well as spectroscopic
aspects. It shall be conducted from a low Earth orbit, into which the Pegasus launcher will
insert the satellite. GALEX will observe by scanning the sky during the orbital night. The all­
sky imaging survey (AIS) phase is baselined to last four months, after which a two­color catalog
of sources to mL #19.4 will be produced. In addition, it is expected that the sky background will
be detected by binning the data to 1 arcmin 2 and after subtracting the point sources. GALEX
will then conduct a deep imaging survey (DIS) of # 200 degrees 2 of the sky, which will include
areas surveyed by the HDF North and South, the ESO Imaging Survey regions, areas of the
Sloan survey, etc. The deep imaging survey will detect sources #4 magnitudes fainter than the
all­sky survey. Some indication of source variability will be obtained from the 80­100 field revisits
done during the AIS and DIS phases.
The spectroscopic survey, to #1 nm resolution, shall be completed during additional four­
month periods by inserting a grism in the converging beam, before the beamsplitter. A shallow
spectroscopic survey to R#150 shall be conducted in the first period over the entire region of
the DIS, after which two additional periods shall be dedicated to deeper surveys of 20 degrees 2 ,
then 2 degrees 2 , for the Medium Spectroscopic Survey and the Deep Spectroscopic Survey,
respectively. The planned mission allows for a four­month contingency/Associate Investigator
programs.
It is now possible to compare the performance of the future missions which will bring sig­
nificant information about the deep UV universe. The three experiments, XMM UV/Optical
monitor, TAUVEX, and GALEX, reach approximately similar limiting magnitudes. GALEX
has a significant advantage in its global sky coverage, but it is restricted to two simultaneous
bands vs. three for TAUVEX. In comparison with the other two instruments, GALEX also
does not include a long­term continuous photometric monitoring option. Its spatial resolution
is advertised to be similar to that of the OUVM, twice better than that of TAUVEX. Both the
XMM UV/Optical monitor and TAUVEX shall operate alongside high energy imagers, allowing
a good determination of the nature of emitting sources. Their deployment in high­altitude orbits
vs. a LEO environment for GALEX will presumably ensure a better S/N, given a lower sky and
particle background.
9.3. WISH LIST
It is clear that ideally ``clean'' observations of the UV sky background should be made far from
the Earth's geocorona, only out of the Galactic plane, to avoid the dust­scattered starlight. For
practical purposes, we shall have to settle in the foreseeable future with observations which avoid
the dust in the Solar System. For this, a telescope must be located beyond the orbit of Saturn,
and this requires a specialized instrument. One possibility is to design multi­purpose missions to
the outer planets. These, during their cruise phase or when not collecting encounter data, could
be used to measure astronomical targets, such as the UV sky background. Toller (1983) and
Toller et al. (1987), for instance, used the Pioneer imaging photopolarimeter to map the optical
emission from the Milky Way in blue (395­485 nm) and in red (590­690 nm) from beyond the
asteroid belt, where the contribution of the zodiacal light was found to be negligible.
The findings by Toller were confirmed by the analysis of Schuerman et al. (1997), where the
boundary beyond which the zodiacal light contribution becomes very small, in comparison with
the di#use Galactic background, was set at 2.8 a.u. Gordon (1997) performed an analysis of the
Pioneer 10 and 11 data from beyond this limit and found an excess red component over that
predicted from the blue light distribution (which could be fully attributed to scattering by dust).

44 NOAH BROSCH
A location beyond the zodiacal dust cloud, therefore, would be ideal for a measurement of the
di#use UVB and of the exceedingly faint UV galaxies. Note also that some missions adopt an
``above the ecliptic'' trajectory to their target; this also would minimize the contamination by
zodiacal light and o#er the deepest astronomical observations.
Two possibilities, one real and another potential, are to include UV observations during the
cruise phases of either the Fast Pluto Flyby (``Pluto Express'') or a possible Neptune orbiter.
The Pluto Express mission was being designed when its funding was frozen in 1996 with a
probable launch date not before the turn of the century. The Neptune Orbiter has not yet
even been proposed o#cially, but would be a logical extension of the Cassini mission. Assuming
that Cassini will reach Saturn by 2004, a Neptune Orbiter could launch by 2010 for a Neptune
arrival 5­10 years after that. Most of its cruise phase could be made available, if so planned,
for astronomical observations. Similar possibilities exist for a putative Interstellar Precursor
mission.
Another, much cheaper option, altough restricted in its spectral band of observation, is to
follow up and expand on the experience gained by the Geneva/LAS group in UV observations
from balloons. Observations from stratospheric altitudes are limited to a spectral band #15 nm
wide centered near 200 nm. For a while, it looked like the SR­71's turned to NASA by the USAF
could be used for this type of obsevation. However, these planes are again no longer available for
scientific research. Despite this limitation, much can be gained from such a mission, if it would
rely on the heritage of long­duration balloon flights with updated hardware.
Super­pressure balloons, flying at 40­45 km altitude, can perform missions of weeks to
months. If launched in the Antarctic, the wind patterns keep the balloon mostly over the Antarc­
tic continent. A launch in the Arctic will probably have the same characteristics, but may intrude
too much in patterns of civilian and military flights over the North Pole. Trans­oceanic flights at
intermediate or equatorial latitudes are also possible and may be more convenient power­wise.
Present day technology allows real­time operation of a long­duration balloon mission by using
the TDRSS data relay satellites while having a high data rate (Israel 1993). Adding solar panels
to a regular long­duration balloon is not a di#cult task, and is one of the possibilities o#ered in
the latest (1997) NASA draft call for ``University Explorer'' missions (note though that super­
pressure, extremely high altitude balloons are not included in the UNEX AO). Such a mission
is therefore doable and will have a reasonable cost. I thus propose to study the possibility of
designing a UV sky survey at 200 nm from a long­duration balloon platform.
A major cost of any sky survey mission lies in building and qualifying space hardware. These
components could have been saved in the cost of such a mission if the GLAZAR­2 telescope
could have been reactivated. This requires first an assessment of the status of the optics, the
tracking mechanism, and the gimbals which compensate motions of the MIR, which were left
attached to the space station for almost one decade with no use. A new focal plane needs to be
designed, built, tested, assembled in place of the damaged one, and the new instrument can be
activated in space. However, the limitation now lies with the MIR itself, which has surpassed its
design lifetime.
Finally, an interesting possibility is to conduct a deep UV survey of a very small fraction
of the sky with the HST. The WFPC­2 has the capability to image #21 mag UV sources to
S/N#10 with deep observations using the F160BW and/or the F218W filters while covering a
significant sky region. It is thus possible to design a survey of UV sources as a ``parallel'' program.
A set of such observations could provide information about the faintest UV sources, beyond the
capability of UIT, FOCA, TAUVEX, or GALEX, with good morphological information, which
could be used to settle the issue of the cosmological UVB. Interestingly, observations with the
F160BW filter will be almost devoid of stars ! Our UV­galaxy model, using the bandpass of

UV SKY SURVEYS 45
F160BW, predicts less than one star/square degree (down to V=20) which could be detected in
#1000 sec exposures.
Using the WFPC­2 exposure time calculator, one finds that a 2,000 sec exposure at high
galactic latitude reaches a S/N of 1.8 with F160BW and 7.8 with F218W for an unresolved V=20
Sbc galaxy. Adopting the FOCA galaxy count from Armand & Milliard (1994) and assuming the
objects have # V #20 mag arcsec -2 , a 10,000 sec exposure with F218W will yield a S/N#1 per
WFPC­2 pixel. At least one such object should appear on almost any deep F218W exposure. This
will be, in principle, even deeper with the STIS, once parallel observations with this instrument
will be allowed. The HST has therefore the potential of making important observations in the
mode of an unbiased survey for faint stars and galaxies.
A possible instrument studied for the 2002 HST servicing mission is the Wide Field Camera
3 (WFC­3), which will replace WFPC­2. Its CCDs will probably have high quantum e#ciency
in the UV, perhaps an order of magnitude better than WFPC2. The existence of this camera,
with a field of view of 160''â160'' and a pixel size of 0.04 arcsec, will make a UV survey with
the HST even more attractive.
In the domain of wishes valid for every branch of astronomy one must include a wavelength­
sensitive, photon­counting imaging detector. One line of such detectors was described by Perry­
man et al. (1992, 1993, 1994). They consist of superconducting tunnel junctions (STJs) in which
a photon creates a cloud of charge carriers, which provide essentially a low resolution imaging
spectrometer with #
## #30. Presumably such detectors would be even more useful in the UV,
because of the higher energy of the analyzed photons, allowing a better energy resolution. A
single­pixel device of this type was already tested successfully in the laboratory (Peacock et al.
1996) and a quantum e#ciency of #50% was demonstrated for ## 200­1000 nm, with a spectral
resolution of 45 nm. A large­format array of such devices, possibly mounted at the WFPC po­
sition on HST, would provide an imaging spectrometer with unique properties. This apparently
has been proposed to ESA as a follow­up instrument for the 2002 HST refurbishing mission
(Peacock et al. 1997) but was not selected, presumably because the di#culties of a cryogenic
system to maintain the detector at T#1K in the HST environment seen daunting.
10. Conclusion
I have shown here that while UV astronomy flourished in the early and mid­70's, the field
stagnated, at least where sky surveys are concerned, from then on. In the early days of UV
studies, a valuable resource appeared with the production of the TD--1 catalog. This is still, 20
years after its publication, the only all­sky survey in the UV. The TD--1 survey detected mostly
stars, but more sensitive instruments revealed very interesting extragalactic sources (AGNs,
star­bursting galaxies, etc.). The most sensitive instrument now available for UV imaging is the
HST, which can hardly qualify as a survey instrument. In terms of surveys, a success of the
GIMI and UVISI missions will extend the TD--1 survey by about five magnitudes, and even
then, only few extragalactic objects would be accessible. The big step forward will come with
the GALEX, TAUVEX and XMM UV/Optical monitor surveys.
The EUV range was explored by the ROSAT WFC and by the EUVE, but due to the
opacity of the ISM, only relatively nearby objects were detected. The outstanding sources are
viewed through ``windows'' in the ISM, where low HI column densities are encountered. At
present, there are no prospects for an EUVE follow­up mission, but the extension of the EUVE
operations promises more and interesting data.
It is clear that, with the existing attitude toward pure research and space activities in general,
it will be extremely di#cult to channel funds for new space missions which will perform additional

46 NOAH BROSCH
UV sky surveys deeper than and beyond that by GALEX. Therefore, the improvement in human
knowledge of the astronomical UV sky can only come about through judicious use of existing,
or multi­purpose platforms, where the UV science piggy­backs on other spectral ranges. Not
only astronomy missions must be considered as able to yield good UV data, but also planetary
or upper­atmosphere science platforms are eligible. Fully new, revolutionary designs tend to be
costly; at the present stage they should be avoided. The heritage of past missions should be fully
realized before embarking on new adventures. In this light, the extension of the EUVE mission
in a low­cost mode is commendable; the continuation of the right­angle surveys should reveal
more faint EUV sources.
Science needs the multi­spectral all­sky UV imaging survey to 19­20 monochromatic mag­
nitude which will be provided by GALEX, as well as a combination of cheaper alternatives,
including a long­duration, very high­altitude balloon with a FOCA or ASCHOT­type telescope
and a high­resolution detector with electronic readout for a sky survey in the 200 nm band.
A mini­survey with HST in the UV, perhaps with an STJ camera, will take advantage of the
exquisite imaging capabilities of this facilty. An extended EUV all­sky survey, # 100â more
sensitive than EUVE and with better angular resolution, is also required. The inclusion of a
UV sky survey phase in planetary missions bound for the outer Solar system, to observe from
beyond the zodiacal dust cloud, is advocated.
Acknowledgements
UV research at Tel Aviv University is supported by special grants from the Government of
Israel the Ministry of Science and Arts, through the Israel Space Agency, and from the Aus­
trian Friends of Tel Aviv University, as well as by a Center of Excellence Award from the
Israel Science Foundation. I acknowledge support from a US­Israel Binational Award to study
UV sources imaged by the FAUST experiment, and the hospitality of NORDITA, the Danish
Space Research Institute, and the Space Telescope Science Institute, where parts of this review
were prepared. I am grateful for the help of many individuals in producing this review. Benny
Bilenko calculated sky models, Liliana Formiggini recalculated the optical­to­UV transformation
and compared the predicted and actual UV brightness for Virgo, Hrant Tovmassian explained
intricacies of the GLAZAR series, and Je# Bloch from the ALEXIS team supplied sky charts and
information. Bruno Milliard from LAS Marseille clarified a number of points related to FOCA
observations and allowed me to look at a FOCA UV image and optical spectra of a few galaxies
in it, Michael Lampton from the Berkeley's SSL/CEA explained intricacies of the EUVE source
count, Jesse Hill from the UIT team explained UIT results, David Bersier from the Observa­
toire de G’en‘eve added information on M51 and supplied its calibrated FOCA UV image, Jayant
Murthy added information about UVISI, Luciana Bianch provided GALEX details in advance
of publication, and David Israel from GSFC explained the TDRSS­balloon connection; for these
I am very grateful. Mark Hurwitz supplied information and a list of publications from the OR­
FEUS flights; Bill Waller did the same for UIT, read an early version of the paper, and provided
some useful remarks. I thank Alan Dressler for some information on his committee's report,
Marie Treyer for calculating the contribution to the UV background from FOCA galaxies, and
Prab Gondhalekar for constructive remarks upon reading one of the final drafts. Boris Shustov
kindly supplied details about the Spectrum­UV mission. I am grateful to the referee for very
constructive remarks.

UV SKY SURVEYS 47
References
Alloin, D., Santos­Lleo, M. et al. 1995 Astron. Astrophys. 293, 293.
Almoznino, E. & Brosch, N. 1997a Mon. Not. R. astr. Soc., submitted.
Almoznino, E. & Brosch, N. 1997b Mon. Not. R. astr. Soc., submitted.
Armand, C. & Milliard, B. 1994 Astron. Astrophys. 282, 1.
Armand, C., Milliard, B. & Deharveng, J.­M. 1994 Astron. Astrophys. 284, 12.
Bahcall, J.N. 1991 Science 251, 1412.
Bahcall, J.N. & Soneira, R.M. 1980 Astrophys. J. Suppl 44, 73.
Barber, C.R., Roberts, T.P. & Warwick, R.S. 1996 Mon. Not. R. astr. Soc. 282, 157.
Belyaev, V.P., Kurt, V.G., Melioranskii, A.S., Smirnov, A.S., Sorokin, L.S. & Tilt, V.M. 1971 Cosmic Res. 8, 677.
Bersier, D., Blecha, A, Golay, M. & Martinet, L. 1994 Astron. Astrophys. 286, 37.
Bianchi, L. & Martin, C. 1997 in ``The Ultraviolet Astrophysics beyond the IUE final Archive'' (R.Harris, ed.)
ESA SP, in press.
Bilenko, B. 1995 MSc thesis, Tel Aviv University.
Bixler, J., Bowyer, S., Deharveng, J.­M., Courtes, G., Malina, R., Martin, C. & Lampton, M. 1984 Science 225,
184.
Bloch, J. 1995 in ``Astrophysics in the Extreme Ultraviolet'' (S. Bowyer & R.F. Malina, eds.) Dordrecht: Kluwer
Academic Publishers, pp. 7­14.
Boggess, A. & Dunkelman, L. 1959 Astrophys. J. 129, 236.
Boggess, A. et al. 1978 Nature 275, 372.
Boggess, A. & Wilson, R. 1987 ``The History of IUE'', in ``Exploring the Universe with the IUE satellite'' (Y.
Kondo & W. Wamsteker, eds.), Dordrecht: Reidel Publishing Co., p. 3­20.
Bohlin, R.C., Hill, J.K., Stecker, T.P. & Witt, A.N. 1982 Astrophys. J. 255, 87.
Bohlin, R.C., Cornett, R.H., Hill, J.K., O'Connell, R.W. & Stecker, T.P. 1990 Astrophys. J. 352, 55.
Boksenberg, A., Evans, R.G., Fowler, R.G., Gardner, I.S.K., Houziaux, L., Humphries, C.M., Jamar, C., Macau,
D., Malaise, D., Monfils, A., Nandy, K., Thompson, G.I., Wilson, R. & Wroe, H. 1973 Mon. Not. R. astr. Soc.
163, 291.
Boyle, B.J. 1989 Mon. Not. R. astr. Soc. 240, 533.
Bowyer, S. 1990 in ``Observatories in Earth Orbit and Beyond'' (Y. Kondo, ed.), Kluwer Academic Publishers,
p.153.
Bowyer, S. 1990 in ``The Galactic and Extragalactic Background Radiation'' (S. Bowyer and Ch. Leinert, eds.),
Dordrecht: Reidel, p. 171.
Bowyer, S. 1991 Ann. Rev. Astron. Astrophys. 29, 59.
Bowyer, S. & Malina, R.F. 1991 in ``Extreme Ultraviolet Astronomy'' (R.F.Malina and S. Bowyer, eds.), New
York: Pergamon Press, p. 397.
Bowyer, S., Freeman, J., Paresce, F. & Lampton, M. 1977 Appl. Opt. 16, 756.
Bowyer, S., Kimble, R., Paresce, F., Lamton, M. & Penegor, G. 1981 Appl. Opt. 20, 477.
Bowyer, S., Sasseen, T., Lampton, M. & Wu, X. 1993 Astrophys. J. 415, 875.
Bowyer, S., Sasseen, T., Wu, X. & Lampton, M. 1995 Astrophys. J. Suppl 96, 461.
Bowyer, S., Lieu, R., Lamton, M., Lewis, J., Wu, X., Drake, J.J. & Malina, R.F. 1994 Astrophys. J. Suppl 93,
569.
Bowyer, S., Lampton, M., Kewis, J., Wu, X., Jelinsky, P. & Malina, R.F. 1996 Astrophys. J. Suppl 102, 129.
Broadfoot, A.L., Sandel, B.R., Shemansky, D.E., Atreya, S.K., Donahue, T.M., Moos, H.W., Bertaux, J.L., Bla­
mont, J.E., Ajello, J.M. & Strobel, D.F. 1977 Space Sci. Review 21, 183.
Brosch, N. 1991 Mon. Not. R. astr. Soc. 250, 780.
Brosch, N., Shemi, A., Netzer, H., Blasberger, A. & Topaz, J.M. 1994 in ``Advances in multilayer and grazing
incidence X­ray/EUV/FUV optics'' (R.B. Hoover and A.B. Walker, eds.), SPIE 2279, p. 469.
Brosch, N., Almoznino, E., Leibowitz, E.M., Netzer, H., Sasseen, T.P., Bowyer, S., Lampton, M. & Wu, X. 1995
Astrophys. J. 450, 137.
Brosch, N., Formiggini, L., Almoznino, E., Sasseen, T., Lampton, M. & Bowyer, S. 1997 Astrophys. J. Suppl.
111, 143.
Brosch, N., Ofek, E., Almoznino, E., Sasseen, T., Lamton, M. & Bowyer, S. 1998 Mon. Not. R. astr. Soc., in press.
Buat, V., Donas, J. & Deharveng, J.­M. 1987 Astron. Astrophys. 185, 33.
Buat, V., Deharveng, J.­M., & Donas, J. 1989 Astron. Astrophys. 223, 42.
Buat, V., Vuillemin, A., Burgatella, D., Milliard, B. & Donas, J. 1994 Astron. Astrophys. 281, 666.
Burnasheva, B.A., Gershberg, R.E., Zvereva, A.M., Il'in, I.V. & Shakhovskaia, N.I. 1989 Astron. Zhurnal 66, 328.
Burstein, D., Bertola, F., Buson, L.M., Faber, S.M.& Lauer, T.R. 1988 Astrophys. J. 328, 440.
Byram, E.T., Chubbs, T.A., Friedman, H. & Kupperian, J.E. 1957 in ``The Threshold of Space'' (M. Zeliko#,
ed.), London: Pergamon Press, p. 203 (quoted in Boggess & Wilson 1987 ``The History of IUE'', op.cit.)
Carbary, J.F., Darlington, E.H., He#ernan, K.J., Harris, T.J., Meng, C.­I., Mayr, M.J., Mcevaddy, P.J. & Peacock,
K. 1994 in ``Aerial Surveilance Sensing Including Obscured and Underground Object Detection'' (I. Cindrich,

48 NOAH BROSCH
N.K. DelGrande, S. Gowrinathan, P.B. Johnson, and J.F. Shanley, eds.) SPIE 2217, pp. 204­212.
Carnochan, D.J., Dworetsky, M.M., Todd, J.J., Willis, A.J. & Wilson, R. 1975 Phil. Trans. R. Soc. London A
279, 479.
Carnochan, D.J. & Wilson, R. 1983 Mon. Not. R. astr. Soc. 202, 317.
Carruthers, G.R. 1973 Appl Opt 12, 2501.
Carruthers, G.R., Heckathorn, H.M. & Gull, T.R. 1980 Astrophys. J. 237, 438.
Carruthers, G., Morril, J., Dohne, B.C. & Reeb, S.A. 1993 SPIE 1764 ``UV Technology IV'', p. 21.
Carruthers, G., Dohne, B.C., Shephard, K.K., Reeb, S.A. & Schmidt, E.G. 1994 SPIE 2282 ``UV Technology V'',
p. 184.
Carruthers, G.R. & Page, T. 1976 Astrophys. J. 205, 397.
Carruthers, G.R. & Page, T. 1983 Astrophys. J. Suppl. 53, 623.
Carruthers, G.R. & Page, T. 1984a Astrophys. J. Suppl. 54, 271.
Carruthers, G.R. & Page, T. 1984b Astrophys. J. Suppl. 55, 101.
Carruthers, G.R. & Page, T. 1984c Pub. Astr. Soc. Pacific 96, 447.
Carruthers, G.R. & Seeley, T.D. 1996 in ``Ultraviolet Atmospheric and Space Remote Sensing: Methods and
Instrumentation'' (R.E. Hu#man & C.G. Stergis, eds.) SPIE 2831, 65­93.
Chavez, M., Stalio, R. & Holberg, J.B. 1995 Astrophys. J. 449, 280.
Chechik, R. & Breskin, A. 1988 Nucl. Instrum. Methods A264, 251.
Chen, P.C. et al. 1992 Astrophys. J. Lett. 395, L41.
Cheng, K.­P. et al. 1992, Astrophys. J. Lett. 395, L29.
Code, A. 1983 in ``Problems of calibration of absolute magnitude and temperature of stars'' (B. Hauck and B.
Westerlund, eds.) Dordrecht: Reidel, p. 131.
Code, A., Houck, T.E., McNall, J.F., Bless, R.C. & Lillie, C.F. 1970 Astrophys. J. 161, 377.
Cohen, M. 1994 Astron. J. 107, 582.
Cohen, M., Sasseen, T. & Bowyer, S. 1995 Astrophys. J. 427, 848.
Cornett, R.H., Jenkins, E.B., Bohlin, R.C., Cheng, K.­P., Gull, T.R., O'Connell, R.W., Parker, R.A.R., Roberts,
M.S., Smith, A.W., Smith, E.P. & Stecher, T.P. 1992 Astrophys. J. Lett. 395, L9.
Cornett, R.H., Greason, M.R., Hill, J.K., Parker, J.Wm. & Waller, W.H. 1997 Astron. J. 113, 1011.
Court’es, G., Viton, M., Sivan, J.P., Decker, R. & Gary, A. 1984 Science 225, 179.
Court’es, G., Petit, H., Hua, C.T., Martin, P., Blecha, A., Huguenin, D. & Golay, M. 1993 Astron. Astrophys.
268, 419.
Courvoisier, T.J.­L., Orr, A., Buehler, P., Zehnder, A. & Henneck, R., Stau#acher, F., Biakhowski, J., Schlumpf,
N., Schoeps, W. & Mchedlishvili, A. 1993 Experimental Astronomy 4, 117.
Craig, N., Abbott, M., Finley, D., Jessop, H., Howell, S.B., Mathioudakis, M., Sommers, J., Vallerga, J.V. &
Malina, R.F. 1997 Astrophys. J.S 113, 131.
Crotts, A.P.S., Landsman, W.B., Bohlin, R.C., O'Connell, R.W., Roberts, M.S., Smith, A.M. & Stecher, T.P.
1992 Astrophys. J. Lett. 395, L25.
Davidsen, A.F., Kriss, G.A., Ferguson, H.C., Blair, W.P., Bowers, C.W. & Kimble, R.A. 1991 Nature 351, 128.
Davidsen, A.F., Long, K.S., Durrance, S.T., Blair, W.P., Bowers, C.W., Conard, S.J., Feldman, P.D., Ferguson,
H.C., Fountain, G.H., Kimble, R.A., Kriss, G.A., Moos, H.W. & Potocki, K.A. 1992 Astrophys. J. 392, 264.
Davis, R.J. et al. 1972 in ``Scientific Results from the Orbiting Astronomical Observatory (OAO­2)'' (A.D. Code,
ed.) NASA SP­310, p. 1.
Davis, R.J., Deutschman, W.A. & Haramundanis, K.L. 1973 ``Celescope Catalog of Ultraviolet Stellar Observa­
tions'', Washington: Smithsonian Institution Press.
de Boer, K.S. 1982 Astron. Astrophys. Suppl. 50, 247
Deharveng, J.­M., Riviere, G.P., Monnet, G. Moutonnet, J., Courtes, G., Deshayes, J.P. & Berges, J.C. 1979
Space Science Instrumentation 5, 21.
Deharveng, J.­M., Sasseen, T.P., Buat, V., Bowyer, S., Lampton, M. & Wu, X. 1994, Astron. Astrophys. 289,
715.
Donas, J., Deharveng, J.­M., Laget, M., Milliard, B. & Huguenin, D. 1987, Astron. Astrophys. 180, 12.
Donas, J., Milliard, B., Laget, M. & Deharveng, J.­M. 1981 Astron. Astrophys. 97, L7.
Donas, J., Milliard, B. & Laget, M. 1995 Astron. Astrophys. 303, 661.
Dressler, A. 1996 ``Exploration and the Search for Origins: A Vision for Ultraviolet­Optical­Infrared Space As­
tronomy'', Washington: Association of Universities for Research in Astronomy.
Ellis, R.S. 1997 Ann. Rev. Astron. Astrophys. 35, 389.
Ellis, R.S., Gondhalekar, P.M. & Efstathiou, G. 1982 Mon. Not. R. astr. Soc. 201, 223.
Fanelli, M.N., O'Connell, R.W. & Thuan, T.X. 1987 Astrophys. J. 321, 768.
Fix, J., Craven, J. & Frank, L. 1989 Astrophys. J. 345, 203.
Giavalisco, M., Livio, M., Bohlin, R.C., Machetto, F.D. & Stecher, T. 1966 Astron. J. 112, 369.
Gondhalekar, P.M. 1990 in ``The Galactic and Extragalactic Background Radiation'' (S. Bowyer & Ch. Leinert,
eds.), Dordrecht: Kluwer Academic Publishers, p. 49.
Gondhalekar, P.M. & Wilson, R. 1975 Astron. Astrophys. 38, 329.

UV SKY SURVEYS 49
Gondhalekar, P.M., Phillips, A.P. & Wilson, R. 1980 Astron. Astrophys. 85, 272.
Gordon, K. 1997 Pub. Astr. Soc. Pacific 109, 1190.
Gry, C., York, D.G. & Vidal­Madjar, A. 1985 Astrophys. J. 296, 593.
Gurzadyan, G.A, Ohanesyan, J.B., Rustambekova, S.S. & Ephremyan, R.A. 1985 ``Gatalogue of 900 faint­star
ultraviolet spectra'', Yerevan: Publishing House of the Armenian Academy of Sciences.
Guzman, R., Gallego, J., Koo, D.C., Phillips, A.C., Lowenthal, J.D., Faber, S.M., Illingworth, G.D. & Vogt, N.P.
1997 Astrophys. J. 489, 559.
Haisch, B., Linsky, J.L., Lampton, M., Paresce, F., Margon, B. & Stern, R. 1976 Astrophys. J. 213, L119.
Harkala, L.K., Mattila, K., Bowyer, S., Sasseen, T.P. & Lampton, M. 1995 Astrophys. J. 443, L33.
Hayakawa, S., Yamashita, K. & Yoshoka, S. 1969 Ap. Space Sci. 5, 493.
Hayes, D.S. and Latham, D.W. 1975 Astrophys. J. 197, 593.
Heck, A., Egret, D., Jaschek, M. & Jaschek, C. 1984 ``IUE Low­Dispersion Spectra Reference Atlas ­ Part 1.
Normal Stars'', Paris: ESA SP­1052.
He#ernan, K.J., Heiss, J.E., Boldt, J.D., Darlington, E.H., Peacock, K., Harris, T.J. & Mayr, M.J. 1996 Johns
Hopkins APL Technical Digest 17, 198.
Henize, K.G., Wray, J.D., Parson, S.B., Benedict, G.F., Bruhweiler, F.C., Rybski, P.M & O'Callaghan, F.G. 1975
Astrophys. J. Lett. 199, L119.
Hennessy, G.S., O'Connell, R.W., Cheng, K.­P., Bohlin, R.C., Collins, N.R., Gull, T.R., Hintzen, P., Isensee, J.E.,
Landsman, W.B., Roberts, M.S., Smith, A.M., Smith, E.P. & Stecher, T.P. 1992 Astrophys. J. Lett. 395, L13.
Henry, P., Cruddace, R., Paresce, F., Bowyer, S. & Lampton, M. 1975a Astrophys. J. 195, 107.
Henry, P., Cruddace, R., Paresce, F., Lampton, M. & Bowyer, S. 1975b Rev Sci Instr 46, 355.
Henry, P., Cruddace, R., Lampton, M., Paresce, F. & Bowyer, S. 1975c Astrophys. J. Lett. 197, L117.
Henry, R.C. 1977 Astrophys. J. Suppl. 33, 451.
Henry, R.C. 1982 10­th Texas Symposium for Relativistic Astrophysics, New York: Ann NY Acad. Sci., p. 428.
Henry, R.C. 1991 Ann. Rev. Astron. Astrophys. 29, 89
Henry, R.C., Weinstein, A., Feldman, P.D., Fastie, W.G. & Moos, H.W. 1975 Astrophys. J. 201, 613.
Henry, R.C., Swandic, J.R., Shulman, S.D. & Fritz, G. 1977 Astrophys. J. 212, 707.
Henry, R.C., Landsman, W.B., Murthy, J., Tennyson, P.D., Wo#ord, J.B. & Wilson, R. 1988 ``Atlas of the UV
sky'', Baltimore: The Johns Hopkins University Press.
Henry, R.C. Feldman, P.D., Kruk, J.W., Davidsen, A.F. & Durrance, S.T. 1995 Astrophys. J. 454, L69.
Hill, R.S., Hill, J.K., Landsman, W.B., Bohlin, R.C., Cheng, K.­P., Hintzen, P.M.N., O'Connell, R.W., Roberts,
M.S., Smith, A.M., Smith, E.P. & Stecher, T.P. 1992a Astrophys. J. Lett. 395, L17.
Hill, J.K., Pfarr, B.B., Bohlin, R.C., Isensee, J.E., O'Connell, R.W., Ne#, S.G., Roberts, M.S., Smith, A.M. &
Stecher, T.P. 1992b Astrophys. J. Lett. 395, L33.
Hill, J.K., Bohlin, R.C., Cheng, K.­P., Hintzen, P.M.N., Landsman, W.B., Ne#, S.G., O'Connell, R.W., Roberts,
M.S., Smith, A.M., Smith, E.P. & Stecher, T.P. 1992c Astrophys. J. Lett. 395, L37.
Hill, J.K., Bohlin, R.C., Cheng, K.­P., Fanelli, M.N., Hintzen, P., O'Connell, R.W., Roberts, M.S., Smith, A.W.,
Smith, E.P. & Stecher, T.P. 1993 Astrophys. J. 413, 604.
Hill, J.K., Isensee, J.E., Cornett, R.H., Bohlin, R.C., O'Connell, R.W., Roberts, M.S., Smith, A.W. & Stecher,
T.P. 1994 Astrophys. J. 425, 122.
Hill, J.K., Isensee, J.E., Bohlin, R.C., Cheng, K.­P., Hintzen, P., O'Connell, R.W., Roberts, M.S., Smith, A.W.,
Smith, E.P. & Stecher, T.P. 1995a Astrophys. J. Suppl. 98, 595.
Hill, J.K., Cheng, K.­P., Bohlin, R.C., O'Connell, R.W., Roberts, M.S., Smith, A.W. & Stecher, T.P. 1995b
Astrophys. J. 446, 622.
Hill, J.K., Waller, W.H., Cornett, R.H., Bohlin, R.C., Cheng, K.­P. Ne#, S.G., O'Connell, R.W., Roberts, M.S.,
Smith, A.M., Hintzen, P.M.N., Smith, E.P. & Stecher, T.P. 1997 Astrophys. J. 477, 673.
Holberg, J.B. 1986 Astrophys. J. 311, 969.
Holberg, J.B. 1990 in ``Observatories in Earth Orbit and Beyond'', pp. 49­57.
Holberg, J.B. 1991 in ``Extreme Ultraviolet Astronomy'', (R. Malina and S. Bowyer, eds.), New York: Pergamon
Press, p.8.
Holt, S.S. & Mundy, L.G. 1996 in ``Star Formation Near and Far: Proceedings of the Seventh Astrophysics
Conference'', AIP Conference Proceedings no. 393, Woodbury, NY: AIP Press.
Huguenin, D. & Magnan, A. 1978 in ``European Sounding Rocket, Balloon and Related Research, with Emphasis
on Experiments at High Latitudes'' Proceedings of the ESRANGE Symposium, Ajaccio 1978, ESA SP­135,
403.
Humphries, C.M., Jamar,C., Malaise, D. & Wroe, H. 1976 Astron. Astrophys. 49, 389.
Hurwitz, M., Bowyer, S. & Martin, C. 1989 in ``Galactic and Extragalactic Background Radiation'' (S. Bowyer
and Ch. Leinert, eds.), Dordrecht: Kluwer.
Hurwitz, M., Bowyer, S. & Martin, C. 1991 Astrophys. J. 372, 167.
Hurwitz, M., Bowyer, S., Kudritzki, R. & Lennon, D. Astrophys. J. 450, 149.
Hurwitz, M. & Bowyer, S. 1996 Astrophys. J. 465, 296.
Israel, D.J. 1993 IEEE AES Systems Magazine (Feb. 1993), 43.

50 NOAH BROSCH
Jakobsen, P., Bowyer, S., Kimble, R., Jelinsky, P., Grewing, M. & Kramer, G. 1984 Astron. Astrophys. 139, 481.
Jelinsky, P., Vallerga, J.V. & Edelstein, J. 1995 Astrophys. J. 442, 653.
Jenkins, E.B., Reale, M.A., Zucchino, P.M. & Sofia, U.J. 1996 Astrophys. Space Sci. 239, 315.
Joseph, C.L. 1995 Experimental Astron. 6, 97.
Joseph, C.L., Arbright, V., Abraham, J., Dieball, D., Franka, S., Styonavich, M., Van Houten, C., Danks, T. &
Woodgate, B. 1995 SPIE 2551, 248.
Joubert, M., Masnon, J., Lequeux, J., Deharveng, J.­M. & Cruvellier, P. 1983 Astron. Astrophys. 128, 114.
Katsova, M.M. & Livshits, M.A. 1989 Astron. Zhurn. 66, 307.
Kennedy, M.J., Friedmann, S.D., Barkhouser, R.H., Hampton, J. & Nikulla, P. 1996 in ``Space Telescopes and
Instruments IV'' (P.Y. Bely & J.B. Breckenridge, eds.) SPIE 2807, pp. 172­183.
Keshi­Kuha, R.A., Osantowski, J.F., Leviton, D.B., Saha, T.T., Wright, G.A., Boucarut, R.A., Fleetwood, C.M.
& Madison, T.J. 1997 Opt. Engineering 36, 157­161.
Kinney, A.L., Bohlin, R.C., Calzetti, D., Panagia, N. & Wyse, R.F.G. 1993 Astrophys. J. Suppl. 86, 5.
Kiszkurno­Koziej, E. & Lequeux, J. 1987 Astron. Astrophys. 185, 291.
Kodaira, K., Watanabe, T., Onaka, T. & Tanaka, W. 1990 Astrophys. J. 363, 422.
Kroupa, P., Tout, C.A. & Gilmore, G. 1993 Mon. Not. R. astr. Soc. 262, 545.
Kumar, S., Paresce, F., Bowyer, S. & Lampton, M. 1974 Appl. Optics 13, 575.
Kung, P., Saxler, A., Zhang, X., Walker, D., Razenghi, M. & Ulmer, M.P. 1996 in ``Photodetectors: Materials and
Devices'' (G.J. Brown & M. Razenghi, eds.) SPIE 2685, 126.
Kupperian, J.E. & Milligan, J.E. 1957 Astron. J. 62, 22.
Kupperian, J.E., Boggess, A.B. & Milligan, J.G. 1958 Astrophys. J. 128, 453.
Kurucz, R.L. 1991 NATO ASI Ser. C., Math. Phys. Sci., 341, 441.
Kurucz, R.L., Peytremann, E. & Avrett, E.H. 1974 ``Blanketed Model Atmospheres for Early­Type Stars'' (Wash­
ington: Smithsonoan Institution Press).
Laget, M. 1980 Astron. Astrophys. 81, 37.
Laget, M., Burgarella, D., Milliard, B. & Donas, J. 1991a ``Astrophysics at FUV and EUV Wavelengths'' (B.
Aschenbach, ed.), Adv. Space Res. 11, 139.
Laget, M., Burgarella, D., Milliard, B. & Donas, J. 1991b Astron. Astrophys. 259, 510.
Lallement, R. 1993 Adv. Space Res. 13, 113.
Lampton, M., Margon, B., Paresce, F., Stern, R. & Bowyer, B. 1976 Astrophys. J. Lett. 203, L71
Lampton, M. 1987 SPIE vol. 868 ``Optoelectronic Technologies for Remote Sensing from Space'', p.12.
Lampton, M., Lieu, R., Schmidt, J.H.M.M., Bowyer, S., Voges, W., Lewis, J. & Wu, X. 1997 Astrophys. J. Suppl.
108, 545.
Landsman, W.B. 1984 PhD thesis, Johns Hopkins University.
Landsman, W. et al. 1992 Astrophys. J. Lett. 395, L21.
Landsman, W.B., Roberts, M.S., Bohlin, R.C., O'Connell, R.W., Smith, A.M. & Stecher, T.P. 1992 Astrophys.
J. Lett. 401, L83.
Landsman, W., Simon, T. & Bergeron, P. 1996 Pub. Astr. Soc. Pacific 108, 250.
Levasseur­Regourd, A.C. & Dumont, R. 1980 A&A 84, 277.
Leinert, Ch., Bowyer, S., Haikala, L., Hanner, M., Hauser, M.G., Levasseur­Regourd, A.­Ch., Mann, I., Reach,
W.T., Schlosser, W., Staude, J., Toller, G.N., Weiland, J.L., Weinberg, J.L. & Witt, A.N. 1998 Astron.
Astrophys. Suppl. 127, 1.
Leitherer, C., Alloin, D. et al. 1996 Pub. Astr. Soc. Pacific 108, 996.
Lipovetsky, V.A. 1992 in ``Astronomy from Wide­Field Imaging'' (H.T. MacGillivray, E.B. Thomson, B.M. Lasker,
I.N. Reid, D.F. Malin, R.M. West and H. Lorenz eds.), Dordrecht: Reidel, p. 3.
Liubimkov, L.S. 1990 Astron. Zhurn. 67, 480.
Longo, G., Ceriello, A. & Capaccioli, M. 1991 Astron. Astrophys. Suppl. 90, 375.
Longo, G. & Capaccioli, M. 1992 ``IUE ULDA Access Guide No. 3: Normal Galaxies'', ESA SP­1152.
MacKenty, J. W., Baggett, S.M., Biretta, J., Hinds, M., Ritchie, C.E., Feinberg, L.D. & Trauger, J.T. 1995 Proc.
SPIE 2478, p.160­166.
Madau, P., Ferguson, H.C., Dickinson, M.E., Giavalisco, M., Steidel, C.G. & Fruchter, A. 1996 MNRAS 283,
1388.
Maoz, D., Barth, A.J., Steinberg, A., Filippenko, A.V., Ho, L.C., Machetto, F.D., Rix, H.­W. & Schneider, D.P.
1996 Astron. J. 111, 2248.
Margon, B., Lampton, M., Bowyer, S., Stern, R. & Paresce, F. 1976 Astrophys. J. Lett. 210, L79.
Margon, B., Szkody, P., Bowyer, S., Lampton, M. & Paresce, F. 1978 Astrophys. J. 224, 167.
Martin, C. 1998 in ``The Ultraviolet Universe at Low and High Redshifts'', AIP Press, in press.
Martin, C., Hurwitz, M. & Bowyer, S. 1991 Astrophys. J. 379, 549.
Mason, K.O., Cropper, M.S., Hunt, R., Horner, S.D., Priedhorsky, W.C., Ho, C., Cordova, F.A., Jamar, C. &
Antonello, E. 1996 in ``EUV, X­ray, and Gamma­ray Instrumentation for Astronomy VII'' (O.H. Siegmund &
M.A. Grummin, eds.) SPIE 2808, pp. 438­447.
Maucherat­Joubert, M., Deharveng, J.­M. & Cruvellier, P. 1980 Astron. Astrophys. 88, 323.

UV SKY SURVEYS 51
McQuade, K., Calzetti, D. & Kinney, A.L. 1995 Astrophys. J. Suppl 97, 331.
Meier, R.R. 1991 ``Ultraviolet spectroscopy and remote sensing of the upper atmosphere'' Space Sci. Rev 58,
1­185.
M’endez, R.A., Minniti, D., de Marchi, G., Baker, A. & Couch, W.J. 1996 Mon. Not. R. astr. Soc. 283, 666.
Merkulova, N.I., Metik, L.P., Pronik, V.I. & Pronik, I.I. 1990 Sov. Astron. 34, 225.
Miller, G.E. & Scalo, J.M. 1979 Astrophys. J.S 41, 513.
Milliard, B., Laget, M., Donas, J., Deharveng, J.­M., Huguenin, D. & Viton, M. 1983 ``A sky survey at balloon­
ultraviolet wavelengths'' 6th ESA Symposium, Interlaken, ESA SP­183, p. 409.
Milliard, B., Donas, J. & Laget, M. 1991 ``Astrophysics at FUV and EUV Wavelengths'' (B. Aschenbach, ed.)
Adv. Space Res. 11, 135.
Milliard, B., Donas, J., Laget, M., Armand, C. & Vuillemin, A. 1992 Astron. Astrophys. 257, 24.
Morgan, D.H., Nandy, K. & Thompson, G.I. 1976 Mon. Not. R. astr. Soc. 177, 531.
Morton, D.C. & Spitzer, L. 1966 Astrophys. J. 144, 1.
Murthy, J. & Henry, R.C. 1995 Astrophys. J. 448, 848.
Murthy, J., Hall, D., Earl, M., Henry, R.C. & Holberg, J.B. 1996 ``An Analysis of 17 Years of Voyager Observa­
tions'', in preparation.
Napiwotzki, R., Hurwitz, M., Jordan, S., Bowyer, S., Koester, D., Weidemann, V., Lampton, M. & Edelstein, J.
1995 A&A 300, L5.
Ne#, S.G., Stecher, T.P. & the UIT Science Team 1995 BAAS 186, 20.10N.
Noter, Y., Nahor, G., Lifshitz, Y., Saar, N. & Braun, O. 1993 Proc. SPIE 1971, p. 276.
O'Connell, R.W. 1987 Astron. J. 94, 876
O'Connell, R.W. 1990 in ``Windows on galaxies'' (G. Fabbiano, J.S. Gallagher and A. Renzini, eds.), Dordrecht:
Kluwer Publishing Co., p. 39.
O'Connell, R.W. 1991 ``Astrophysics at FUV and EUV Wavelengths'' (B. Aschenbach, ed.), Adv. Space Res. 11,
71.
O'Connell, R.W. 1992 in ``The stellar population of galaxies'' (B. Barbuy and A. Renzini, eds.) D. Reidel Publishing
Co., p. 233.
O'Connell, R.W., Bohlin, R.C., Collins, N.R., Cornett, R.H., Hill, J.K., Hill, R.S., Landsman, W.B., Roberts,
M.S., Smith, A.M. & Stecher, T.P. 1992 Astrophys. J. Lett. 395, L45.
O'Connell, R.W. & Marcum, P. 1996 in ``HST and the high redshift universe'' (N.R. Tanvir, A. Aragon­Salamanca,
and J. Walsh, eds.).
Oke, J.B. & Gunn, J.E. 1983 Astrophys. J. 266, 713.
Okumura, K. 1993 PhD thesis, University of Paris, unpublished.
Onaka, T. 1990 in ``The Galactic and Extragalactic Background Radiation'' (S. Bowyer and Ch. Leinert, eds.),
Dordrecht: Kluwer, p. 379.
Onaka, T., Tanaka, W., Watanabe, T., Watanabe, J., Yamaguchi, A., Nakagiri, M., Kodaira, K., Nakana, M.,
Sasaki, M., Tsujimura, T. & Yamashita, K. 1989 Astrophys. J. 342, 235.
Page, T. & Carruthers, G.R. 1981 Astrophys. J. 248, 906.
Page, T., Carruthers, G.R. & Heckathorn, H.M. 1982 ``Revised S201 Catalog of Far­Ultraviolet Objects'', NRL
Report 8487.
Paresce, F., McKee, C. & Bowyer, S. 1980 Astrophys. J. 240, 387.
Parker, J.W., Hill, J.K., Bohlin, R.C., O'Connell, R.W., Ne#, S.G., Roberts, M.S., Smith, A.M. & Stecher, T.P.
1996 Astroph. J. 472, 29.
Parsons, S.B., Wray, J.D., Kondo, Y., Henize, K.G. & Benedict, G.F. 1976, Astrophys. J. 203, 435.
Peacock, A., Verhoeve, P., Rando, N., van Dordrecht, A., Taylor, B.G., Erd, C., Perryman, M.A.C., Venn, R.,
Howlett, J., Goldie, D.J., Lumley, J. & Wallis, M. 1996 Nature 381, 135.
Peacock, A., Verhoeve, P., Rando, N., Perryman, M.A.C., Taylor, B.G. & Jakobsen, P. 1997 Astron. Astrophys.
Suppl. 123, 581.
Petit, H., Hua, C.T., Bersier, D. & Courtes, G. 1996 Astron. Astrophys. 309, 446.
Perryman, M.A.C., Foden, C.L. & Peacock, A. 1992 in ``Photon Detectors for Space Instrumentation'' (ESA­SP
356, p. 21.
Perryman, M.A.C., Foden, C.L. & Peacock, A. 1993 Nucl. Instr. and Meth. A325, 319.
Perryman, M.A.C., Peacock, A., Rando, N., Van Dordrecht, A., Videler, P. & Foden, C.L. 1994 in ``Frontiers of
Space and Ground­Based Astronomy'' (W. Wamsteker et al. , eds.), Dordrecht: Kluwer, p. 537.
Pica, A.J., Smith, E.P., Bohlin, R.C., O'Connell, R.W., Roberts, M.S., Smith, A.M., Fanelli, M.N. & Stecher,
T.P. 1993 BAAS 181 113.09P.
Pounds, K. & Wells, A.A. 1991 ``Astrophysics at FUV and EUV Wavelengths'' (B. Aschenbach, ed.) Adv. Space
Res. 11, 125.
Pounds, K.A., Allan, D.J. et al. 1993 Mon. Not. R. astr. Soc. 260, 77.
Priedhorsky, W., Bloch, J.J., Cordova, F.A., Smith, B.W., Ulibarri, M.D., Siegmund, O.H.W., Marshall, H.L.,
Vallerga, J.V & Vedder, P.W. 1991 in ``Extreme Ultraviolet Astronomy'', (R. Malina and S. Bowyer, eds.)
New York: Pergamon Press, p. 464.

52 NOAH BROSCH
Pye, J.P. 1995 in ``Astrophysics in the Extreme Ultraviolet'' (S. Bowyer & R.F. Malina, eds.) Dordrecht: Kluwer
Academic Publishers, pp. 1­4.
Pye, J.P., McGale, P.A., Allan, D.J., Barber, C.R., Bertram, D., Denby, M., Page, C.G., Ricketts, M.J., Steward,
B.C. & West, R.G. 1995 Mon. Not. R. astr. Soc. 274, 1165.
Reichen, M., Kaufman, M., Blecha, A., Golay, M. & Huguenin, D. 1994 Astron. Astrophys. Suppl 106, 523.
Rifatto, A., Longo, G. & Capaccioli, M. 1995a Astron. Astrophys. Suppl. 109, 341.
Rifatto, A., Longo, G. & Capaccioli, M. 1995b Astron. Astrophys. Suppl. 114, 527.
Roach, F.E. & Gordon, J.L. 1973 ``The Light of the Night Sky'', Dordrecht: Reidel Publishing Co.
Roussel­Dupr’e, D.C. & Ameduri, F.P. 1993 Proc. SPIE 1742, p. 621.
Roussel­Dupr’e, D.C. & Bloch, J.J. 1996 in ``Cataclysmic variables and related objects'' (A. Evans & J.H. Wood,
eds.), Dordrecht: Kluwer Academic Publishers, p. 257.
Roussel­Dupr’e, D.C., Bloch, J.J., Theiler, J., Pfafman, T. & Beauchesne, B. 1996 in ``Astronomical Data Analysis
Software and Systems V'' ASP Conference Series 101 (G.H. Jacoby & J. Barnes, eds.), pp. 112­115.
Russell, J.L., Lasker, B.M., McLean, B.J., Sturch, C.R. & Jenkner, H. 1990 Astron. J. 99, 2059.
Salpeter, E.E. 1955 Astrophys. J. 121, 161.
Sandel, B.R., Shemansky, D.E. & Broadfoot, A.L. 1979, Astrophys. J. 227, 808.
Sasseen, T.P., Lampton, M., Bowyer, S. & Wu, X. 1995 Astrophys. J. 447, 630.
Sasseen, T.P. & Deharveng, J.­M. 1996 Astrophys. J. 469, 691.
Schmidt, E.G. & Carruthers, G.R. 1993a Astrophys. J. 408, 484 (erratum 1994 Astrophys. J. 435, 920).
Schmidt, E.G. & Carruthers, G.R. 1993b Astrophys. J. Suppl 89, 259.
Schmidt, E.G. & Carruthers, G.R. 1995 Astrophys. J. Suppl 96, 605.
Schuerman, D.W., Weinberg, J.L. & Beeson, D.E. 1997 BAAS 9, 313.
Siegmund, O.H.W., Gummin, M.A., Stock, J. & Marsh, D. 1992 in ``Photon Detectors for Space Instrmentation'',
ESA SP­356, pp. 89­96.
Smith, A.M. 1967 Astrophys. J. 147, 158.
Smith, A.M. & Cornett, R.H. 1982 Astrophys. J. 261, 1.
Smith, A.M., Cornett, R.H. & Hill, R.S. 1987 Astrophys. J. 320, 609.
Smith, E.P., O'Connell, R.W., Bohlin, R.C., Cheng, K.­P., Cornett, R.H., Hill, J.K., Hill, R.S., Hintzen, P.,
Landsman, W.B., Ne#, S.G¿, Roberts, M.S., Smith, A.M. & Stecher, T.P. 1992 Astrophys. J. Lett. 395, L49.
Smith, E.P., Bohlin, R.C., O'Connell, R.W., Roberts, M.S., Smith, A.M. & Stecher, T.P. 1995 BAAS 186, 07.03S.
Smith, E.P., Pica, A.J., Bohlin, R.C., Cornett, R.H., Fanelli, M.N., Landsman, W.B., O'Connell, R.W., Roberts,
M.S., Smith, A.M. & Stecher, T.P. 1996 Astrophys. J. Suppl. 104, 287.
Smith, D.A., Ne#, S.G., Bothun, G.D., Fanelli, M.N., O#enberg, J.D., Waller, W.H., Bohlin, R.C., O'Connell,
R.W., Roberts, M.S., Smith, A.M. & Stecher, T.P. 1996b Astrophys. J. 473, L21.
Stecher, T.P. & Milligan, J.E. 1962 Astrophys. J. 136, 1.
Stecher, T.P., Baker, G.R. et al. 1992 Astrophys. J. 395, L1.
Storchi­Bergmann, T., Calzetti, D. & Kinney, A.L. 1994 Astrophys. J. 429, 572.
Taresch, G., Kudritzki, R.P., Hurwitz, M., Bowyer, S., Pauldrach, A.W.A., Puls, J., Butler, K., Lennon, D.J. &
Haser, S.M. 1997 A&A 321, 531.
Terebizh, V.Yu. 1986 Contributions of Special Astrophys. Obs. 50, 79.
Toller, G. 1983 Astrophys. J. Lett. 266, L79.
Toller, G., Tanabe, H. & Weinberg, J.L. 1987 Astron. Astrophys. 188, 24.
Thompson, G.I., Nandy, K., Jamar, C., Monfils, A., Houziaux, L., Carnochan, D.J. & Wilson, R. 1978 ``Catalog
of Stellar Ultraviolet Fluxes'', SRC.
Tovmassian, H.M., Khodajayants, Yu.M., Krmoyan, M.N., Kashin, A.L., Zakharian, A.Z., Hovhannessian, R.Kh.,
Mkrthian, M.A., Tovmassian, G.H., Huguenin, D., Bootov, V.V., Romanenko, Yu.V., Laveikin, A.I. & Alexan­
drov, A.P. 1988 Pis'ma v Astron. Zh. 14, 291. (Sov. Astron. Lett. 14, 123).
Tovmassian, H.M., Khodjayants, Y.M., Krymoyan, M.N., Isadjanian, R.V., Gasparian, O.N., Zakharian, A.Z. &
Huguenin, D. 1991a in ``Astronomy from Wide­Field Imaging'' (H.T. MacGillivray et al. , eds.), p. 55.
Tovmassian, H.T., Hovhannessian, R.Kh., Epremian, R.A. & Huguenin, D. 1991b in ``Astronomy from Wide­Field
Imaging'' (H.T. MacGillivray et al. ., eds.), p. 473.
Tovmassian, H.M., Oganesyan, R.Kh. Epremian, R.A., Huguenin, D., Viktorenko, A.S. & Serebrov, A.A. 1991c
Sov. Astron. 35, 470­478.
Tovmassian, H.M., Lorenz, H., Priebe, A., Richter, G.M., Schmidt, K.­H., Khodzaiants, Iu.M., Krmoian, M.N.,
Isadzanian, R.V. & Zakharian, A.Z. 1991d AN 312, 281.
Tovmassian, H.M., Hovhannessian, R.Kh., Epremian, R.A., Huguenin, D., Khodjayants, Y.M., Krmoyan, M.N.,
Kashin, A.L., Alexandrov, A.P. & Romanenko, Y.V. 1992 Astrophys. Space Sci. 188, 205.
Tovmassian, H.M., Oganesyan, R.Kh. Epremian, R.A. & Huguenin, D. 1993a Astron. J. 106, 627.
Tovmassian, H.M., Hovhannessian, R.Kh., Epremian, R.A. & Huguenin, D. 1993b Astron. Astrophys. Suppl. 100,
501.
Tovmassian, H.M., Hovhannessian, R.Kh. & Epremian, R.A. 1994 Astrophys. Space Sci. 213, 175.
Tovmassian, H.M., Hovhannessian, R.Kh., Epremian, R.A. & Huguenin, D. 1994 Mon. Not. R. astr. Soc. 266,

UV SKY SURVEYS 53
337.
Tovmassian, H.M., Navarro, S.G., Cardona, O., Hovhannessian, R.Kh. & Epremian, R.A. 1996a Astron. J. 111,
299.
Tovmassian, H.M., Navarro, S.G. & Cardona, O. 1996b Astron. J. 111, 306.
Treyer, M.A., Ellis, R.S., Milliard, B. & Donas, J. 1997 in ``The UV Universe at Low and High redshits'', AIP
Press, in press.
Ulmer, M.P., Razeghi, M. & Bigan, E. 1995 in ``Optoelectronic integrated circuits, materials, physics, and devices''
(M. Razeghi, Y.­S. Park & G.L. Witt, eds.) SPIE 2397, pp. 210­217.
Vallerga, J. & Lampton, M. 1987 in ``Optoelectronic Technologies for Remote Sensing from Space'' SPIE 868 ,
p.25.
Vallerga, J., Veeder, P.W. & Welsh, B.Y. 1993 Astrophys. J. 414, L65.
Vallerga, J., Siegmind, O.H., Dalcomo, J. & Jelinsky, P.N. 1997 in ``Solid State Sensor Arrays: Development and
Applications'' (M.J. Blouke, ed.) SPIE 3019, pp. 156­167.
Van Duinen, R.J., Aalders, J.W.G., Wesselius, P.R., Wildeman, K.J., Wu, C.C., Luinge, W. & Snel, D. 1975
Astron. Astrophys. 39, 159.
Vennes, S., Chayer, P., Hurwitz, M. & Bowyer, S. 1996 Astrophys. J. 468, 898.
Vuillemin, A., Buat, V., Donas, J., Milliard, B., Courtes, G. & Golay, M. 1991 Adv. Space Res. 11, 143.
Yamashita, K. 1968 Astrophys. Space Sci. 2, 4.
Zvereva, A.M., Severny, A.B., Granitsky, L.V., Hua, C.T., Cruvellier, P. & Courtes, G. 1982 Astron. Astrophys.
116, 312.
Waller, W.H., Marsh, M., Bohlin, R.C., Cornett, R.H., Dixon, W.V., Isensee, J.E., Murthy, J., O'Connell, R.W.,
Roberts, M.S., Smith, A.M. & Stecher, T.P. 1995 Astron. J. 110, 1255.
Waller, W.H., Bohlin, R.C., Cornett, R.H., Fanelli, M.N., Freedman, W.L., Hill, J.K., Madore, B.F., Ne#, S.G.,
O#enberg, J.D., O'Connell, R.W., Roberts, M.S., Smith, A.M. & Stecher, T.P. 1997 Astrophys. J. 481, 169.
Weller, C. 1983 Astrophys. J. 268, 899.
Wesselius, P.R., van Duinen, R.J., Aalders, J.W.G. & Kester, D. 1980 Astron. Astrophys. 85, 221.
Wesselius, P.R., van Duinen, R.J., de Jonge, A.R.W., Aalders, J.W.G., Luinge, W. & Wildeman, K.J. 1982 Astron.
Astrophys. Suppl. 49, 427.
White, N. 1991 in ``Extreme Ultraviolet Astronomy'' (R. Malina and S. Bowyer, eds.), New York: Pergamon Press,
p. 15.
Wilkinson, E., Green, J.C., McLean, R. & Welsh, B. 1996 BAAS 188, 5802.
Witt, A.N., Friedmann, B.C. & Sasseen, T.P. 1997 Astrophys. J. 481, 809.
Wu, C.­C., Ake, T.B., Boggess, A., Bohlin, R.C., Imho#, C.L., Holm, A.V., Levay, Z.G., Panek, R.J., Schi#er,
F.H. & Turnrose, B.E. 1983 ``The IUE Ultraviolet Spectral Atlas'', IUE NASA Newsletter No. 22.
Wu, C.­C., Crenshaw, D.M., Blackwell, J.H., Wilson­Diaz, D. Schi#er, F.H., Burstein, D., Fanelli, M.N. &
O'Connell, R.W. ``IUE Ultraviolet Spectral Atlas'', IUE NASA Newsletter No. 43.