The airglow emissions vary considerably
with time, on short (minutes) and long timescales, mainly due to changes in
the atmosphere and in solar activity. They also depend on
geomagnetic latitude, with a distinctive tropical
brightness enhancement. The brightness values given
below therefore are only indicative of the typical intensities.
Many of the airglow emissions arise in the ionospheric E layer at
90 km, some in the F region above 150 km (see Fig. 18),
some, like Ly
and H in the
Geocorona. The phenomenological side of airglow, which is the part of
interest for the night sky brightness, has for the visual region in large part
been studied in the
sixties and seventies, which reflects in the list of references.
Typical brightness values
of main airglow lines are summarised in Table 13.
Figure 18: A typical height profile of airglow volume emission, as measured
from the satellite OGO II. The peak near 90 km is due to OH emission,
the extended peak at higher altitudes to [OI] emission at 630 nm.
From Reed & Blamont (1967)
| Source | Wavelength | Height of | Intensityb |
| emitting layer | |||
Ly | 102.6 nm | geocorona | 10 R |
|
Ly | 121.6 nm | geocorona | 3 kR(night) - 34 kR(day) |
| OI | 130.4 nm | 250 -300 km | 40 R (in tropical airglow) |
| OI | 135.6 nm | 250 - 300 km | 30 R (in tropical airglow) |
| O2 (Herzberg bands) | 300 nm - 400 nm | 90 km | 0.8 R/Å |
| [OI] | 557.7 nm | 90 km | 250 R |
| Na D | 589.0 nm, 589.6 nm | 92 km | 30 R (summer) |
| to 100 R (winter) | |||
| [OI] | 630.0 nm | 250 - 300 km | 60 R |
| [OI] | 636.4 nm | 250 - 300 km | 20 R |
|
H | 656.3 nm | geocorona | 4-6 R (night) |
| pseudocontinuum | 400 nm - 700 nm | 90 km | 0.3 R/Å |
| O2 | 864.5 nm | 80 km | 1 kR |
| OH | 600 nm - 4.5  m | 85 km | 4.5 MR(all bands) |
Broadfoot & Kendall (1968) give the spectrum of the airglow
from 300 nm to 1 m (see Fig. 20). It is based on photoelectric
observations at Kitt Peak near zenith and within 30
of
the galactic pole. The spectral resolution is 5 Å, the scan step
four times smaller. The [OI] lines at 630 nm and 636.4 nm and also
H are weaker than average in these observations.
Figure 19: Spectra of the nightglow from 800 Å to 1400 Å
at 3 Å resolution. The data were obtained from the space shuttle at an
altitude of 358 km on December 5, 1990. Two spectra
are shown, of which the upper one was taken closer to the dusk
terminator. It therefore also shows OII 834 and HeI 584 (in second order),
which are features belonging to the dayglow. The zenith distance
was 85 and
90 for the upper and lower spectrum, respectively.
Ly is a geocoronal line. The continuum at 911 Å is due
to O+ recombination to the ground state.
From Feldman et al. (1992)
Figure 20: Spectrum of the airglow from 300 nm to 1 m (from
Broadfoot & Kendall 1968)
Ultraviolet astronomical observations mostly are taken from above the
atmosphere by rockets or satellites. In this context it is relevant
to know the airglow as seen from such spacecraft positions.
Results obtained at typical
altitudes are shown in Figs. 19 and 21.
The strength of the main emission lines
is also summarised in Table 13. For the OI 130.4 nm and 135.6 nm lines
enhanced values observed in the tropical airglow (Barth & Schaffner
1976) are given. At mid latitudes they are less intense by about one
order of magnitude. Apart from the main emission lines shown in Fig. 19,
the ultraviolet region between 850 Å and 1400 Å is thought
to be free of nightglow emission.
The viewing line of spacecraft on the night side of the atmosphere may cross
the terminator and continue through the sunlit parts of the atmosphere.
Under these twilight conditions, dayglow features become important.
E.g. the NO 
bands then are excited by resonance fluorescence
and then are much stronger,
the N2 Lyman-Birge-Hopfield bands are clearly visible,
and the forbidden [OII] emission at 247 nm is strong. Figure 22 shows
ultraviolet airglow emission observed under such conditions.
An excellent review on observations and modelling of both
dayglow and nightglow ultraviolet emissions has been given by
Meier (1991).
Figure 21: Left: Spectrum of the nightglow from 1250 Å to 1700 Å
at 17 Å resolution. The data were obtained from the space shuttle
at a height of 330 km in January 1986 at minimum solar activity.
The oxygen OI lines at 1304 Å and 1356 Å are the brightest features.
For the weakly visible Lyman-Birge-Hopfield bands the dashed curve
shows a predicted spectrum.
Right: Spectrum of the ultraviolet nightglow from 170 nm to 310 nm
at 29 Å resolution obtained on the same flight.
The solid line shows an appropriately scaled solar spectrum
and is assumed to show the contribution to zodiacal light. From
Morrison et al. (1992)
Figure 22: Ultraviolet twilight airglow spectrum, as observed during
a rocket flight on September 24, 1979. Left: from 1200 Å
to 1500 Å at 20 Å resolution. Ly is at left. "LBH"
refers to the Lyman-Birge-Hopfield bands. These observations were done
in the height range 100 km - 200 km. - Right: From 170 nm to 310 nm
at 25 Å resolution. The dotted line shows the zodiacal light contribution.
These observations refer to rocket heights of 170 km - 246 km. -
The field of view of the experiment was oriented 23 from the sun and essentially in the horizontal plane (0.2 elevation).
For conversion to absolute fluxes, a solid line is given with both parts of
the figure. It indicates which signal would be produced at each wavelength by
a monochromatic source of a given brightness (100 R for the short-wavelength
part, 18 R for the longer wavelengths). For continuum emission this
would correspond to 5.0 R/Å and 0.72 R/Å, respectively. From
Cebula & Feldman (1982, 1984)
From 1 m to 3 m, OH in a layer around 90 km height
dominates the airglow emission. There is a gap in the OH
spectrum around 2.4 m (see Fig. 27) which is important for balloon
observations and also for the low background observations possible
from Antarctica (see Sect. 4.3 (click here)).
Seen from the ground, longward of 2.5 airglow is
only a small addition to the thermal emission from the troposphere
(compare Fig. 11 in Sect. 4 (click here) above).
Figures 25 and 26 show the near-infrared OH spectrum at two resolutions,
once with a low spectral resolution of 
= 160 Å,
and once with a higher resolution of 
. Wavelength lists and intensities for the individual OH bands
can be found in Ramsay et al. (1992) and Oliva & Origlia
(1992). Obviously, the near-infrared airglow is dominated by the OH
bands. They primarily also determine the night sky brightness in the
J (1.2 m) and H (1.6 m) bands (Fig. 11, Sect. 4.3 (click here)).
In absence of atmospheric extinction, a thin homogeneously
emitting layer at height h above the Earth's surface shows a
brightness increase towards the horizon, which is given by the
so-called van Rhijn function
where R = 6378 km is the radius of the earth. E.g., for h = 100 km
results (Roach & Meinel 1955).
This situation typically
applies for balloon experiments. Figure 23 shows an example. For observations
from the ground, extinction and scattering change the behaviour
in particular for zenith distances > 40. Around
a maximum airglow increase by about a
factor of about four
may be expected at 
, with the brightness
decreasing again towards the horizon (see Fig. 24 for an observation
and Roach & Meinel (1955) for a selection of predicted
profiles). For shorter wavelengths, with stronger scattering and extinction,
this decrease starts already at higher elevations. However, appropriate
models (based on realistic assumptions, including multiple scattering
in a spherical atmosphere and going down to
the horizon) to account for the observed brightness profiles from the
zenith to the horizon have not yet been calculated. The results given
in Sect. 5 (click here) do not claim to be accurate near the
horizon.
Figure 23: Increase of airglow brightness at 2.1 m towards the horizon
observed from a balloon at 30 km altitude on October 23, 1972. Dots
represent the measurements, the line gives the van Rhijn function for a
height of the emitting layer of 92 km. From Hofmann et al. (1977)
Figure 24: Zenith angle dependence of sky brightness observed at 530 nm
from Mt. Haleakala, Hawaii (Kwon et al. 1991).
The points represent an average
normalised profile. The thin lines are the curves predicted by
Barbier in 1944 for heights of the airglow emitting layer of
50 km (higher maximum) and 200 km, respectively. The solid line fitting the data
is an ad-hoc modification of Barbier's formula
Figure 25: Near-infrared airglow spectrum as seen from the ground at 160 Å
resolution (for > 1.2 m). The OH bands mainly
contributing to the emission have been identified in the figure.
"differential" simply means "per micron". From
Harrison & Kendall (1973)
Figure 26: Near-infrared airglow spectrum as observed from Mauna Kea
at spectral resolution . In regions with atmospheric transmission 
0.75 the flux
has been arbitrarily set to zero. Longward of
2.1 m thermal atmospheric emission takes over. Note that 1000 of the
units used correspond to 
, 
,
and 
W/m2srm at 1.25 m, 1.65 m
and 2.2 m, respectively. From
Ramsay et al. (1992)
Figure 27: Spectral distribution of near-infrared zenith airglow showing
the gap in airglow emission around 2.4 m. The airglow measurements
have been performed from a balloon at 30 km altitude during flights
in 1972 and 1974. Variations from flight to flight and during one
night were less than a factor of two. From Hofmann et al. (1977)
Figure 28: Correlation between the diffuse sky emission at 467 nm
(Strömgren b) and at = 525 nm. The brightness variations
in both bands are mainly due to airglow. From Leinert et al.
(1995)
Figure 29: Variation of OH airglow, observed from Mauna Kea.
Left: Short term variations (minutes) caused by the passage of wavelike
structures. Right: Decrease of OH airglow during the course
of a night, shown for several bands separately. From Ramsay et al.
(1992)
Airglow emission is often patchy and varying in brightness and spatial distribution with time. Roach & Gordon (1973) demonstrate this by showing airglow maps in time steps of 15 minutes on the right upper corner of odd pages, thus enabling a "thumb-cinema" look at these spatio-temporal variations. Quantitative examples for variation during one night or variation with solar cycle can be seen in Figs. 8 and 10 in Sect. 4 (click here). Often a systematic decrease of airglow emission during the course of the night is observed, explained as result of the energy stored during day in the respective atmospheric layers.
Figure 29 shows this for the OH emissions and also gives an example for the wavelike structures often apparent in these emissions.
These examples do not give at all a full overview on airglow variability but just demonstrate that it is a typical property of this source of night sky brightness.
In the visual spectral region, correlations between the prominent [OI] and NaD airglow emission lines and "pseudocontinuum" bands at 367 nm, 440 nm, 526 nm, 558 nm, 634 nm and 670 nm have been studied by Barbier (1956) who established three "covariance groups". E.g., the correlation between the 557.7 nm line and the "pseudocontinuum" at 502 nm has been used by Dumont (1965) to eliminate the airglow contribution from his zodiacal light measurements. Sometimes such correlations can be quite tight (see Fig. 28).
Above 1000 km, the earths atmosphere changes to a
composition of mainly neutral hydrogen with some
ionised helium, the density falling off gradually over a few earth radii.
Two telling images of the geocorona in Ly,
including the globe of the earth, are shown by
Frank et al. (1985, see p. 63).
This geocorona is optically thick to the solar Lyman lines. Typical
intensities of the emissions observed from ground (in the visual)
or from earth orbit are given in Table 13, with the data taken from
Caulet et al. (1994) and Raurden et al. (1986) for
Ly, Meier et al. (1977) for Ly, Levasseur
et al. (1976) for H.
Solar radiation is scattered by neutral interstellar gas atoms
which are coming from the solar apex direction and are pervading the
solar system until ionized. The emitting region is a sort of cone
around the apex-Sun line. The observed emission depends on the
position of a spacecraft with respect to this cone (see, e.g. the
review by Thomas 1978). Typical patterns observed for the
Ly and He 584 Å lines are shown in Fig. 30.
Figure 30: Interplanetary emission in the Ly (left) and He 584 Å
lines (right) observed by the Mariner 10 UV spectrometer (Broadfoot
& Kumar 1978). The observations were performed on January 28, 1974,
while the spacecraft was at a heliocentric distance of
0.76 AU and 
from the apex-Sun axis. The
brightness units are Rayleighs. From Thomas (1978)
Depending on altitude and solar activity, satellites produce additional light emissions by interaction with the upper atmosphere (Shuttle glow). Photometric measurements thus may be affected. These light phenomena are relatively strong in the red and near-infrared spectral regions, but are noticeable towards the ultraviolet as well.
For instance, during the Spacelab 1 mission the emissions of the
N2 Lyman-Birge-Hopfield bands were found to be in the range of
(Torr et al. 1985). These observations at
250 km altitude were performed under conditions of moderate solar activity.
During minimum solar activity and at 330 km, Morrison et al.
(1992) observed no such emissions. The GAUSS camera onboard the
German Spacelab mission D2 (296 km, moderate solar activity) observed a
patchy glow with 
W m-2sr-1nm-1
at 210 nm (Schmidtobreick 1997).
Taking into account the appropriate conversion factor,
the observed glow intensity amounts to about 0.4 R/Å in its
brightest parts. Although these three observations were made at
somewhat different wavelengths, the overall increase of emission
intensity I with surrounding air density 
is in agreement with
an 
law.
