JOURNAL OF GEOPHYSICAL
RESEARCH, VOL. 85, NO. A13, PAGES 7861-7870, DECEMBER
30, 1980
Morphologyof the VenusUltraviolet Night Airglow
A. I. F. STEWART,J.-C. GERARD,l D. W. RUSCH, AND S. W. BOUGHER
Laboratoryfor Atmospheric
and SpacePhysicsandDepartmentof Astrogeophysics
Universityof Colorado,Boulder,Colorado80309
Imagesof thenightside
of Venusin the(0,1)/Jbandof nitricoxidehavebeenobtainedby the Pioneer
Venusorbiterultravioletspectrometer
(OUVS).Theemission,
whichisproduced
by radiativeassociation
of N and O, showsa brightspotreaching-•5 kR andlocatedat 2 a.m.localsolartimejust southof the
equator.The emittinglayeris at 111+ 7-kmaltitude.A onedimensional
verticaltransportmodelshows
that the hemisphericaveragebrightness
of 0.8 kR is consistent
with the orbiterneutralmassspectrometer
(ONMS)measurements
of N andO near167km,andthatthealtitudeof theemittinglayeris consistent
with the eddymixingmodelproposedto explainthe daysideheliumprofilemeasuredby the busneutral
massspectrometer.
In themodel,N reaches
a peakof 7 x 108cm-3 at 114km, andO reaches
a peakof
2.6 x 10• cm-3 at 106km. Thereis a fair degreeof consistency
betweenthe ONMS, OUVS, and other
airglow measurements,
exceptas regardsthe local time dependence.
INTRODUCTION
The ultraviolet night airglow of Venus was first detectedby
Barth et al. [1968] from Mariner 5. Subsequently,the visual
Hemberg II and Chamberlain bands of molecular oxygen
were seen from Venera 9 and 10 [Krasnoœolsky
et al., 1976;
Lawrenceet al., 1977;S!angerand Black, 1978]. The ultraviolet nightglowwas again detectedfrom PioneerVenus [Stewart
et al., 1979] and subsequentlyidentified as the gamma and
delta bands of nitric oxide [Feldrnanet al., 1979; Stewart and
Barth, 1979] excited by the radiative associationof atoms of
nitrogen and oxygen. Since in the spectrumof the night airglow the emissions
that might be expectedto accompanythe
dissociativeproductionof O and N are weak or absent [Stewart and Barth, 1979; Stewart, 1980], it is evident that the night
airglow is a result of the transportof atomsfrom the dayside
by the generalcirculationof the thermosphere.
Stewart and Barth [1979] reported that the maximum emission apparentlyoccurrednear the equator at about 2 a.m. local solar time and that bright patcheswere occasionallyseen
elsewhere.Since then, the precessionof the orbit of the Pioneer Venus spacecraftthrough local time has allowed the ultraviolet spectrometerto obtain numerousimagesof the dark
side of Venus showing the spatial and temporal variation of
the emission. In the present paper we describe these images
and their analysis.We alsodeterminethe altitude of the emitting layer, using data from an earlier phase of the mission
when periapsiswas on the nightside.
We investigate the production of the emission using a
simpleone dimensionalsteadystatemodelof the transportof
N and O downward through the nightsidethermosphere.The
comparisonof this model with our observationsconfirmsthe
plausibilityof the conceptof the daysideorigin of the atoms
and confirmsalso that vertical mixing on the nightsideis very
vigorous.
ultraviolet spectrometer observations of the planet [Stewart,
1980]. The essentialcharacteristicsare presentedin Figure 1.
The slant range from the spacecraftto the limb seenby the ul-
traviolet spectrometervariesfrom 31900 km at P- 2:03 (2 h,
3 min before periapsis)to 11700 km at P- 0:33, and from
6100 km at P - 0:16 to < 1000 km at periapsisto 4100 km at P
+ 0:10. The planet is in view for a maximum of 36ø of spin
during the first view period and for a maximum of 186ø during the second.
The imagesthat concernus here were obtained from orbit
167 (May 20, 1979) throughorbit 210 (July 2, 1979), during
which period the subspacecraft
point at 30øN (---P - 0:55)
moved from 2225 to 0300 local time. The images were
acquiredin a seriesof swaths,normally 36 s but occasionally
24 s apart. The integrationperiod usedwas 8 ms, giving an
angular resolutionacrossthe planet of 0.23ø. Resolutionin
the north-south direction was determined by the length of the
slit, 1.38ø; there was considerable overlap of successive
swaths.The wavelengths
usedwere 1980/• (the 0,1 delta
band)and2365/• (the0,1gammaband).We concentrate
on
the 1980-/•images,
whicharelesssubject
to contamination
by
internally scatteredlight near the terminatorsthan are the
2365-/• images.
Images obtained during the secondview period were obtained with a 32-ms integration period (0.94ø of spin). Because of the lower spacecraftaltitude, successiveswaths do
not overlap, and in addition, the changesin range are much
more severe.Several such imageswere analyzed by Stewart
and Barth [1979];none are consideredhere. However, swaths
from these images taken near periapsiswere used to determine the altitude of the emitting layer. Near periapsisthe long
axis of the slit aligns with the limb and this, combined with
the low altitude and short slant ranges,givesgood altitude resolution on the limb. Although •<5-km resolution is possible
usinga 4 msintegrationperiod,smalluncertainties
in the data
timing preventan absolutealtitudedeterminationat this reso-
OBSERVATIONS
The orbit of the Pioneer Venus spacecraft has been describedelsewhere[Colin, 1979a, b], as has the geometry of the
lution. The method used was to selectswathsthat clearly show
both limbs in the limb-brightened airglow emission,and to resolve the uncertaintiesin timing by assumingthat the emis-
•On leave from Institut d'Astrophysique,Universit6 de Liege,
Liege, Belgium.
sion layer altitude is the same on both limbs. The disadvantage here is the 32-ms integration period, which
degradesthe altitude resolutionto --•20 km. The remaining
uncertainty in the emissionaltitude arises from the small
Copyright¸ 1980by the AmericanGeophysicalUnion.
Paper number 80A0731.
0148-0227/80/080A-0731501.00
7861
7862
STEWART ET AL.' MORPHOLOGY OF VENUS UV NIGHT AIRGLOW
l
m
m
mm •
•
ill
m
m
i
m
m
Plate1. Unrectified
imageat 1980A obtainedonorbit 186.The colorcodeis shownat thelowerleft andthescaleis l
kR per color level. The sunlit crescentis set to 15 kR. The blue spotsoutsidethe planet discare due to instrumentbackground.
STEWART ET AL.: MORPHOLOGY
Plate 2.
OF VENUS UV
NIGHT
AIRGLOW
Consecutiveunrectifiedimagesobservedbetweenorbits 181 and 196 The empty casescorrespondto orbits when
no NO • or ¾observations
wereavailable.The colorscaleis the sameasin Plate1.
7863
7864
STEWART ET AL.: MORPHOLOGY OF VENUS UV NIGHT AIRGLOW
P-33
<•--•
P-123
The color scaleis the same as in Plate 1. The displacementof
the terminatorwith respectto the limb is due to the revolution
of the planet about the sun by 1.60ø/24 hours. Occasional
dark stripesacrossthe discare due to a lossof telemetrydata
P
or to scansmade in other wavelengthsthan the NO &(0, 1)
band. Large variationsof the airglow are readily seenboth in
intensity and morphology. From these and other images, we
find no clear periodicity in either the average or the peak
brightness(cf. Figure 2). The morning sectorbright spot is
Fig. 1. Drawing (not to scale) of the OUVS observinggeometry. usually present, and the equator is generally brighter than
The dotted line representspart of a typical trace acrossthe disc (the other latitudes.Random bright patchesare seento occurat all
remainder is on the other side of the planet). The dashedlines representthe envelopesof the endpointsof thesetraces,i.e., the limb of the latitudes and local times, with the exception of the mid-latitude regionnear the eveningterminator (in the upper right of
planet as seenby OUVS.
P+IO
'•
these images).
Maps in eclipticlatitude and longitudecoordinates
have
changesin spin period and in spin axis orientation that occur
near periapsisdue to aerodynamic drag and to the deploy-
been made for orbits 167 to 210, following the procedure outlined below. Each individual monochromatic image is first
ment of the radar antenna.
correctedfor the emissionangle. The angle from nadir is calThe presenceof the diffuselyreflectingcloud layer beneath culated for each image element, and the correspondinginthe airglow layer enhancesthe observedsignalat wavelengths tensity is convertedinto a vertical emissionrate by applying
>2000 A, but at 1980A the single-scattering
albedoof CO2is the appropriateChapmanfunction;the topsidescaleheightof
0.25, and the enhancement is less than 10%. The instrument
the airglow distributionis taken to be 5 kin, and the peak of
the layer is assumedto be locatedat 110 kin. The image elementsare subsequentlysortedout into latitude and longitude
RESULTS
bins of 3ø x 4.8ø size, respectively.Ecliptic longitudesare fiApoapsisimagesobtainedbetweenMay 20 and July 2, 1979, nally convertedinto hour angleby shiftingthe longitudescale
in the wavelength of the (1,0) & band have been analyzed to by an amountdependingon the date of observation.The map
study the Venus nitric oxide nightglow emission. Plate 1 obtained on orbit 186 is shown in Plate 3, superimposedon a
showsan exampleof a falsecolorunrectified
imageat 1980A latitude and hour angle grid. In this casethe color scaleis 400
obtained during orbit 186 on June 8, 1979. The intensity scale R per color level from 0 to 6 kR, and as before,intensitiesexhas 15 levelsranging from 1 (dark blue) to 15 (bright yellow) ceeding the maximum value have been set to color level 15.
kR. The color scaleis displayedat the lower left of the image. Becauseof the viewing geometry the grid cannot be entirely
In this particular representation,intensitiesequal to or larger filled in since the most southern latitude seen by the OUVS
than 15 kR are set to the 15th color level. The yellow crescent instrumentis 60ø and other parts of the planet are not reached
near the top of the image correspondsto the sunlit portion of by the field of view. The high intensitylevelsalongthe 90øN
the discwhere the intensity of the scatteringof solar radiation latitude line and the 90ø and 270ø hour angle meridans are
on the dayside greatly exceedsthe 15-kR limit. The coordi- causedby contaminationof the observationsby scatteredlight
nates of the image are spin phase angle in abscissaand time originatingfrom the dayside.The centerof the bright airglow
from periapsisin ordinate. The tick marks along the vertical spotseenin Plate 1 is locatednear 15øSlatitude and 0130 LT.
axis correspondto 4-min time intervals from periapsisfrom P The fainter bright area near the limb in Plate 1 is also vis- 126 to P - 32 min. Some averaging of adjacent data points ible on the map, but the maximum emissionrate is no longer
has been performed in order to present the image in a 128 x
situated on the limb as a consequenceof the correction ap128 pixel format. The distorted aspectof the planet's disc is plied for the emissionangle.
causedby the motion of the spacecraftalong its inclined orbit
In order to investigatethe statisticalmorphologyof the NO
during the --•90min necessaryto scan the planet.
& nightglow,a latitude hour angle map was constructedfor
The disc of the planet is readily distinguishedfrom the each of 35 available images between orbits 167 and 210. The
background where the count rate is substantiallysmaller than daysideand the scatteredlight contributionswere removed
on the disc, indicating the presenceof the nightside airglow. from each map, and the resultingnightsidemaps were added
The airglow distribution is structured with emission rates together.The intensitywas then normalizedby the total numranging from lessthan I to 15 kR in this particular image. A
ber of entriescontributing to each bin of the final map. This
bright area is clearly visible in the southern hemisphere,and global map is shownin Plate 4 on a 200 R per color level
the attitude data indicatesthat it is locatednear the equator in
scale.As was expectedfrom the individual maps, it showsa
the morning sector.
bright patch reachinga maximum brightnessof 7.5 kR and
Since one apoapsisimage of the planet is obtained every 24 centerednear 10øS and 0200 LT. In addition, the equatorial
hours, the opportunity is given to examine the variability of regionin the premidnightand a high-latitudearea in the postthe airglowstructure.
A setof 14imagesat 1980A fromorbit midnight sectoralso exhibit secondaryintensity maxima. A
181 to 196 is displayedin Plate 2 in chronologicalorder. relativelydark regionnear40øN and 2000LT is alsoseen.The
The missingimagescorrespondto orbits 182 and 190 when Ly
areas located north of 75øN or having hour angles greater
a observationswere made. In order to insert the imagesinto a
than 60ø and less than 285ø have been excluded by the re128 x 128 array, each individual image was compressedlinmoval of the daylight. The averagevertical emissionrate deearly by a factor of 4. Consequently, 16 elementsof the un- duced from this map is 780 R for the NO & (0, 1) band, which
rectified imagesare averagedto make one element of Plate 2. correspondsto 2.7 kR for the total NO • system.
sensitivity
at 1980A is 1.5c/kR/32ms.
STEWARTET AL.: MORPHOLOGYOF VENUS UV NIGHT AIRGLOW
Since large day-to-day variations are observedin the air-
7865
where K is the eddy coefficient and H is the ambient scale
glow morphology,a systematicstudy of its variability was height.If we adoptthe form K-- A/M •/2 [vonZahn et al., this
made,usingthe setof orbitsincludedin Plate 4. The average issue],where M is the ambient number density and A is a conintensityin a rectangleboundedby 60øS and 66øN and by stant, then the solution to theseequationsis
2000 LT and 0400 LT was calculated. The result is shown in
Figure 2 by the curvelabelled 'dark disk average,'where the
dots representthe average emissionrate as a function of the
orbit number. The intensity variation of the brightest area
of the nightsideis also plotted at the top of Figure 2 on the
same intensity scale. The emissionrate is obtained by av-
eragingthe intensityovera 9ø of latitudex 15o of hour angle
area centeredon the brightestimage element. Finally, the
F10.7 cm solar flux is plotted at the bottom. The fluxes are
measured from the earth and shifted in date to account for the
fact that the two planetsare at differentsolarlongitude.This
delay is given in daysby AT = 27 a/360, where a is the difference in longitudebetweenthe two planets.
The curvesshow fluctuationsby almosta factor of 2 of the
averagebrightness.The variationsof the maximumbrightness
are evenlarger but generallyin phasewith the total nightside
intensity.Thesefluctuationsare statisticallysignificantdue to
the large number of countsincludedin each map. Each raw
image of the nightside contains .•104 counts, and consequently,the standarddeviationfor each point of the average brightnesscurve is on the order of 1%, much smaller than
the measuredday-to-day variations. The standard deviation
for the points of the maximum brightnesscurve is on the order of 5%. There is no evidence of correlation between the two
brightness
curvesand the solarflux, and in fact, the average
n ....
2qbH
K
• B.M
where B is a constantof integration(this solutioncan be verified by substitution).For a minor constituentflowing downward from a high-altitude sourceto sink at lower altitudes, the
appropriatevalue for B is zero. Thus for atomic oxygenintroducedby high-altitudehorizontal transportfrom the dayside, we have at altitudes above its recombination sink
[O]-- 21q,
oIH/g
where [O] and q•oare the number densityand vertical flux of
atomic oxygen,respectively.We assumethat this expression
describesthe oxygendensityat the altitudeswhereN is being
consumedby radiative associationwith O. A good rule of
thumb in problems of downward transport into a chemical
sink region is that the minor constituentreachesits maximum
density at the altitude where the transportand chemical lifetimes are equal [Rishbethand Barron, 1960];that is, for atomic
nitrogen, where
H2/K = (k[O])-•
Using our expressions
for [0] and K it is readily seenthat at
this altitude,
brightnesskeepsincreasingafter orbit 198 when the solar ac-
M = Me = A2/2klq,
oltP
tivity decreases.
This trend is causedby the changein local
time of the orbital plane sincethe early imagestend to emphasizethe eveningsector,whereasthe later onesmostlyinclude the morningside.The intensityof the brightestspotremains almost constantsince the bright area near 2 a.m. is
within the area coveredby all the imagesusedin Figure 2.
[o1 -- [Ol•-- (21q,
ol/kH) '/•
INTERPRETATION
We have used a one dimensional
chemical diffusion model
similar to that employed by Ruschand Cravens[1979] to solve
for the density distributionsof CO, N, O, 02, NO, and 03 on
the nightside of Venus. The upper and lower boundaries in
the model are at 140 km and 70 km, respectively.Table 1 details the chemical reactions employed and the adopted rate
coefficients.We compare the results of the model with the
OUVS airglow measurements of the emission in the nitric
oxide delta bands formed by reaction 1.
The downward flowing atomic nitrogen is consumedby radiative associationwith atomic oxygen, with a rate coefficient
k; the atomic oxygen subsequentlyrecombinesat a lower altitude. In this situation the N and O satisfy some simple relationships,which we demonstrateby using a one dimensional
analytic model. For simplicity, we assumethat the thermosphereis isothermal and that eddy transport greatly exceeds
transport by molecular diffusion. Then, in the absence of
sourcesand sinks,the vertical flux • and number densityn of
a minor constituentsatisfythe equation
d4'_ O
dz
and the chemical
lifetime
of N is
(r•), = (k[O],)-• = (H/2klrkol)•/2
Thus the altitude of the maximum N density dependson A
(i.e., on the strength of eddy mixing) and on the downward
flux of O, while the chemicallifetime of N at its peak depends
on the flux of O but not on the eddy mixing coefficient. For
example,for A = 2 x 10•3[yonZahnet al., 1979]and I½ol-- 1
X 10•2cm-2 s-l, we find M• = 1.5x 10•4cm-3, corresponding
to about 110 km; at this altitude, rs-- 8 x 104 s or about 1
day.
We have performeda detailedvertical transportcalculation
to extend the above arguments.We chose a vertical (down-
ward) flux of 1 x 10•2cm-2 s-• for O and 1 x 10•øcm-2 s-• for
N. The value for q•ois representativeof valuescalculatedby
Dickinsonand Ridley [1977], while the value for N is simply
scaledfrom this by the ratio of the N and O densities[Niemann et al., this issue].The resultsof the calculation are shown
in Figure 3 for a valueof K•4o-' 3 x 108cm2 s-• (i.e.,A -- 1.6
x 10•3).The atomic oxygen and nitrogen measurementsof
Niemann et al. [this issue]extrapolatedto 140 km assuming
diffusive equilibrium agree with our values there to within a
factor of 2. The atomic oxygenprofile has a maximum density
at 106km of 2.6 • 10• cm-3, and the maximumnitrogendensityis 7.3 x 10• cm-3 at 114km. The atomicoxygendensityat
the nitrogenpeak is 2 x 10• cm-3 in reasonableagreement
with the value 3 x 10• cm-3 predictedby our simpletheoretical treatment. The curve labeled P(O2) is the productionrate
of molecular oxygen by reaction 4 of Table 1. We will discuss
its importance later in this section.
7866
STEWART ET AL.' MORPHOLOGY
OF VENUS
UV
NIGHT
AIRGLOW
II
Plate 3. Mercator projectionof the image from orbit 186 (Plate 1). Lines are plotted every 30ø of latitude from 90øN
to 90øS and hour angle from 270ø (right) to 90ø (left). The color scaleis 400 R per color level and the yellow regionsare
contaminated by instrumentally scatteredlight.
STEWART ET AL..' MORPHOLOGY
OF VENUS
UV
NIGHT
AIRGLOW
Plate4. Map of the NO/• (0, 1) bandverticalintensitydistributionobtainedby summingup 35 orbits.The latitudeand
hour angle grid are similar to Plate 3. The color scaleis 200 R per color level.
7867
7868
STEWART ET AL.: MORPHOLOGY OF VENUS UV
PV OUVS'.Nitric Oxide8 (0,1) Band,1980
mize and the altitude
AIRGLOW
of the maximum
volume emission rate.
This dependenceis shown in Figure 4. The curve labeled
P(O2) representsthe altitude where the productionof 02 from
reaction 4 is maximum. The nitrogen density and nitric oxide
I0-
Brightest
Spot
5 -
NIGHT
emission rate maxima
are more sensitive to K than the atomic
oxygen and 02 production rate maxima. For the values of K
used the maximum density of N changesfrom 133 km to 107
km, a differenceof 26 km, while the O density maximum
Dork
changes from 117 to 102 km, a 15 km difference. Also the
curve for the delta band maximum lies parallel to that for the
nitrogen density. Thus a determination of the altitude at
which the nitric oxide emissionrate maximizesis a fairly sensitive indicator of the eddy diffusion coefficient.We have de-
D•sc
Averoge
I
254:
-
eeel
el
I
ee
terminedthe altitude of the emissionlayer near periapsison 8
orbits. The resultsare shown in Figure 5; the mean value is
150111 km with a standard deviation of 7 km. We use only the
I
I
21o
170
180
190
200
values obtained before periapsis,since the pulse of aerody16o
220
ORBIT
NUMBER
•'
namic drag at periapsis perturbs both the spin rate and the
Fig. 2. Variation of the average(lower curve)and maximum (up- spin axis orientation and renders the pointing information
per curve)brightness
in the (0, 1)&band of NO over 35 orbits.The suspectafter periapsis.Using Figure 4, this mean altitude imlowercurveis averagedoverall latitudesbetween60øSand 66øN and pliesKi4o-•3 x 108cm2 s-l, (,4 -• 1.6x 1013);
thisvalueis conall localtimesbetween8 p.m. and 4 a.m. The uppercurveis averaged
sistent
with
the
value
,4
=
2
x
1013
derived
by
vonZahn et al.
over a 9ø (latitude)x 15ø (hour angle)box containingthe brightest
[1979]
near
the
morning
terminator.
spot.
ße
20C
eeeeeeeeeeee
ß
lee#eel eel
eel
ß
F 10.7
ß
eeee
DISCUSSION
The volume emissionrate profile in the nitric oxide delta
bandspeaksat 110km with an intensityof 3.6 x 103ph cm-3
s-l. The integrated vertical intensity is 2.8 kR, which is in
good agreementwith the measuredaverageintensityon the
nightsideof 2.6 kR. In the model the efficiencyof nitrogen
atoms in producingexcited NO moleculesis determinedby
the fraction that are lost through reaction 1.
Nitrogen atomsare also lost by reactions2 and 3. Reaction
3 is a potential sourceof gamma and delta emission,but the
efficiencyfor production of an excited NO molecule is extremely small and we ignore its contribution.In an equilibrium calculation, half of the N atoms must be lost through reaction 2, the remainder by reactions 1 and 3. Reaction 1
accounts for 45% of the loss,and reaction 3 accountsfor 5% of
the lossin this model. Thus a photon is producedby 45% of N
atoms subsidingon the nightside.This efficiencyis independent of the N flux but is dependenton the strengthof the verti-
cal mixing. An increasein K resultsin a deeperpenetrationof
N atomsinto the atmosphereand a subsequent
increasein the
importanceof the three body reaction,reaction3 relativeto
reaction 1. Additional calculations show the efficiency of N
atomsin producingphoton changesfrom -•50% for Ki4o = 1
X 107 cm2 S--I to 36% for Ki4o -- 1 X 109cm2 S--l.
The strengthof the vertical mixing affectsthe altitude at
which the atomic oxygen and nitrogen densityprofilesmaxi-
TABLE
Reaction
1.
Table
Our observationshave shown that the nitric oxide night airglow typically showsa bright spot at about 2 a.m., that this
spot is -•5 times brighter than the hemisphericmean brightness,and that the emitting layer is near 111 km. Our one dimensionalsteadystatetransportmodel showsthat the zenith
emissionrate in the NO & systemis -•0.25 times the downward
flux of N atoms, that the column production rate of 02 is 0.5
times the downward flux of O (as expected),and that the altitude of the emitting NO layer dependson the flux of O and on
the strengthof eddy mixing. A first-orderinterpretation of our
measurementswould be that the brightnessmap is also a map
of the downward
flux of N. Since the N and O have similar
atomic weights, we might expect their velocity fields in the
thermosphericcirculation pattern to be similar, so that the N/
O ratio would be fairly uniform over the night hemisphere, at
least at altitudes above the chemical
sinks.
Several other data sets bear on this question. First, the
ONMS in situ measurementsof N and O [Niemann et al., this
issue]show that at 167 km, the N and O densitiesreach a minimum near 2 a.m., rather than a maximum as the above simple
interpretation suggests.However, the density scale heights
also reach a minimum near midnight, and if the ONMS measurementsare extrapolated to 140 km (the upper boundary of
our model), we find that both the N and O densitiesrise from
dusk to maxima near 10 p.m. and decline toward dawn; the
of Chemical
Reactions
Rate Coefficient
Source
(1) O + N -• NO
(2) N + NO -• N2+ O
(3) O + N + CO2-• NO + CO2
(4) O+ O+CO2•->O2+CO 2
(5) O + CO + CO2-• CO2+ CO2
2.5 X 10--17cm3 S-1
2.1X 10-ll cm3 s-I
1.1X 10-32(T/300)-1/2cm6$--1
4.7X 10-33(T/300)-2cm6s-l
2 x 10-37cm6s-l
Mandelrnan
etal. [1973]
Phillips
andSchiff[1962]
Baulchetal. [1973]
Campbell
andGray[1973]
Sirnonaites
andHeicklen
[1972]
(6) O + 02 + CO2-• 03 + CO2
(7) O + 03 -• 02 + 02
1.4x 10-33(T•v/300)-2'5cm6s-l
Kaufman[1969]
1.9x 10-ll exp(-2300/7) cm3s-1 Harnpson
[1973]
STEWART ET AL.' MORPHOLOGY
OF VENUS UV
I01
102
103
AIRGLOW
ALTITUDE
PRODUCTION
RATE(CM-;5
SEC-I)
i00
140
NIGHT
104
7869
OF EMI'I-I'ING
LAYER
105
130
--
120
--
130
'•
120
--
::3 i10
LU 110--
i00
-J
9O
8O
107
, , , ,,
108
D9
DIO
100--
12
90--
I011
10
13
13
14
10
'-• OF SAMPLES
I01•
DENSITY(CM-3)
I
-8
Fig. 3. Calculated profiles of the densitiesof atomic oxygen and
atomic nitrogen. The volume emissionrate of the NO & bands is due
to reaction 1, and the production of 02 moleculesis due to reaction 4
for an eddydiffusioncoefficient
of 3 x 108cm2 s-' at 140km.
-6
!
-4
!
1
-2
0
TIME FROM PERIAPSIS (MIN)
Fig. 5. Measurements of the altitude of the emitting layer,
groupedin 1-min bins. At P- 6.5 min, the limbs are at 35øN; at P .5 min, they are at 12øN.
maxima are quite pronounced in that the l0 p.m. densities
("•1 X 108cm-3 and --1 x 10'ø cm-3 for N and O) exceed
thoseat 8 p.m. and at 2 a.m. by a factor of about 2. From this
we would anticipate the peak in the airglow emissionto occur
before midnight rather than after.
Second, the Herzberg II emission from molecular oxygen
observedfrom Venera 9 [Krasnopolskyet al., 1976] shows a
peaknear 11 p.m. and a steadydeclinefromthat time until 4
a.m. This is consistentwith the extrapolatedONMS measurements of O but not with the average NO brightnessmap. We
may note, however, that we occasionallysee NO airglow
brightnessdistributionssimilar to the Venera distribution;on
orbit 184,for example,the bright spotlies at or slightly before
midnight.It would be of greatinterestto know what the average distributionwasfor the severaldawn-to-duskpassesmade
by Venera 9. The averageintensityof--5 kR seenby Venera 9
a value an order of magnitude smaller than this, but the uncertainty is large.
A third relevant observation is the detection of the 1.27 /•
emissionfrom molecular oxygen on the nightside of Venus by
Connes et al. [1979]. Their field of view lay in the postmidnight sectorand the mean zenith intensity was --600 kR.
If we assumea unit efficiency for the production of 1.27 /•
photons upon recombination of molecular oxygen, then the
necessary
flux of O is 1.2 x 10'2 cm-2 s-', entirelycompatible
with our model and, by implication, with the ONMS measurementsalso. Thus given the simplicity of our model and
the possibilityof other sourcesfor the 02 emissions[Conneset
al., 1979] there is a fair degree of consistencybetween the
measured O and N densities and the measured NO and 02
represents
a quantumefficiencyof--10 -2 photons/Oatom, airglow intensities.
givenour O flux of 1 x 10'2 cm-2 s-'. A laboratorydetermia,
aThe discrepanciesbetween the local time dependenciesof
tion (G. M. Lawrence,privatecommunication,1979)suggests the NO and 02 emissions and the ONMS measurements are
difficult to assessin the absence of an adequate dynamical
model. The descendingpart of the thermosphericcirculation
might be displacedtoward the morning terminator if the 4day retrograderotation seen at the cloud top (--65 km) level
persiststo higher altitudes.However, our current ideascannot
account for morphological differencesbetween the NO and
02 emissions,and we are forced to invoke the day-to-day variability of the morphology to explain the observed discrepv
120
-
ancy.
The variations in morphology and overall brightnessare
themselves of interest. Our simple model shows that the
brightnessof the NO emissiondependsonly on the downward
flux of N atoms;thus the observedpatchinessof the emission
must result from spatial irregularitiesin the high-altitude circulationpattern. This interpretationis consistentwith the variability of the nightsideN and O concentrationsobservedby
the ONMS. By the same token, the variations in overall
i07
i0B
iob
brightness
imply variationsin the supplyof N atomsfrom the
K140 (CM
2 SEC-I
dayside,perhapsindicativeof variationsin the relative imporFig. 4. The dependenceof the altitudesof the maximum densities
of oxygenand nitrogen,the maximum & band emissionrate, and the tance of global and local transport processesin the dayside
maximum 02 productionrate as a function of the eddy diffusion coef- thermosphere. Such variations would certainly affect the
ficient.
global transportof N (and O), sincethe dissociationof N2
•_ •o
7870
STEWARTET AL.: MORPHOLOGYOF VENUS UV NIGHT AIRGLOW
peaksat altitudesnear the locationof the homopauseinferred
by yon Zahn et al. [!979].
Lawrence,G. M., C. A. Barth, and V. Argabright, Excitationof the
Venus night airglow, Science,195, 573, 1977.
Mandelman, M., T. Carrington, and R. A. Young, Predissociation
andits inverse,usingresonance
absorption
NO(C2II) • N + O, J.
Acknowledgments.
This researchwas performedin part under
Chem. Phys., 58, 84, 1973.
NASA contractNAS2-9477and supportedin part by NASA grant
NGL 06-003-052.The National Center for AtmosphericResearch Niemann, H. B., W. T. Kasprzak, A. E. Hedin, D. M. Hunten, and
N. W. Spencer,Mass spectrometricmeasurementsof the neutral
(NCAR) is acknowledged
for providingcomputingservices.
NCAR is
gas compositionof the thermosphereand exosphereof Venus, J.
sponsored
by the National ScienceFoundation.J.-C. G. is supported
Geophys.Res., this issue.
by the BelgianFoundationfor ScientificResearch(FNRS).
The Editor thanks D. E. Anderson and P. D. Feldman for their as-
sistancein evaluatingthispaper.
REFERENCES
Barth, C. A., J. B. Pearce,K. K. Kelly, L. Wallace,and W. G. Fastie,
Ultraviolet emissionsobservednear Venus from Mariner V, Science, 158, !675, 1968.
Baulch,D. L., D. D. Drysdale,D. G. Horne,and A. C. Lloyd, Homogenousgasphasereactionsof the He-N2-O2system,in EvaluatedKineticDatafor High Temperature
Reactions,
vol. 2, pp. 1557, Chemical Rubber Company Press,Cleveland,Ohio, 1973.
Phillips,L. F., and H. I. Schiff,Massspectrometerstudiesof atom reactions,1, Reactionsin the atomic nitrogen-ozonesystem,J. Chem.
Phys., 36, 1509, 1962.
Rishbeth,H., and D. W. Barron,Equilibrium electrondistributionsin
the ionosphericF2-1ayer,J. Atmos. Terr. Phys., 18, 234, 1960.
Rusch, D. W., and T. E. Cravens, A model of the neutral and ion ni-
trogenchemistryin the daytimethermosphereof Venus, Geophys.
Res. Lett., 6, 791, 1979.
Simonaites, R., and J. Heicklen, Kinetics and mechanisms of the reac-
tionof O(3p)withcarbonmonoxide,
J. Chem.Phys.,56,2004-2011,
1972.
Slanger,
T. G., and G. Black,The O2(C3Au
--> a'Ag)bandsin the
Campbell,I. M., and C. N. Gray, Rateconstants
for O(3p)recombinightglowspectrumof Venus, Geophys.Res. Lett., 5, 947, 1978.
nationandassociation
with N(4S),Chem.Phys.Lett.,18,607, 1973. Stewart,A. I. F., Designandoperationof the PioneerVenusOrbiter
Colin, L., Encounter with Venus, Science,203, 743, 1979a.
Ultraviolet Spectrometer,IEEE Trans. Geosci.RemoteSensing,
Colin, L., Encounterwith Venus:An update,Science,
205,44, 1979b.
GE-18, 65, 1980.
Cormes,P., J. F. Noxon,W. A. Traub,and N. P. Carleton,O2('A) Stewart,A. I., and C. A. Barth,Ultravioletnight airglowof Venus,
emission
in the day andnightairglowof Venus,Astrophys.
J. Lett.,
Science, 205, 59, 1979.
in press, 1980.
Dickinson,R. G., and E. C. Ridley,Venusmesosphere
and mesospheretemperaturestructure,2, Day-night variations,Icarus, 30,
163-178, 1977.
Feldman,P. D., H. W. Moos, J. T. Clarke, and A. L. Lane, Identification of the UV nightglowfrom Venus,Nature,279, 221, 1979.
Hampson, R. F. (Ed.), Survey of photochemicaland rate data for
twentyeightreactions
of interestin atmospheric
chemistry,
J. Phys.
Chem. Ref Data, 2, 267-312, 1973.
Kaufman,F., Elementarygasreactions,
Ann.Rev.Phys.Chem.,20,
Stewart,A. I., D. E. Anderson,Jr., L. W. Esposito,and C. A. Barth,
Ultraviolet spectroscopy
of Venus:Initial resultsfrom the Pioneer
Venus orbiter, Science,203, 777, 1979.
von Zahn, V., K. H. Fricke, H.-J. Hoffmann, and K. Pelka, Venus:
Eddy coefficients
in the thermosphere
and the inferredhelium content of the loweratmosphere,Geophys.
Res.Lett., 6, 337-340, 1979.
vonZahn,V., K. H. Fricke,D. M. Hunten,D. Krankowsky,K. Mau-
ersberger,
andA. O. Nier, The upperatmosphere
of Venusduring
morning conditions,J. Geophys.Res.,this issue.
45-90, 1969.
Krasnopolsky,
V. A., A. A. Krysko,V. N. Rogachev,
andV. A. Parshev,Observations
of Venusfrom the interplanetary
spacecraft
Venera9 and Venera 10, Soy.SpaceRes.,14, 789, 1976.
(ReceivedJanuary2, 1980;
revisedMay 28, 1980;
acceptedMay 29, 1980.)