The dark count level was low, averaging
counts s
Å
and 
counts s
Å
, respectively,
for the two detector voltages used during the mission. The dark count
increased by as much as a factor of 30 during passages through the South
Atlantic Anomaly, but remained too low to have a significant effect on
observations of bright targets planned during these periods.
Due to its large apertures, HUT is particularly sensitive to airglow.
Flying near solar minimum for Astro-2 led to a dramatic decrease in many
airglow lines relative to the near-solar maximum conditions of Astro-1.
Optically thick lines such as Ly
and O I 
were
reduced by a factor of two to three in intensity relative to Astro-1.
Typical orbital night intensities for Ly
were 
1.5 kR
(1 Rayleigh = 
) with variations
depending on the particular line of sight.
Optically thin lines such as O I 
were down by nearly
an order of magnitude relative to Astro-1.
Most airglow lines are weak enough that they only affect regions of the spectrum
corresponding to the width of the aperture used for a particular observation.
For the frequently used 
aperture this is 
15 Å.
Line profiles derived from observations of blank fields suffice to characterize
and remove these lines when rescaled to the observation of interest.
Light from geocoronal Ly
scattered within the
spectrograph is a more serious problem.
While the holographically ruled grating has extremely low scattering
properties, 
of the total intensity far from
the line center, the strength of the Ly
airglow emission gives
a scattered component to the background that is comparable to the detector
dark count during orbital night. During orbital day and in the near wings
of Ly
it dominates the background.
Observations of blank fields during orbital night were summed and fitted with
a smoothly varying function to derive the Ly
scattering profile for
each aperture used during the Astro-2 mission.
Figure 6 shows the scattering profile obtained for 7146 s
of integration on blank sky in four separate pointings
through the 
aperture.
Since the cores of Ly
and of O I 
are affected
by dead-time corrections and a continuously decreasing QDE, the best match
for removing the scattered Ly
profile is obtained by the scaling the
model profile to the near wings in the 1159--1211 Å and 1221--1273 Å
ranges.