Figure 1: Overview on the brightness of the sky
outside the lower terrestrial atmosphere and at high ecliptic and
galactic latitudes. The zodiacal emission and scattering as well as
the integrated light of stars are given for the South Ecliptic
Pole (
, 
). The bright magnitude
cut-off for the stellar component is V = 6.0 mag for 0.3 - 1 
m.
In the infrared, stars brighter than
15 Jy between 1.25 and 4.85 m and brighter than
85 Jy at 12 m are excluded.
No cut-off was applied to the UV data, 
0.3 m.
The interstellar cirrus component
is normalized for a column density of 1020 H-atoms cm-2 corresponding
to a visual extinction of 0.053 mag. This is close to the values at the
darkest patches in the sky.
Source for the long-wavelength data, 
1.25 m,
are COBE DIRBE and FIRAS
measurements as presented by Désert
et al. (1996). The IR cirrus spectrum is according to the model of
Désert et al. (1990) fitted to IRAS photometry.
The short-wavelength data, 1.0 m, are from the following
sources:
zodiacal light: Leinert & Grün (1990);
integrated starlight: 0.3 m, Gondhalekar
(1990), 
m, Mattila
(1980); cirrus: 
= 0.15 m, Haikala et
al. (1995), 
m, Mattila &
Schnur (1990), Mattila (1979). The
geocoronal Lyman 
(121.6 nm) and the OI(130.4, 135.6 nm) line
intensities were as measured with the Faint Object Camera of the Hubble
Space Telescope at a height of 610 km (Caulet et al.
1994). The various references for the airglow emission
can be found in Sect. 6 (click here)
This paper is concerned with the night sky brightness from the far UV
(
100 nm) to the far infrared ( 200 m).
Quite a few sources contribute to the diffuse brightness of the
moonless sky (
) in this wavelength range:
-airglow from the upper atmosphere (
).
-Zodiacal light, both as scattered sunlight and thermal emission
of interplanetary dust particles, from interplanetary space
(
). (In the far UV interplanetary Ly emission is
important.)
-Integrated starlight (
) of the stars not individually
accounted for
-diffuse galactic light (
), in the UV and visual mainly
reflections off interstellar dust particles. Their infrared thermal emission
is known as "cirrus'' since the pioneering IRAS observations.
It dominates the sky brightness in the far-infrared. Interstellar gas
contributes line emissions over all of
our wavelength range.
-Extragalactic background light (
) in addition to the
radiation of individually detected galaxies.
The combined light of these radiations is attenuated by atmospheric
extinction, while tropospheric scattering of the infalling flux
adds a non-negligible brightness component (
).
Formally, the above statements may be expressed as
It should be noted that the "extinction coefficient'' 
(which
depends on wavelength , zenith distance z, height of the
observer and change of the atmospheric conditions with time) for
diffuse sources has a value different from that determined for stars.
The scattered light not only contains additional contributions
due to stars and galaxies otherwise accounted for individually,
but, increasingly more important, the light pollution due to the
ever-growing man-made lighting.
For space observations atmospheric extinction and scattering are irrelevant, but other complexities like instrumental stray light of lunar, terrestrial or solar radiation may arise. For low orbits, spacecraft-induced glow phenomena may be present.
Quite understandably then, extracting accurate brightness values from Eq. (1 (click here)) is a difficult task, and the past has seen a measure of disagreement between individual determinations. In the following we want to summarise what consensus has been obtained in this field during the last years, in order to provide a basis for easier reference and comparability.
The aim of this article is to provide the reader with comparatively easy access to agreed-upon or at least recommended values of night sky brightness. Inevitably this requires smoothing and interpolating of data. Therefore we want to give at the same time sufficient information on original publications to give an impression on the grade of agreement or disagreement of the available data and to allow the reader who wants to do so to draw his own conclusions.
We will go through the components basically in the order in which they appear in Eq. (1 (click here)), and for each component try to provide information on the visual, infrared and ultraviolet wavelength ranges.