THE EUVE GUEST OBSERVER PROGRAM
                               Anne B. Miller
           EUVE Guest Observer Program, Center for EUV Astrophysics,
                 2150 Kittredge St., University of California,
                       Berkeley, California 94720, USA


ABSTRACT

   The EUVE Spectrometer is operated under the EUVE Guest Observer Program by
NASA and UC Berkeley.  The Spectrometer covers the wavelength range of 70-760
A in three overlapping bands at a resolution of 100-400.  Guest observers
receive time-tagged photon event lists, and instrument and spacecraft house-
keeping data.  Data is delivered with software tools that run in the Image
Reduction and Analysis Facility and are tailored for handling EUVE satellite
data.  Processing software sorts all data into a set of  time-ordered tables
that can be listed or displayed graphically.  Instrumental effects are removed
from the data, and photon events are tagged with corrected position and wave-
length, producing imagelike event lists which preserve arrival times and other
information.  Software for data quality selection, spectrum extraction, time-
series processing and spectral model comparison can be used directly on the
event list files.


1. INTRODUCTION

   The EUVE science payload includes four imaging systems and a three-channel
extreme ultraviolet (EUV) spectrometer [1].  After completion of the EUVE all-
sky survey, the balance of the mission is being conducted as a Guest Observer
program, under which the astronomical community may propose to make pointed
spectroscopic observations in the EUV.  The EUVE Guest Observer (EGO) Program
is supported by the EGO Center at the Center for Extreme Ultraviolet Astrophy-
sics, the University of California, Berkeley.  (For information about the EGO
Program, contact Dr. Carol Christian, Center for EUV Astrophysics, 2150 Kit-
tredge St., Berkeley, CA 94720, USA email:  egoinfo@cea.berkeley.edu).  The EGO
Program may extend to 3 years or more.


2. THE EUVE SPECTROMETER

   The spectrometer is a radially symmetric slitless objective design based on
variable line-space grazing-incidence reflection gratings [2].  Spectra are
accumulated simultaneously in three bandpasses from 70-760 A.  The instrument
is pictured in cross section in fig. 1, and its physical characteristics are
summarized in table 1.
   The telescope is a Wolter-Schwarzschild type II grazing-incidence design
with gold-coated surfaces [3].  It is f/3.3, with a total geometric area of
over 450 cm^2.  Each of three spectrometer channels has a 2 deg field of view
and receives light from one-sixth of the telescope mirror through a section in
the annular front aperture.  This division gives each channel a geometric area
of ~75 cm^2.  The remaining light feeds a parallel imaging system.  The medium
and long wavelength channels have wire-grid collimators which limit the spec-
tral width of diffuse light to 20 arc minutes, FWHM.
   After the mirror, each converging beam strikes one of three variable line-
space gratings, placed at 120 deg intervals around the optic axis, which pro-
duce nearly stigmatic spectra using straight, mechanically ruled grooves [4].
The short wavelength grating produces first order spectra from 70 to 190 A,
the medium wavelength from 140 to 380 A, and the long wavelength from 280 to
760 A.  The gratings are coated with rhodium or platinum  to enhance grazing-
incidence reflectivity.
   The spectra are focussed onto three microchannel plate detectors, which
provide two-dimensional imaging and time-tagging of individual photons -- via
electron amplification [5], [6].  Each event position is calculated by instru-
ment software from the division of a charge cloud on a divided anode.  Thin
film filters, completely covering each detector, define broad bandpasses and
screen out diffuse geocoronal and interplanetary light [7].
   Fig. 2 shows the minimum detectable flux in photons cm^(-2) sec^(-1) A^(-1)
for each of the spectrometer channels.  This function combines the effects of
the spectrometer's first order efficiency and interference from local geocoro-
nal airglow [8].


3. PROPOSALS AND DATA ACQUISITION

   The EGO Program is a cooperative effort between NASA and UC Berkeley to
support all activities of the mission's pointed phase.  Announcements of op-
portunity are made yearly, and proposals are submitted to the EUVE NASA project
office.  The practicality of proposed  observations is evaluated by EGO Center
scientists and scientific merit is judged by a NASA peer review committee.
   The EGO Center and EUVE operations staff collaborate  with NASA Flight Oper-
ations Team (FOT) on observation planning and data acquisition.  Observations
are scheduled by Berkeley and all science plans are reviewed by the FOT.  Slews
and instrument operations are commanded from the EUVE Science Operations Center
(ESOC) at Berkeley.  The ESOC staff also monitor observations and instrument
health during real-time data transmissions.  They are advised by EGO Center
scientists, who act on the observer's behalf if problems arise.  To assist GOs
in making simultaneous observations with other instruments, the EGO Center
notifies each principal investigator of the planned date and time of their
observations via electronic mail shortly before execution.
   Most spectroscopic data is obtained during orbit nighttime, in periods of
20 to 33 minutes.  Data acquisition for a 40,000-second integration takes about
two days.  Satellite telemetry is received by the Tracking and Data Relay Sa-
tellite System satellite and relayed to the ESOC by NASA's Goddard Space Flight
Center.  The data are decommutated at the EGO Center and delivered to Guest
Observers in a format compatible with  the Image Reduction and Analysis Fa-
cility (IRAF) and other systems that support FITS images.


4. EGO PROGRAM DATA PREPROCESSING

   Spectrometer data are preprocessed in two stages using the IRAF/EUV packages
developed by the EGO center.  First, the telemetry is restructured to separate
photon events from instrument and spacecraft housekeeping data.  All data are
logged in IRAF ST (Space Telescope) format tables, which preserve the time or-
der of events.  In this tabular form, each engineering monitor can be displayed
as a time-series, much like a mechanical strip chart tape.  This accessibility
is central to the goal of letting each GO process only the data that meet their
science criteria.  Second, photon events are re-sorted into imagelike quick
position-oriented event (QPOE) files, one for each Spectrometer channel.  Each
QPOE file contains a header with global information, followed by a time-tagged
list of photon event records, sorted in position order.  Although QPOEs can be
manipulated and displayed like two-dimensional IRAF images, they contain more
information.  The form of EUVE event records is specified by a flexible pipe-
line program, which uses a Spectrometer calibration database to calculate wave-
length and corrected position for each photon.  The pipeline can be set up to
include other information, including raw positions, pulse height information,
and spacecraft aspect, to name a few.
   Baseline spectra are extracted for each GO using standard apertures and
background subtraction techniques.  A typical spectrum for the calibration
target AU Mic in the medium wavelength channel is pictured in fig. 3, binned
at the size of a resolution element.


5. GUEST OBSERVER DATA ANALYSIS

   GOs may come to Berkeley to use the IRAF/EUV packages, or they may receive
a copy of the software to use in their home IRAF system.  All spectral reduc-
tion tools provided by the EGO Center are interactive IRAF tasks with on-line
documentation, designed to take advantage of the flexibility offered by the
IRAF/QPOE interface.  Tools are provided for data selection, extracting and
combining spectra, doing wavelength calibration, and creating simulated EUVE
spectra from models.

5.1 Data Selection

   EGO Center data reduction software and storage are designed to enable each
observer to use all information in the telemetry that relates to his scientific
data.  The storage of housekeeping data in tables and the retention of time
stamps in QPOE files creates the potential for observers to filter their data
according to instrumental and observing conditions.  Data selection is imple-
mented through the interactive IRAF/EUV task dqselect.  This task allows the
GO to plot any combination of the  engineering monitors as time series, and to
specify an allowable range of values for each monitor.  Using these limits,
dqselect creates a list of "good time" intervals in the data set, when all the
monitors are within range.  These time intervals can then be applied as a macro
to the QPOE event lists using qpmedit, a macro-editing task.  Subsequent pro-
cessing will affect only the desired subset of the data.
   For example, one of the first steps in handling any data set is to create a
filter that excludes events from periods with either very high or very low
count rates.  When the filter is applied to the QPOE before processing, each
photon time stamp is compared to the list of good times and must fall into one
of the intervals to pass.  This removes data from periods when the detectors
were turned off or flooded with high background.  The time filter itself de-
fines the effective integration time of the observation, which is critical if
accurate flux measurements are to be obtained.

5.2 Studying Time-varying Events

   Because the time-stamp for each photon is preserved in the QPOE file, the
event list can also be rebinned in time, to display variations in brightness.
Time series are created with user-specified resolution using the IRAF/EUV task
qpbin, which can be combined with region masks to isolate variations in a sin-
gle spectral feature.  In the following example, the author studied the daily
variations in the diffuse light level from geocoronally excited HeII at 304 A.
First, a time filter and region mask were created to select nighttime 304 A
events in an off-spectrum region from the medium wavelength QPOE file of a
calibration observation.  Second, qpbin was used to make a time series over the
selected region at thirty-second resolution.  A similar series was made for a
region of the same size that sees only constant detector noise and charged par-
ticle detections, and was subtracted from the 304 A series as background.  The
result is a one-dimensional IRAF image showing the modulation of the He II
count rate over the entire observation, and is shown in the top section of fig.
4.  The gaps are caused by detector turnoff during orbit daytimes.  The back-
ground series is shown in the bottom section of the figure.
   The  magnetic latitude of the spacecraft as a function of time was then
calculated using the task backmon, which produces a number of geographic func-
tions of the ephemeris.  The latitude is shown in the middle section of fig. 4.
   Comparison shows a strong anti-correlation between the HeII 304 A count rate
and the magnetic latitude, with variations in the geocoronal intensity of over
+/- 20% around an average of 60-65 counts per second.  Charged particle detec-
tions correlate directly with magnetic latitude but occur at a much lower level.
A GO who wishes to reduce geocoronal background at 304 A might select data from
higher magnetic latitudes, accepting the risk of raising the overall noise from
particle detections by a smaller amount.


REFERENCES

1. B. Haisch, S. Bowyer, and R.F. Malina, "The Extreme Ultraviolet Explorer
	Mission", JBIS, this volume (1993).
2. M.C. Hettrick, S. Bowyer, R.F. Malina, C. Martin, and S. Mrowka, "The Extreme
	Ultraviolet Explorer Spectrometer", Appl. Opt., 24, 1737 (1985).
3. S. Bowyer and J. Green, "The Fabrication, Evaluation and Performance of
	Machined Metal Grazing Incidence Telescopes", Appl. Opt., 27(8), 1414
	(1988).
4. M.C. Hettrick, "Aberrations of Varied Line-Space Grazing Incidence Gratings
	in Converging Light Beams", Appl. Opt., 23, 3221 (1984).
5. O.H.W. Siegmund, R.F. Malina, K. Coburn, and D. Werthimer, "Microchannel
	Plate EUV Detectors for the Extreme Ultraviolet Explorer", IEEE Trans.
	Nucl. Sci., NS-31, 776 (1984).
6. J.V. Vallerga, G.C. Kaplan, O.H.W. Siegmund, M. Lampton, and R.F. Malina,
	"Imaging Characteristics of the Extreme Ultraviolet Explorer Micro-
	channel Plate Detectors", IEEE Trans. Nucl. Sci., 36(1), 881 (1989).
7. J.V. Vallerga, O.H.W. Siegmund, E. Everman, and P. Jelinsky, "The Calibra-
	tion of Thin Film Filters to be Used on the Extreme Ultraviolet Explor-
	er Satellite", Proc. SPIE, 689, 138 (1986).
8. S. Chakrabarti, R. Kimble, and S. Bowyer, "Spectroscopy of the EUV (350-1400
	A) Nightglow", J. Geophysical Research, 89, no. A7, 5660-5664 (1984).


FIGURE CAPTIONS

Figure 1:  The EUVE Spectrometer in cross section.
Figure 2:  Minimum detectable flux/A in the EUVE spectrometer.  Solid lines:
	short and long wavelengths; dashed line:  medium wavelengths.
Figure 3:  Medium wavelength spectrum of AU Mic, approximately 140-380 A.
Figure 4:  Variations in count rate for HeII 304 A emission (top), magnetic
	latitude (middle), and charged particle detections (bottom) during a
	calibration observation.