Data Set Overview
  =================
    The Midcourse Space Experiment (MSX) sampled the mid-infrared
    thermal emission from the zodiacal dust cloud in spectral bands
    centered at 8.3, 12, 15, and 21 microns, with a variety of
    experiments during a 9 month period in 1996.  The experiments and
    their results are described in
 
    Price, S.D., P.V. Noah, D. Mizuno, R.G. Walker, and S. Jayaraman
    2003.  Midcourse Space Experiment Mid-infrared measurements of the
    thermal emission from the zodiacal dust cloud.  Astron. J. 125,
    962-983.  [PRICEETAL2003]
 
    These observations were obtained with the MSX Spatial Infrared
    Imaging Telescope (SPIRIT III).  Data from four different
    experiments are included in this data set, as described below
    (descriptions excerpted from [PRICEETAL2003]).  Stars were removed
    and the detectors combined as described in Price et al. 2003.  The
    file naming convention is as follows:  The four character
    abbreviation for the experiment (in parentheses after each
    experiment below) is followed by the two digit observation number
    within the experiment.  Following this is the letter indicating
    the band, A, C, D, or E.  The B band observations are omitted due
    to non-detection.  Finally a digit following the band letter is
    used to distinguish multiple observations in the same band.
 
    Celestial Background Experiment 1 (CB01)
 
    CB01 obtained five ecliptic pole-to-pole scans that came within
    25-30 degrees of the sun.  The scans started a few degrees behind
    the north ecliptic pole and scanned in ecliptic latitude at
    constant ecliptic longitude (epoch 1996.5) terminating at the
    south ecliptic pole.  The CB01 scans near the sun were sandwiched
    between the highest priority experiments conducted after the
    initial commissioning phase of the mission, and mid-summer of
    1996, when the spacecraft was no longer in eclipse for part of the
    orbit.  The earth was used as an occulting disk to block the sun
    for these observations,
 
    Celestial Background Experiment 4 (CB04)
 
    CB04 simultaneously measured the zodiacal background as it covered
    the two gaps in sky coverage left by the IRAS survey with a set of
    61 scans.  Collectively, the CB04 scans sampled sun-centered
    longitudes between 98 and 295 degrees.   The scans started a few
    degrees behind the north ecliptic pole, but scanned at constant
    1983.5 epoch longitude and ended at 70 degrees south ecliptic
    latitude.  Thus, the focal plane swept a swath 1 degree in
    longitude (at the ecliptic plane) at a nominal scan rate of 0.125
    degrees/sec.  The last 31 scans of CB04 were executed at half this
    scan rate to partially compensate for the higher noise caused by
    increasing focal plane temperature toward the end of the mission.
 
    Lunar Off-Axis Response Experiment (DC32)
 
    The DC32 experiment plan was to obtain a series of seven scans at
    the same Sun-centered longitude, nominally within 15 degrees of
    the Moon, over two and a half days; the middle scan crossed the
    Moon, while the other scans were on either side of it.  The first
    and last scans in the series were programmed to be far enough away
    from the Moon that they can be used to define the zodiacal
    background for the experiment.  Six such experiments were
    executed, four against the full Moon and two against the
    three-quarter Moon at solar elongations of ~180 and ~90 degrees
    respectively.  The last experiment against the full Moon was taken
    late in the mission and was not included in the database because
    of excessive noise.  Each observation consisted of a down
    (decreasing latitude) and up scan at an approximately constant
    Sun-centered longitude at a scan rate of 0.06 degrees/sec.  Unlike
    the CB01 and CB04 experiments, these scans were executed with the
    columns parallel to the scan direction.  Thus, a very narrow swath
    was swept out in longitude but with a very large number of samples
    within a given latitude interval.
 
    Zodiacal observation in the Mirror Scan Mode (EM10)
 
    This experiment scanned 170 degrees along the ecliptic plane, from
    a sun-centered longitude of ~160 degrees to ~330 degrees, at a
    rate of 0.3 degrees/sec.  The sensor was repositioned to ecliptic
    latitude of -50 degrees and a latitude profile obtained at a
    sun-centered longitude of ~329 degrees.  These data were taken in
    mirror scan mode, with the mirror scan direction perpendicular to
    the spacecraft motion, that is, in latitude for the scan along the
    plane and in longitude for the latitude scan.  The amplitude of
    the mirror scan was 1.5 degrees.  The geometry was such that
    contiguous coverage was obtained along the center line of the
    scan.

Confidence Level Overview
  =========================
    A number of factors influence the absolute accuracy of the
    measured zodiacal radiance:  the absolute radiance calibration of
    the sensor, the uncertainty in the dark offset measurement, the
    biases introduced by the dark offset model, the temperature of the
    focal plane array, and factors external to the telescope such as
    the non-rejected Earth radiation (NRER).  These factors are
    examined in detail as they apply to the present data set, in the
    appendix of Price et al. (2003) q.v.
 
    The dark offset analysis of the EM10 experiment observations were
    based on a comparison between observations of the same ecliptic
    longitudes from different experiments (CB01 and CB04)  coupled
    with a model of the zodiacal background to decouple the effects of
    differing observing epochs between the different experiments
    (section 4.1 of Price, Noah, Mizuno, Walker, and Jayaraman, AJ
    125, 962-983, 2003).  The dark offset calibration for the CB01 and
    CB04 experiments also used observations from overlapping
    observations, but from within each experiment,  with no
    modifications due to the change in the relative position of the
    spacecraft during the epoch of the experiment.  As the solar
    elongation changes when observing a single point near the plane of
    the ecliptic, the column density and temperature of interplanetary
    dust changes and can result in a variation in brightness
    background of order 1% on timescales of 12 hours (c.f., Hauser et
    al., Ap J Letters, 278, L15-L18, 1984).This result suggests that
    the dark offset matrices derived assuming a constant zodiacal
    background in the overlapping regions within a given  CB
    experiment may be in error.  Although the effect on the data from
    not taking into account this small change in the zodiacal
    background  is probably small, this notice is included to make the
    user aware of the assumptions used in the dark offset data
    analysis procedures.