NOTES FOR USERS OF UNIFIED ABSTRACT FILE DATA
FROM THE ULTRAVIOLET NITRIC OXIDE (UVNO) EXPERIMENT
ON ATMOSPHERE EXPLORER C (AE-C)
A. EXPERIMENT RATIONALE
The purpose of the UVNO experiment is to measure the
distribution of, and variations in, nitric oxide molecules in
the altitude range 100-250 km. Nitric oxide is produced in
this region by the reaction of N(4S) and N(2D) with O2 and is
removed by reaction with N(4S) or, less importantly, by diffusion.
The atomic nitrogen is produced by the dissociation of N2
following the absorption of EUV sunlight by the atmosphere.
The distribution of NO depends on temperature (through the strong
temperature dependence of the reaction between N(4S) and O2),
on composition (through the competition between the reaction between
N(2D) and O2 and the quenching of N(2D) by O), and on transport
processes.
B. INSTRUMENT SUMMARY
The UVNO instrument is a fixed-grating Ebert-Fastie spectrometer
which accepts light in a 15A bandpass centered on the (1,0) Gamma
band of NO at 2149A. It measures sunlight which has been resonantly
scattered in this band by ambient NO molecules. The
detector is a photomultiplier with a sapphire window and a CsTe
cathode; its output is fed into pulse-counting electronics.
The line of sight is perpendicular to the S/C spin axis, which
in turn is parallel to the orbit normal vector. The field of view
is 4 x l/4 (degrees), with the long axis parallel to the S/C spin axis
and therefore parallel to the viewed limb. On despun orbits
the instrument looks perpendicular to the line joining the
center of the earth and the S/C, in the wake direction. On
spinning orbits, the instrument scans the limb of the Earth
once per spin. Its sampling time is 21 msec, giving an angular
smearing (at 4 rpm) of 1/2 degree.
C. DATA WORDS
UVNO entries in the AE-C Unified Abstract File consist of
5 words per 15 sec:
#67 - H(SC) The NO scale height at the S/C altitude (km)
#68 - NO(SC) The NO density at the S/C altitude (/cm**3)
#69 - NO(150) The NO density at 150 km (/cm**3)
#70 - NO(M) The maximum NO density observed during a given
limb scan (/cm**3)
#71 - Z(M) The altitude at which this maximum occurs (km)
See Section I for more details.
D. DATA CATALOG
Attached is a catalog of those February 1974 orbits for which
UVNO data is available. On most orbits measurements exist for
S/C altitudes less than about 250 km. On spinning orbits,
measurements of NO at 150 km may exist at S/C altitudes up to
350 km.
E. CALIBRATION
The instrument was calibrated before launch in the laboratory.
A diffusing screen was illuminated by light from a Xenon lamp
which had been passed through a monochromator. The screen was
then measured by the instrument and by a field-limited standard
phototube. In this way the sensitivity to an extended source
was measured as a function of wavelength. In flight, the
detector sensitivity was periodically monitored using an
internal lamp. No direct monitoring of the reflectivity of the
optics was performed, but no systematic changes in the signal
levels produced by viewing the Rayleigh scattering atmosphere were
observed during the period between launch (in December 1973) and
February 1974.
F. RELIABILITY OF DATA
1. Despun orbits. On despun orbits, the accuracy of UVNO
results depends on the success of the background subtraction
technique (see Section I), but more importantly on the accuracy
of the S/C altitude information. On an unknown number of orbits,
the instrument pointing was subject to about 2 1/2 degrees error (due to an
ambiguity in the S/C ACS system); on these orbits, the "true"
NO density would be about 60% larger than those entered in the UAF. A
list of orbits on which this problem was known to occur is
attached. (The UAF entries have not been adjusted).
2. Spinning orbits. No attitude errors exist because the
UVNO detection of the limbs of the earth provided pointing
information to a very high accuracy. However, the curve-fitting
algorithm used to analyze spinning-orbit data (section 1)
strongly weights the low-altitude, strong-signal region at the
expense of the spacecraft-altitude, weak-signal region; so the
reported NO density at S/C altitude, and the NO scale height
(words 68 and 67) are not reliable.
G. EVENTS AFFECTING DATA
Except as noted in (F) above, the UVNO entries in the UAF
have already been screened for spurious or misleading results.
H. ANOMALIES IN DATA
See (G) above.
I. DATA REDUCTION
1. Noise removal. The cosmic-ray and energetic particle
environment of the experiment in space produced obvious spikes
in about 20% of the samples. These were removed by rejecting
samples lying above a 2*sigma limit calculated from 5 neighboring
samples. The remainder of the spurious counts were assumed to
be random, and were removed (along with the genuine "dark
counts") by background subtraction techniques.
2. Temperature and optical thickness correction.
The response of the instrument to a slant column containing
N mols/cm**2 at some temperature T is
C = INTEGRAL(S*g dN) counts/21msec
N=0 to N
where g is the number of solar photons scattered per molecule per
second, and S is the instrument sensitivity. A theoretical study
of the scattering and detection processes, using the measured
instrument slit-function, was performed. The results can be
approximated by
S*g = P/(Q+N) where P = 10*(70+T)
Q = P*5.7E12*(1+T**2/3.8E6)
Thus C = P*ln(1+N/Q)
or, rearranging, N = Q*[exp(C/P)-1] (1)
3. Despun orbits. For despun orbits, the data was averaged
in 15 second periods after noise removal. The background count
level was obtained from measuruments at high altitudes on downleg
and upleg, and subtracted; the remaining NO signal was rejected if
it was less than half the subtracted background. NO was assumed
to be distributed according to
NO(Z) = NOsc * exp{ -(Z-Zsc)/Hsc }
where Hsc was chosen to be 33 km (representative of actual values).
Then
N = NOsc * Hsc * Ch(Re+Zsc,H,THETA) (2)
in which Ch(R,H,THETA) is the Chapman function and THETA (the look angle
measured from the zenith) is obtained from S/C attitude data. NOsc
is then obtained from (1) and (2). T was obtained from the formula
T = 900 - 550*exp{ -(Zsc-120)/50 } (3)
The error introduced by assuming this value of T rather than a
measured value is small compared to other uncertainties. The
parameters NO150, NOm and Zm were not obtained on despun
orbits.
4. Spinning orbits. On spinning orbits the background
count level was measured over a 30 degree arc near the zenith, and
subtracted. At altitudes below Zsc, NO was assumed to form a
layer given by
NO(Z) = 2*NOm*[exp{-(Z-Zm)/Hsc} - exp{-2*(Z-Zm)/Hsc}/2]
= 2*NOm*[b*x - (b*x)**2/2] (4)
in which: NOm, Hsc, and Zm are to be determined;
x = exp{-(Z-Zo)/Hsc} where Zo= 90 km
b = exp{(Zm-Zo)/H) = 1/xm
The observed slant column density is then
N(Zk) = 2*NOm * sqrt{2*pi*RH} * [b*x - (b*x)**2 / 2*sqrt(2)] (5)
where Zk is now the look altitude (the lowest altitude reached
by the instrument line-of-sight). From (1) and (5) we have
exp(C/P) = 1+ 1/Q * 2*NOm * sqrt{2*pi*RH} * [b*x - (b*x)**2 / 2*sqrt(2)] (6)
The instrument counts C at each look altitude Zk sampled during
the limb-scan were fittted to the expression
exp(C/P) = 1 + A*x - B*x**2
by a least squares process in which Hsc was varied. Then the
parameters NOm and Zm were readily obtained from A and B, using
(6). The exponential expression (3) for T was used (with Zsc
replaced by Zk) above 120 km, and replaced by
T = 175 + 175 * [(Zk-85)/35]**2
for values of Zk less than 120 km. The actual curve-fitting process
was more complex than implied above. It assumed that NO = 0
wherever expression (4) is negative, and it also included an
iterative correction for Rayleigh scattering contamination of the
signal below 100 km. The parameters NOsc and NO150 were obtained
from expression (4) using the derived values of NOm, Hsc and Zm.
At altitudes so high (>200 km) that the above fitting
process failed due to poor altitude resolution, it was still
possible to find NO150 by assuming an exponential distribution
of NO and analyzing only the data for which Zk lay between 120
and 180 km. This procedure generally failed above about 350 km.
USEFUL REFERENCES
A. Dalgarno, W.B. Hanson, N.W. Spencer, and E.M. Schmerling,
"The Atmosphere Explorer Mission," Radio Sci., 8, 263-266,
1973 .
C.A. Barth, D.W. Ruseh and A. I. Stewart, "The UVNO Experiment
for Atmosphere Explorer," Radio Sci., 8, 279-386, 1973.