Aug 30, 1990 - ments carried out in situ [McElroy et al., 1976; Nier and. McElroy, 1976, 1977; $eiff and Kirk, 1977; Kerzhanovich,. 1977]. In the same way, ...
JOURNAL
OF GEOPHYSICAL
RESEARCH,
VOL. 95, NO. B9, PAGES 14,795-14,810, AUGUST
30, 1990
A Nonsteady One-Dimensional Theoretical Model of Mars' Neutral Atmospheric Composition Between 30 and 200 km R. RODRIGO, E. GARCfA-/•LVAREZ, M. J. LOPEZ-GONZALEZ, AND J. J. L6PEZ-MORENO Instituto de Astrofisica de Andaluc[a, Granada, Spain There has been a big advance in the knowledge of the composition of the atmosphere of the planet Mars since its exploration by different missionsin the 1970s, and this will be deeply increased in the following years as the upcoming programs to Mars develop. In this context, we have elaborated a
modeloftheMars'neutral atmosphere including thefollowing compounds: O(3p),O(1D),02, 03, H, H 2, OH, H20, HO2, H202, CO, and CO2, between 30 and 200 km of altitude. The model is carried out for middle latitudes in equinox conditions and with moderate solar activity and provides the day-to-night evolution of the atmosphere.The scarcity of observationscorrespondingto the nightside of the planet has made it necessary to calculate the atmospheric temperature profile based on the available
observations
and on theoretical
estimations.
The model includes a detailed
treatment
of both
the photochemicaland the dynamical processes.In this sense, the most recent values of the reaction rates and photodissociationcross sections have been used, and a new height profile of the eddy diffusion coefficient has been computed which is able to explain the vertical distribution of carbon monoxide. The concentration profiles obtained show, in general, a very good agreement with the available experimental measurements.
1.
INTRODUCTION
In the last few years, the exploration of the solar system has provided a more precise knowledge of the structureand compositionof the planetary atmospheres.New experimental data on Mars's atmosphere from ground-basedobservations, remote sensingexperiments from flyby and orbiting spacecraft, and in situ measurementshave become available allowing a better estimation of some atmospheric parameters. Simultaneously, new theoretical models have been developed in an attempt to achieve a reasonableunderstanding of the photochemical and transport processesresponsible for the altitude profiles of the atmosphericspeciesin the middle and upper Martian atmosphere.Although our present understandingof the Martian atmospherecan be characterized by a continuouslyimproved knowledge of the composition of the lower Martian atmosphereat daytime, there are, however, quite a large number of fields which still remain poorly explored and unknown. Thus there is very little information on the mean state of the atmosphereat nighttime, on the energy balance in the upper atmosphere,and on the compositionand dynamical processesin the atmosphere above 40 km.
Following the deep-space missions to Mars, different theoretical models of the structure and compositionof the Martian atmosphere have been developed. Most of them have been made to interpret, in a general way, the measurements carried out in situ [McElroy et al., 1976; Nier and McElroy, 1976, 1977; $eiff and Kirk, 1977; Kerzhanovich, 1977]. In the same way, models have been constructed for the study of some specificproblems and compounds,either odd-oxygen constituents [Kong and McElroy, 1977a; Krasnopolsky and Parshey, 1979; $hirnazaki and $hirnizu, 1979; Traub et al., 1979; $hirnazaki, 1981] or nitrogen compounds [Hunten, 1974; Yung et al., 1977; Krasnopolsky et al., 1979; Copyright 1990 by the American Geophysical Union. Paper number 90JB00158. 0148-0227/90/90JB-00158505.00
Fox and Dalgarno, 1980]. On the other hand, stationary composition models [Liu and Donahue, 1976; Marov et al., 1976; McElroy et al., 1977; Izakov and Krasitskii, 1977] and one nonsteady model [Krasitskii, 1978] have been developed to study the altitude distribution of the different constituents in the Martian atmosphere. Modelling taking into account the dust content and the annual cycle [Leovy et al., 1985] and their influence on the abundanceof the atmospheric constituents [Lindner, 1988] and on thermal regime [Zurek, 1978, 1982] have made it possible to improve the general understandingof the chemical and transport processescausingthe complex interrelations between the different atmospheric constituents.
The aim of this paper is to present a nonsteady onedimensionaltheoretical model of the Mars' atmosphere for middle latitudes in equinox conditions with moderate solar activity in the 30-200 km region by using equations not restricted to minor constituents, which includes a complete photochemical scheme and a reasonable treatment of transport phenomena.The model describedhere is in line with the previous work of Rodrigo et al. [1981], Battaner and Rodrigo [1981], and Rodrigo et al. [1986] for the terrestrial atmosphere. The model incorporates the latest measurements of different atmospheric parameters and includes sophisticatedsolar effect calculations taking advantage of more recent estimates and experimental measurements of solar fluxes and photodissociationcross sections. The model computes the abundances of neutral atmospheric compoundsas a function of the height and solar zenith angle. The
compounds considered areCO2,CO,O (atthestates3pand 1D),02, 03, H202, H2, H20, andthemainspecies derived from the water vapor photolysis: OH, H, and HO2. From a general point of view, the model uses a semiempirical description of the diurnal variations of the atmospheric temperature altitude profile. On the other hand, it is considered vertical transport by molecular and turbulent diffusion which is parametrized, as usual, by an eddy diffusion coefficient previously computed.
14,795
14,796
RODRIGOET AL.' MARS' NEUTRAL ATMOSPHERICCOMPOSITIONBETWEEN 30 AND 200 KM 2.
MODEL
lOO
CALCULATIONS
The mathematical transport equations used in the model are those developedfor O, 02, and N2 by Stubbe [1973] to study the ionic constituents in the terrestrial atmosphere. Later on, Battaner [ 1975] extended the equationsto a neutral atmosphere containing oxygen and hydrogen compounds. Rodrigo et al. [1981] and Battaner and Rodrigo [1981], making use of a generalization of the equations, studied the influence of the horizontal transport in the polar terrestrial thermosphere,and Rodrigo et al. [ 1986]applied them to the study of a carbon-oxygen-hydrogenatmosphere. The continuity and momentum equations for every con-
'6o
0 120
,
I 140
stituent may be expressed as
ot
= Pi - rlili --
180
200
220
240
26O
Temperature (K)
Oili
•
160
(tlilJi)
(1)
OZ l+aior T Oz (•iOtli K(•iOtl i 10T •)
Fig. 1. Temperature profiles of the Martian atmosphere. Solid curve, from Viking 1 entry science data [Seiff and Kirk, 1977]; dashed curve, from Viking 2 entry science [Seiff and Kirk, 1977]; dotted-dashed curve, from Mars 6 observations; dotted curve, from COSPAR [ 1982].
Vi Vi Di
-
Oz
+ -
T Oz
+
(2)
where i is the ith constituent, t is local time, ni is concentration, Pi is photochemicalproduction, l i is specificphotochemicalloss, z is altitude, vi is mean vertical velocity, Vi is friction velocity, D i is moleculardiffusioncoefficient,a i is thermal diffusioncoefficient,T is temperature,H i is individual scale height, K is eddy diffusion coefficient, and H is atmospheric scale height. 2.1.
Atmospheric Temperature
To determine the diurnal variations of the compound concentrations, it is necessary to include the day-to-night variations of the atmospheric temperature. As experimental information on the nighttime thermospheric temperature is not available, we have to compute the temperature variations based on theoretical developments. Based on the available data, Stewart and Hanson [1982] has adopted in its Mars reference atmosphere a mean atmospheric temperature profile which can be taken as representative of mean conditions in the dayside martian atmosphere [Stewart and Hanson, 1982]. The profile was taken from the stationary model of McElroy et al. [ 1977] and is shown in Figure 1, together with those measured by Mars 6 and Viking 1 and 2. Both of Viking's temperature profiles are very similar with small differences associated to the different season, latitude, and local time at which measurements were taken. The thermospheric temperatures are in both cases unusually low, 180 K and 130 K for Viking 1 and 2, respectively. These values are much lower than those obtained from radio occultation and dayglow measurements on Mariner missions [Fjeldbo et al., 1970; Anderson and Hord, 1971; Kliore et al., 1972; Barth et al., 1972]. Viking temperature profiles were measured during a period of exceptionally low solar activity and when Mars was near its aphelion, and this could explain the lower values measured for the temperature at high altitudes. The COSPAR profile showsa subadiabaticlapserate of---2.5 K/km up to about 30 km. At higher altitudes, the atmosphere is essentially isothermal up to 100 km. In this range of altitudes, the atmo-
sphere is practically in radiative equilibrium. At altitudes above 100 km, the increase of the temperature with height, followed by a transition to an isothermal profile, is mainly determined by absorption of ultraviolet solar radiation and by conductive transfer of heat to lower altitudes down to a height where part of the energy escapesto space through infrared radiation losses. In this way, there is no doubt that the thermospheric temperature varies considerably with time and a-day-to-night variation of--•150 K in the mean exospherictemperature has been claimed by many authors [seeBougher et al., 1988, and references therein]. To compute the day-to-night variation of the temperature profile, we have applied the same method developed by Hedin et al. [1983] and Hedin [1983, 1987] for the Venusian and terrestrial atmospheres. Thus, for a determined local time, the temperatureprofile is calculatedfrom some certain height,z0, up to the upperlimit consideredin the model, z = Zmax.The initial altitude, z0, hasto be chosenin sucha way that at lower altitudes the temporal variations of the temperature and its gradient with height are practically zero. In this way, a value for z0 between 60 and 100 km would be adequate, because of experimental data show that an isothermal atmosphereexists at lower altitudes and variations with local time can be considered as negligible. At altitudes lower than z0, a time-independenttemperatureprofile can be consideredand the values recommended by COSPAR [ 1982] were adopted. As we have mentioned above, the temperature profile has a positive gradientwith heightfrom --•100km up to asymptotically reaching the exospheric temperature from an inflection point at altitude Za. In this model, the upperboundaryis placedat Zmax-- 200 km at which it canbe assumedthat the temperature at the top of the atmosphere, T•, has already been practically reached. The exospheric temperaturecan be consideredas a function of local time, latitude, and solar activity. Becausethe diurnal variations of the thermospherictemperature dependbasically on those of the exospheric temperature, we first calculate the time dependenceof T•. This dependencewith local time can be evaluated making use of mean values and an expansion function in Legendre's polynomials. Thus T•(t) can be expressed as
RODRIGO ET AL.' MARS' NEUTRAL ATMOSPHERICCOMPOSITIONBETWEEN 30 AND 200 KM
14,797
Too(t)= Too[P00 + a10P10(cos X) + allPll(COS 0) sin tot] (3)
whereboth the meanglobalcomponentP0oand the funda-
7oo
mental diurnal componentP ll are considered.Toorepresents
themeanexospheric temperature, anda valueof 300K was adoptedhere in accordancewith COSPAR and correspond-
•oo
ingto a moderate solaractivity.X is solarzenithangle: cosX = sin 0 costo(t- 12), to= 2•r/12Martianhours,0 is
colatitude, t is localtime, and Primare the Legendre's •oo associated
functions.
The a10 and a ll coefficientsare determined"a priori" in such a way that the computed temperature profile adjusts to l OO 5 10 15 20 o the mean profiles obtained by Viking 1 and 2 at z = 120 km, Local Time (hours) where seasonalvariations can be consideredas negligible. We have obtained for a10 and all values of 0.332591 and Fig. 2. Variation of the exospheric temperature along the day. Solid curve, this model; dashed curve, Izakov and Morozov [1976]. -2.030765, respectively. Once Too(t)has been obtained, T(z, t) can be expressed [Hedin, 1983] as
For z -> Za
Morozov [1976] which is greater than that normally accepted of e = 0.16 [Stewart et al., 1972;Bougher et al., 1985;Bittner
T(z, t)= Too(t)- [Too(t)- T•(t)]
and Fricke, 1987].
Figure 3 shows the atmospheric temperature profile at noon and midnight. The obtained values at 200 km are very close to the maximum and the minimum values, respectively, that the atmospheric temperature reaches along a For z < Za complete day. At the upper boundary of the model, the obtained variation is of about 155 K which agrees very well T(Z, t)= TO-(T O- Ta) with that predicted by Izakov and Morozov [1976] whose profiles are also plotted in Figure 3. This day-to-night variation is also in agreement with that of--•150 K recommended by COSPAR and that assumed by most authors [see, e.g., Singhall and Whitten, 1988; Bougher and Dickinwhere z• is the height at which molecular diffusionpredomson, 1988]. The altitude range at which the temperature inates, T• is temperaturecorrespondingto altitude z•, Ta is gradient is positive is shorter at night than during the day, temperature at Za, Tois thatat z0, R•, is the planetradius, and a value for Tooof 220 K is rapidly reached from 130 km and T} is the temperaturegradientat z• altitude and considat night because of the absence of solar radiation. The low ered as time-independent. atmospheric density at higher levels of the atmosphere and In this way, the T(z, t) profiles will be, at any time, absorption of ultraviolet solar radiation during daytime leads asymptotic in both extremes, z0 and Zmax,and with an to a vigorous gradient of temperature which is positive inflection point at Za in such a way that the mean profile between 100 and 160 km. At higher altitudes, the lapse rate consideredshouldbe reproduced.The altitude Za = 126 km becomes smaller until the temperature asymptotically accomplishesthis requirement. A value of 100 km has been reaches a value of around 380 K for the exospheric temperadopted for z0, although the calculationswere carried out ature which is greater than that indicated by COSPAR [ 1982] down to 30 km in order to comparethe obtainedprofile with COSPAR profile, with the result that both profiles are
ßexp (oott)-Tl(t) (z-(Rp+z) T• •----T/z•)(R• +z•).) (4a)
(T[Too(t)-Ta(z-Za)(gp+za),)(ab)
ßexpToo(t)Tl(t)Ta-To (Rp +z)
coincident.A value for zl equalto 140km hasbeenfixed for
theturbopause level.
200
DAY '
The diurnal variation of Toois shown in Figure 2, together with that obtained by Izakov and Morozov [1976] for comparison. These authors determined the diurnal variation of Tooby means of a two-dimensional model of the Martian
thermosphere structurein equinoxconditionswith moder-
ately high solar flux under the assumption ofa100% •00 atmospheric composition and a heating efficiency of e = CO2 0.3. The diurnal variation obtained here is similar to that calcu-
lated by Izakov and Morozov [1976]. A maximum value for the exospheric temperature is achieved near noon in both cases. Although the amplitude of the Toovariation agrees fairly well, there is a small differencein the phasethat could be due either to their computation or to the restricted limit number of terms used in (3). The differences in absolute value could be attributed to the different solar activity considered and to the high value for e used by Izakov and
I
200
•
I
300
I
400
500
Temperature (K)
Fig. 3. Temperature profiles of the Martian atmosphereat noon and midnight. Solid curves, this model; dashed curves, lzakov and Morozov [1976].
14,798
RODRIGOET AL.: MARS' NEUTRAL ATMOSPHERICCOMPOSITIONBETWEEN 30 AND 200 KM TABLE
1.
Chemical
Reactions
Reaction
No.
Reference?
Rate Coefficient*
(R1) (g2)
CO + O + CO 2--->CO 2 + CO 2 CO + OH --> CO 2 + H
(R3) (R4) (R5) (R6) (R7) (R8) (R9) (R10) (Rll) (R12) (R13) (R14) (R15) (R16) (R17) (R18) (R19) (R20) (R21) (R22) (R23) (R24) (R25)
O(3p)+ 0 2 + CO2--'>0 3 + CO2 O(3p)+ 0 3--->0 2 + 0 2 O(3p)+ O(3p)+ CO2 --->0 2 + CO2 O(1D)+ CO2 --->O(3p)+ CO2 O(1D)+ H2 -->OH + H O(1D)+ H20--->OH + OH H + 0 2 + CO2 --->HO 2 + CO2 H+03-->OH+02
O(3p)+ OH -->02 + H O(3p)+ HO2 --->OH + 0 2 OH + HO 2 --> H20 + 0 2 H + HO 2--->H 2 + 0 2 H + HO 2---> H20 + O H + HO 2-->OH + OH HO 2 + HO 2 --->H20 2 + 0 2 OH + OH + CO2 --->H20 2 + CO2
O(3p)+ H202 --->H20 + 0 2 H + H + CO2-->H2 + CO2 HO 2 + 03--->OH + 20 2 H 2 + OH---> H + H20 H20 2 + OH--> HO 2 + H20
CO•- + H2 -->CO2H+ + H CO2H+ +e-->CO 2 + H
kl = 2 x 10-37
Slanger et al. [1972]
k2 = 6.0x 10-13(0.25+ A[M]/(1 + A[M])i
A = 1.82 x 10-20
k3 = 6.0 x 10-34(T/300) -2'3 k4 = 8.0 • 10-12exp(-2060/7) ks = 9.4 x 10-34exp(484/7) k6 = 7.4 x 10-11exp(117/7) k7 = 1.0x 10-1ø k8 = 2.2 x 10-1ø k9 = 1.3x 10-31(T/300) -1'6 kl0 = 1.4x 10-lø exp(-470/7) kl• = 2.2 x 10-11exp(117/7) k•2= 3.0 x 10-11exp(200/.7) k13= 1.7x 10-11exp(416/7) kN = 6.66x 10-12 k15= 2.96x 10-12 k16= 6.44x 10-11 k17= 2.3 x 10-13exp(590/7) kl8 = 1.6x 10-3o(T/300) -ø'8 k19= 1.4x 10-12exp(-2000/7) k20= 2.0 x 10-32 k21= 1.4x 10-14exp(-580/7) k22= 6.1 x 10-12exp(-2030/7) k23= 3.1 x 10-12exp(-187/7) k24= 1.4x 10-9 k25= 3.0 x 10-7
Hampson [1980]
Lindner [ 1988]
Lindner [ 1988]
Hampson [1980]
Krasnopolsky [ 1986] Krasnopolsky[ 1986]
*All values are quoted in the molecule cm s system.
?Except as noted, reaction rate constantsare those recommendedby DeMore et al. [1985].
as a mean dayside exospherictemperatureand seemsto be
1979; Shimazaki and Shimizu, 1979; Shimazaki, 1981]. In all
more consistent with solar maximum
of them, CO2 is regeneratedvia reactionCO + OH --> CO2
2.2.
Photochemical
conditions.
Processes
One of the main problems of Mars aeronomy is to explain the fact that carbon dioxide, which accountsfor about 95% of the Martian atmosphere, predominatesfrom the surface until, at least, 200 km, and this notwithstandingthat CO2 is
+ H as the key process of different cycles. Furthermore, a large eddy diffusioncoefficientis required, of the order of 2-4 orders of magnitude larger than that at equivalent density levels in the Earth and Venus atmospheres. The chemistry of the Mars' atmosphere could be repre-
sentedby the chemistry of a CO2 atmospherecontaining somequantityof water vapor. Becauseof recentadvancesin rapidly photodissociatedat all levels and that three-body the determination of some important reaction rates and recombinationof the photodissociationproducts,CO and O, photodissociationcross sections,it seemsto be convenient is a very slow processfor regeneratingthe atmosphericCO2. to revise the photochemicalprocessestaking place in the Furthermore, atomic oxygen easily forms 0 2 in three-body Martian atmosphere.In this sense,the excellent reviews by
reactions, andthisprocess is about105 timesmorerapid Barth [1985], Wayne [1985], and Krasnopolsky [1986] have
than reaction with CO to form CO2. The relative concentrations observed in the lower atmosphere [Owen et al., 1977; Owen, 1982] are as follows: CO2, 95.5%; CO, 0.07%; 02, 0.13% and according to the Mariner 6 and 7 measurements, the O abundance
to be referred to. Tables
1 and 2 list the chemical reactions
and photodissociation processesconsideredin this model.
is of the order of 1-3% at the level 1.2 x TABLE
10-9 bar(or equivalently at --•130km).However,bearingin mind the above processesand if no other processof CO2 formation were considered, a mixing ratio of about 0.1% of
No.
2. Photodissociation Reaction
Processes
Wavelength, nm
CO and02 wouldbe producedin only 4 years[vonZahn and
CO2 + hv-->CO + O(1D)
A
CO + O
A_> 167
03 + hv-->0 2 + O(1D)
Lyman-aefoda) = 0.7
and 02 abundancesnear valuesof 0.1%. Catalysisof CO2 and 02 (and even O) by odd-hydrogencompoundswas proposed in early models [McElroy and Donahue, 1972; Parkinson and Hunten, 1972] as responsiblefor the CO + O association.These early calculationshave been revised and improved by different authors until photochemical models capable of reasonably explaining the experimental measurements are obtained [Hunten, 1974; Kong and McElroy, 1977a, b; McElroy et al., 1977; Krasnopolskyand Parshev,
167
A < 410efoda)= 0.7 0 3 + hv--> 0 2 + O
410 --< A < 730
02 + hv--->O + O(ID)
Lyman-a A < 175.4 175.4 < A -< 243.9 A < 350 175 -< A < 275
O2 + hv--> O + O H202 + hv--> OH + OH HO 2 + hv-->OH + O H20 + hv--> H + OH
Lyman-a ef = 0.75
H20 + hv--> H 2 + O
Lyman-a ef = 0.25
A < 10
198
RODRIGO ET AL.: MARS' NEUTRAL ATMOSPHERIC COMPOSITION BETWEEN 30 AND 200 KM
The catalytic cycles in which odd-nitrogencompounds(N, NO, and NO2) take part are not included, as their influence on the chemistry of other compoundshas been demonstrated as unimportant [Yung et al., 1977; Krasnopolsky and Parshev, 1979]. Table 1 shows the rate constants for reactions used in this model. Most of the rate coefficients are taken from the excellent review of DeMore et al. [1985]. However,
'",
14,799
",,,•,TO TA L
•-•
',.,.- "., N, x (BFonch
two major problems remain when the photochemistryof the atmosphere of Mars is studied. There is still a great uncertainty on the value for some reaction rates at the very low temperatures of the Mars' atmosphere, and this is particu-
larly relevant for important reactions involving HO2 whose reaction rates are difficult to measure in the laboratory. The second problem is related to the three-body reactions as their effectivenessdependsupon which molecule is used as third body. In general, there is no experimental determination for theserate coefficientswhen CO2 or CO is considered as the third body. However, it can be considered that the ratio of the three-body effectiveness of one molecule to another is approximately the same for all reactions, and therefore when experimental rate constantsare not available for reactionswith CO2 as the third body, it is then possible to use a conversion factor for the correspondingcoefficient measured in the laboratory. In this case, we have used the values considered by Lindner [1988], who took into account the conversion factors and the efficiencies proposed by different authors. Rate coefficients for reactions (R5) and (R20) are those from Harnpson [1980] and for reaction (R1) from Slanger et al. [1972]. The main photochemical process is CO2 photodissociation. The recombinationof the CO2 photodissociationproducts, CO and O, takes place at lower altitudes where both constituents have been transported. This recombination is rapid enough because odd-hydrogencompoundsact as catalysts. One of the main cyclesis throughreactionswith HO 2 and OH (reactions (R12) and (R2) in Table 1) which, together with reactionconvertingH in HO 2 (R9), givesa net resultfor the cycle: CO + O --• CO2. Another important cycle is that proposed by Parkinson and Hunten [1972], which includes reactions (R9), (R17), and (R2) and H202 photodissociation.An indirect photodissociationof 02 occurs producinga net result for the cycle: GO d- 02 d- hv --• GO2 d- O. H202 acts as an odd-hydrogen reservoir, and the effectivenessof the cycle is small because H202 photodissociationis a slow process.Other indirect 02 photodissociationprocessis that via HO2 photodissociation although this presents similar difficulties. The cycle involving reactions (R3), (R10), and (R2) can also be considered as a nonnegligiblemechanismto produceCO2. The above cycles, in which odd-hydrogen compounds play a primary role, also catalyze the recombination of odd-oxygen compounds. The odd-hydrogencompoundsare mainly produced by H20 photodissociationat wavelengths
shorter than197nmandbyreaction ofH20 withO(1D)(R8).
30
10-4
altitudes above 80 km, molecular hydrogen is dissociatedvia reactions (R24) and (R25). Although the main source of odd-hydrogenin the lower atmosphereis H20 photodisso-
1
10
100 1000 '"• 4
Rates of various loss processesfor H•.
hydrogen formation at altitudes above --•100 km and in the case of a very dry atmosphere. The main sinks of odd-hydrogen (see Figure 4) are reactions of HO2 with other HO2 molecule (R17) and with OH (R13) at altitudes near the lower boundary of the model. At altitudes higher than -35 km, reaction of atomic hydrogen with HO2 ((R14) and (R15)) beginsto be competitive, and it becomespredominant at altitudes above -50 km. Reactions (R18) and (R20) can be considered as negligible sinks of odd-hydrogen between 30 and 60 km. Reactions (R16), (R22), and (R23) produce an interchangebetween the different odd- and even-hydrogen species, and they do not play a major role. Figure 5 shows the OH production rates for the different processes considered in the model. The main production mechanism of OH up to 60 km is reaction (R!2) of atomic
oxygenwith HO2. At higher altitudes,H20 2 photodissociation is the predominant mechanismup to --•85 km. At this
altitude,dissociation of H2 by reactionwith O(1D) (R7) becomes important. It is worthwhile noting that reaction of atomic hydrogen with ozone (R10), the well-known Bates
and Nicolet [1950] mechanismcapable of producinghydroxyl molecule in vibrational levels v = 6, 7, 8, and 9, can be consideredas a secondaryproduction mechanism of OH in the altitude range considered here.
Theatomicoxygenelectronically excitedat level1Dplays an important role in the Martian atmosphere. Although its
100
. 80
-
'-•
, i
•,- O( D)+H 2 ".,..,.
_
• 60
H+HO '''-M+MU2•'• ' H2U+h•• ''•
3 O( P)+HO
40
10-4
10-3 0.01
0.1
1
10
100 1000
104
OHProduction Rotes(cm-3 s-1)
ciation,dissociation of H2 via reactionwithO(•D) or with CO•- shouldbe consideredas responsiblefor the odd-
0.1
H• LossRotes (cm-s s-•) Fig. 4.
Water vapor is recovered through reactions (R13) and (R15). These sources of water vapor are in turn associatedto the
lossof molecular hydrogen by reactionwithO(1D)(R7).At
10-3 0.01
Fig. 5.
Rates of production mechanisms of OH.
105
lO6
14,800
RODRIGO ET AL.' MARS' NEUTRAL ATMOSPHERIC COMPOSITION BETWEEN 30 AND 200 KM
8O
ultraviolet region to 10 nm in the visible region. The solar fluxes at the top of the atmosphere were those proposedby Mount and Rottman [1983] suitable for moderate solar activity and conveniently modified to take into account the distance between Sun and Mars. Absorption cross sections
E 60
for 02 and 03 were taken from Ackerman [1971].The values of Thompson et al. [1963] (as used by Kasting et al. [1979])
._
were adopted for the H20 absorption cross sections,and those for HO2 are from Paukert and Johnston [1972]. For H202 cross sectionswe have adopted the values proposed by Schfirgers and Welge [1968] up to 175 nm and from
4O
, , , ..... I
0
1O0
]". .......
I ' 7'.',',. .... I
1000
, , , ,,,,il
104
Molina et al. [1977] for wavelengths greater than 197.3 nm.
,
10s
OxLossRotes(cm-3 s-t) Fig. 6.
Rates of various loss processesfor Ox at noon.
We have usedfor CO2 absorptioncrosssectionsthe values proposedby Inn et al. [1953] up to 175 nm and by Shemansky [1972] up to 203 nm, assuming that at greater wavelengthsthe observedcross sectionsare only due to Rayleigh
410nmisthemainproduction mechanism of O(•D),whereas
scattering in accordance with DeMore and Patapoff[1972]. Calculationfor 02 photodissociationin the fine structure of the Schumann-Runge bands were made following the detailed analysis of Nicolet and Peetermans [1980] and Nicolet [1984]. Average absorption cross sectionsfor H20,
collisionaldeactivationwith CO2 is the main lossprocess.A
H202, and CO2 in this spectralregionare thoseproposedby
abundanceis very low, it is capableof dissociatingH20 and H2 molecules.Ozone photolysisat wavelengthsshorterthan
nonnegligible contribution to theO(•D) concentration is due Kockarts [1976], while for 03 the values of Ackerman [1971] to CO2 photodissociationat A < 170 nm and to the photodissociationof 02 at wavelengthsbetween 134.2 and 175.4 rim.
Molecular oxygen is mainly formed by reaction of atomic oxygen with odd-hydrogen compounds via reactions (Rll) and (R12) and by ozone photolysis. It can be mentioned, as secondarysourcesof 02, atomic oxygenrecombinationand the cycle involving reactions (R3), (R10), and (R11). In this cycle, odd-hydrogen compounds act as catalysts. On the other hand, 02 is photodissociatedat wavelengthsshorter than 244 nm and reacts with atomic hydrogen to produce
HO2 and with atomic oxygento produce03. Atomic oxygen is mainly producedby CO2 and 02 photodissociation.
It is well known that a redistribution
of the
concentration of the active odd-oxygen species actually existsbetweenO and 03. As illustratedin Figure6, the main loss mechanismof odd-oxygen, at altitudes below 40 km, is the reaction of atomic oxygen with HO2 (R12). This process regeneratespart of the molecular oxygen and produces OH which reacts with CO producingCO2 or with O to produce 02 (Rll). At altitudes above 40 km, atomic oxygen easily recombinatesto form 02 in three-body reactions(R5). This can be consideredas the main loss processof odd-oxygenat these altitudes. At nighttime, a maximum of odd-oxygenloss occurs at about 50 km because of, basically, the action of reactions, (R5), (R11), and (R10). These reactions, together with reaction (R12) are the main odd-oxygen loss mechanisms at altitudes
were adopted. CO2 absorptioncrosssectionsstronglydependon temperature. Their values at room temperature are double those at 200 K [DeMore and Patapoff, 1972]. This effect has an importantinfluencenot only on the computationof the CO2 photodissociationcoefficient but also in the calculation of the photodissociationcoefficientsof absorbingspeciesin the same spectral range [Parisot and Zucconi, 1984]. We have
taken into account the temperaturedependenceof CO2 absorptioncrosssectionsmaking use of the relation given by Yung and DeMore [1982]. In order to examine the importance of the incident solar radiation in the computation of the different photodissociation coefficients,the total spectral region was regroupedin various ranges dependingon the absorbing species. In this way, the noon photodissociationcoefficientsof 02 in four spectral regions are shown in Figure 7. The Herzberg continuum and the Schumann-Rungebands are mainly responsiblefor the 02 photodissociationup to about 40 km. The contribution of the Shumann-Rungebands is still impor-
2OO
150
below 40 km.
Ozone is mainly producedby recombinationof atomic and molecular oxygen in three-body reactions, and its loss is dominated by photolysisprocessesmore than by reactions
1
3
2
E1 oo
with atomic hydrogenand with HO2. Photodissociation coefficients (see Table 2) were calcu-
lated, following the sameformalism of Rodrigo et al. [1986], for CO2, 03, 02, H202, HO2, and H20 in three spectral regions between 116.3 and 730 nm (116.3-175, ShumannRunge bands and 197.3-730 nm) and consideringLyman a line separately in all of the photochemical processes involved. The ranges 116.3-175 nm and 197.3-730 nm were divided into spectral intervals ranging from 0.9 nm in the far
' •0 10-9 10-• ....... 10'-• 1'....... 10-
10-8
10-7
10-6
1
02photodissociotion coefficients (s-1) Fig. 7. Noon photodissociationcoefficientsof 02 within four spectralregions: 1, Lyman a; 2, 116.3-175.4 nm; 3,175.4-198 nm' 4, 198-243.9 nm; T, total.
RODRIGO ET AL.' MARS' NEUTRAL ATMOSPHERIC COMPOSITION BETWEEN 30 AND 200 KM
tant up to a height of--•60 km. Above this altitude, the contribution from wavelengths shorter than 175.4 predomi-
nates. The effect of Lyman a radiation on the 02 photodissociation can be ignored in the range of altitudes considered here. These results agree well, qualitatively speaking, with those obtained in the terrestrial atmospherewith semiempirical and theoretical determinations[see, e.g., Nicolet, 1981a; Rodrigo eta!., 1986] with the exception of the role played by Lyman a radiation which has a nonnegligible contribution to the total photodissociation coefficient between 70 and 80 km in the Earth's atmosphere. Water vapor and carbon dioxide practically absorb the solar radiation in the same spectral region and with a similar behavior (see Figure 8). This is explained by the influence of
the absorptioncross sectionsof CO2 on the photodissociation coefficientof H20. For both compounds,photodissociation is mainly produced, at altitudes below 50 km, by absorption in Schumann-Runge bands region. At higher altitudes, absorption in Lyman a and Schumann-Runge continuum predominates, becoming photodissociationcoefficient for CO2 in the Schumann-Rungecontinuum greater by a factor of 5 at altitudes above --•100 km. Water vapor photodissociationby absorption of Lyman a becomes important at a height of--•70 km, predominating at altitudes above 85 km, as in the terrestrial atmosphere at equivalent density levels [Nicolet, 1981b, 1985; Rodrigo et al., 1986]. The radiation between 198 and 310 nm, where the effective cross section reaches its maximum, plays the main role in
the O3 photodissociation. A constant valueof 3.6 x 10-3 s-• in all thealtituderangeconsidered here is obtained for the photodissociationcoefficient of process3 in Table 2. The
production of O(•D) fromphotolysis of O3by absorption of Lyman a and in the Schumann-Runge continuum and bands system is negligible in comparison with that of the Hartley bands. The values considered for the efficiency to produce
O(•D), 0.7 for A < 198and 1.0for 198-< A -< 310nm, are those proposed by L6pez-Moreno et al. [1988] and L6pezGonzdlez et al. [ 1989]. Chappuis bands (process4 in Table 2)
play a secondary role in the 03 photodissociation.The photodissociationcoefficient of this process reaches a constant value of 1.5 x 10-4 s-1. The radiation between 198 and 310 nm plays the main role in the H20 2 photodissociation,and this role is played by the
14,801
2OO
150
1oo 5o
lOO
1000
10'•
105
106
107
108
109
101ø
Diffusion Coeff,c,ents (cm2 S-1) Fig. 9.
Eddy and molecular diffusion coefficients profiles at noon, as defined
in the text.
radiation at wavelengths shorter than 243.9 nm in the case of HO 2. Contribution due to absorption of Lyman a can be considered as negligible for both compounds. The low concentration of these constituents, together with a slow
photodissociation coefficients, 2.6 x 10-4 s-1 and 4.7 x 10-5 s-• for HO2 andH202, respectively, yielda small odd-hydrogen compound concentrations taking part in the catalytic cycles for the O and CO recombination and for the
indirect photodissociationof 02 already discussed. 2.3.
Diffusion Processes
Vertical transport affecting the atmospheric composition is mainly produced by two different processes:turbulent and molecular diffusion, the former being more effective at lower levels. In equation (2), which is not restricted to minor constituents, the friction velocities can be calculated according to the Stubbe's [1973] formulation. The method of calculation is applicable to any number of gaseous species and corresponds to the first order approximation of the diffusion theory [see, e.g., Chapman and Cowling, 1970]. The molecular diffusion coefficientsD i were calculated following the same formulation. The thermal diffusion factor was assumedto be zero except for atomic hydrogen. The values obtained for D i at noon have been plotted in Figure 9. The diffusion coefficient correspondingto atomic hydrogen is the largest, as it corresponds to the lightest
mass,andreachesa valueof 3.7 x 109cm2 s-1 at 170km. Molecular hydrogen also has a higher diffusion coefficient at 150
3
1
2
170km (2.7 x 109cm2 s-1) thanthatcorresponding to the othercompounds, whichtakevaluesof about9 x 108cm2
T
•
.
• 00 ...../....•.
10-12
10-11
10-1ø
10-9
10-8
.-,.z..
10-7
10-6
0-5
Photodissociotion Coefficients (s-1) Fig. 8. Noon photodissociationcoefficients within three spectral regions: 1, Lyman a; 2, 116.3-175.4 nm' 3, 175.4-198 nm; T, total. (Left)CO2. (Right) H20.
s-• at the same altitude. The different processes able to produce and propagate turbulence have been studied by several authors [see, e.g., Ebel, 1980, and referencestherein] usingthe concept of eddy diffusion. When one-dimensionalmodels of the atmosphere are developed, the turbulent diffusion is parameterized by means of an eddy diffusion coefficient K [Lettau, 1951; Colegrove et al., 1965, 1966]. Although in the last few years, considerable
efforts
have been made to determine
the tem-
poral, seasonal, and height variation of the eddy diffusion coefficient, it still remains the least known factor when it comes to constructing models of the atmosphere. Nevertheless, a great number of studies, mainly on the role of
14,802
RODRIGO ET AL.' MARS'NEUTRALATMOSPHERIC COMPOSITION BETWEEN30 AND200 KM
20O
150
turbopause level, between 130 and 140 km, is also in agreement with other determinations. A more detailed description of the method and its validity is given by Rodrigo et al. [1990]. 2.4.
Numerical Solution and Boundary Conditions
_•oo
Introducing equations (2) into equations (1) results in a system of parabolic second-order differential equations which is solved by means of a modified Gaussian algorithm with a high degreeof accuracy and stability. The time step At 5O for numerical integration must be smaller than the characteristic time of any atmosphericprocess. At the region below 106 50 km, the photochemical characteristic times of some Eddydiffusion Coefficient (cm2 s-1) compounds are very short, and thus the application of a Fig. 10. Different eddy diffusion coefficientprofiles: 1, McElroy systematicand generalnumericalprocedure would require a and Donahue [1972]; 2, Parkinson and Hunten [1972]; 3, Zurek computation time of the order of days. This problem can be [1976]; 4, Krasnopolskyand Parshey [1979]; 5, $himazaki [1981]; 6, solved, however, by combiningthe detailed integrationof (1) $himazaki and Shimizu [1979]; 7, Nier et al. [1976]; 8, Nier and for the long-lived compoundswith photochemical equilibMcElroy [1976]; 9, Krasitskii [1978]; and 10, this model. rium conditionsfor those compoundswith short lifetimes. In this sense, a good simplificationconsistsin the introduction, turbulence in the terrestrial atmosphere, have contributed to as usual, of two dummy compounds that respectively inour knowledge on turbulence and its influence on minor clude the odd-oxygen and odd-hydrogen specieswhich are compounds. Kong and McElroy [1977a, b], Izakov [1978], not stable. Atomic oxygen and ozone were included in the yon Zahn et al. [1980], and yon Zahn and Hunten [1982] odd-oxygen function and H, OH, and HO2 in the oddhave presented a critical analysis of the eddy diffusion hydrogen function. The lower boundary of the model at 30 km allows us to coefficient profiles used in different Martian atmosphere models. In most of these, an eddy diffusion coefficient was neglect the possibleinteraction between the surface and the chosen to adjust the abundance of certain compounds [see, atmosphere [Huguenin, 1982; Lindner and Jakosky, 1985] e.g., Nier and McElroy, 1977]. and permits us to considerthat 02, CO2, H20, H202, and As Krasnopolskyand Parshey [1979] claimed, the study of H2 are fully mixed in ratios accordingto thosegivenby the the distribution of the atmospheric minor compoundsis the concentration profiles initially considered. These mixing most reliable procedure for the determination of K. To ratios areCO2,96.5%;02, 1.65x 10-3;H20, 1.1x 10-8; derive a time-independentprofile for K, we have used the H202, 2.25x 10-8;H2, 7.5 x 10-6. For odd-oxygen and method describedby Battarter [ 1975]and appliedby Rodrigo odd-hydrogen functions mixing at the lower boundary, we et al. [1990] to the atmosphere of Mars. Basically, the adopt the formalism used by Rodrigo et al. [1981]. The upper boundary, placed high in this model at 200 km, procedure consists of the integration of the coupled continuity and motion equationsfor one determined compound, and the use of equationsnot restricted to minor compounds atomic oxygen or carbon monoxide in this case, to determine allow us to assumediffusive equilibrium for any constituent which value of K explainsthe experimentalO or CO profiles. other than atomic hydrogen. An escape velocity for H is We have chosen O and CO because their abundances are assumed, in accordance with Mariner measurements [Barth found to be very much affected by turbulent transport. The et al., 1972; Anderson, 1974; Lewis and Prinn, 1984]. K profiles obtained for both O and CO profiles are very Nonthermal escape of atomic oxygen was not directly included in the model. However, no appreciable variation is similar [see Rodrigo et al., 1990]. In the development of the atmosphere theoretical non- produced in the final results when this flux is imposed as steady model we have adopted the K profile obtained by upper boundary condition. The integration was made by first analyzing the behavior using CO concentrations because experimental results on CO abundances are available and because CO is a constituof every constituent in all the altitude range. Between 30 and lifetimeof O(3p) is very short,in the ent chemically less active than the atomic oxygen. Below 90 40 km, the chemical canbe determined making km, where the method is not applicable, a constant value 102-103s range.O concentration
equalto 2.1 x 107cm2s- • waschosen for K in accordance use of the odd oxygen function once 03 concentrationis with different theoretical and semiexperimentalresults indi-
catingthatvaluesfor K around107cm2s- • arerequired to explain the aeronomy of the lower Martian atmosphere. In Figure 10, the K profile proposed is plotted, together with others reported in the literature for comparison.Our profile showsa typical shapewith different turbulent layers, and its absolutevalues are, in general, in good agreementwith those obtained by other authors. Up to -130-140 km, eddy diffusion dominates transport processes, as can be seen in Figure 9. At about 160 km the molecular diffusion is the fastest process, ensuring diffusive equilibrium at 200 km, where the upper boundary of the model was placed. The
known. The values of the odd-oxygen function can be evaluated by means of the general algorithm. Because the photochemicalcharacteristictime of 03 in this region is shorter than its diffusion characteristic time, it can be
assumed that photochemical equilibrium accomplishesfor
03 and thus 03 concentrationcan be calculatedfrom its continuity equation neglecting the transport terms. In the
sameway, O(lD) is not affectedby transport,and its concentration is computed by means of its photochemical equilibrium equation. A similar situation is found for the hydrogencompounds.H2, H20, and H202 have chemical lifetimes large enough to compute their concentration by
RODRIGOET AL.'.MARS'NEUTRALATMOSPHERIC COMPOSITION BETWEEN30 AND 200 KM
means of the general algorithm. Atomic hydrogen, because of its short lifetime (between 50 and 500 s), has to be computed making use of the odd-hydrogen function and the photochemicalequilibrium condition for HO2 and OH. 02, CO, and CO2 are chemically stable compounds,and their concentration can be calculated by means of the general algorithm. Above 40 km, the chemical lifetime of the atomic oxygen is of the order of 1 hour, and the general algorithm can be then applied to calculate its concentration,whereasthe same scheme
of resolution
is maintained
for
the
rest
of the
200
150
-
:1 O0 50
compounds. For the same reasons, atomic hydrogen concentration can be computed from the general algorithm at altitudes
14,803
0 1001000104 105 106 107 108 109101ø1011101210151014101510161017
above 50 km.
Concentrotion (cm-3)
At nighttime, the situation appears completely different. Fig. 11. Concentration profiles (30-200 km) for the major and The photochemistryof the atmospherechangessubstantially after the sunset mainly because of the absence of photodis- odd-oxygen compounds. Solid symbols, Viking 1; open circles, Krasnopolsky [1975]; curves, this model; solid curves, noon; dashed sociation processes. This is associated with the absence of curves, midnight.
O(•D) and with a rapid decreaseof the O(3p) and H
concentration at lower altitudes as there is no production of these species.Thus 03 has no important chemical loss and will be affected by transport. In this way, atomic oxygen concentration was calculated assuming chemical evolution.
mixing ratios practically constants with height and with a
valueof0.96and1.6x 10-3 , respectively. Theexperimental
OnceO(3p)andodd-oxygen functionareknown,03 con- profiles obtained by Viking 1 sounder by means of mass centration can easily be calculated. At altitudes greater than
40 km, O and 03 concentrationwere calculateddirectly by usingthe general algorithm applied to their coupledcontinuity and momentum equations. The altitude and time variation of the H compounds were obtained by considering the odd-hydrogen function and chemical evolution for atomic hydrogen at altitudes below 50 km. The general algorithm was again used at higher altitudes. It is also necessaryto establishtemporal conditions.From the initial profiles, the calculationof the final time-dependent model was performed until a stablediurnal solutionresulted. The final convergence was reached after 8 days of integration with a height step of Az = 1 km and a time step of At = 15 min in most of the cases. The choice of the initial profiles is somewhat arbitrary. In general, previously computed profiles by means of a stationary model are chosen. As the final resultsof the model are obtained in an iterative process, the choice of the initial profiles has no practical influence on the final profiles. We have adopted as initial profiles in this
spectrometry between 115 and 200 km are also shown in Figure 11. These measurements were made with an unusually low exospheric temperature, To• = 180 K. Kong and McElroy [1977a] have studied the influence of To•upon the concentration of the atmospheric constituents. At noon, the value considered in this model for To•is 377 K, almost 200 K greater than that measured by Viking 1, and this, together with a lower turbulence and a higher solar radiation, could
explain the discrepancybetween the CO2 concentrationat higher levels as measured by Viking 1 and that calculated here.
An altitude profile of CO2 concentration was deduced by
Krasnopolsky [1975]from dayglowmeasurements of CO•bands made by the ultraviolet spectrometerson board Mariner 6 and 7 near solar maximum conditions, and this is also plotted in Figure 11. On board the same spacecraft,Kliore et al. [1971] obtained CO2 densitiesfrom the electron density deduced from radio occultation measurements. The CO2 concentrations obtained by means of both techniques agree
modelthoserecommended by COSPAR.The CO•- concen- verywell, with valuesof 6.8 x 10•øand3.1 x 108cm-3 at tration profile, which is not computed in the model, was considered stationary during daytime and was taken from Viking 1 measurements [Hanson et al., 1977]. At sunrise and at sunset, the time step for numerical integration is too long to be able to withstand the rapid changesin the concentration of certain compoundscaused by a sudden change in physical conditions. Thus, at these times, if an unstable situation develops, the time step is automatically divided, as many times as necessary, by a factor of 10 until stability is once again achieved.
120 and 200 km, respectively. These values are very similar
to thoseobtainedat noonin thismodel,7.2 x 101øand3 x 108cm-3 atthesamealtitudes. Themodelpredicts a greater concentration at noon than at midnight at 200 km by a factor of 1.4, and this is in agreement with a lower exospheric temperature during nighttime. The absence of CO2 photodissociationduring night is compensatedby an inversion in the vertical flux, being upward during daytime. Barker [1972] detected first 02 in the Martian atmosphere from observation of very weak absorption lines in the 762 nm
band,andan 02 mixingratioof 1.3 x 10-3 [Carletonand 3.
RESULTS
AND DISCUSSION
Figure 11 shows the concentration profiles at noon and
midnight for the major compounds,CO2, CO, and 02, at heightsof 30-200 km. Above -• 120 km, the profiles are very regular showing characteristics of a diffusive equilibrium situation. In the lower atmosphere,the CO2 and 02 distribution is mainly determined by mixing up to -100 km, with
Traub, 1972] was given for the lower atmosphere. Recently, Trauger and Lunine [1983] have deduced, with the same
technique, a valueof 1.1x 10-3 forthe02 mixingratio.The modelgivesfor thisratioa valueof 1.6 x 10-3 constant with height up to -• 120 km both at noon and midnight, and this confirmsthat the turbulent processesare mainly responsible for the 02 distribution in the lower and middle atmosphere. Mass spectrometric measurementson board Viking landers
14,804
RODRIGOET AL.' MARS' NEUTRAL ATMOSPHERICCOMPOSITIONBETWEEN 30 AND 200 ICM 200
[Owen et al., 1977]providedan 02 mixingratio in the range
(1-4) x 10-3. Ourresults givea valueat noonfor the02 mixingratioof 2 x 10-3 at 140km,andthisis in agreement with that evaluated from Viking landersmeasurements[Nier
•H2
150
et al., 1976].
The importantloss of 02 by photodissociation duringday is balancedby upward vertical transportwhich requiresa
fluxfor02 of 5 x 109cm-2 s-1 at 80kmin orderto maintain the 02 concentrationaccordingto a diffusive equilibrium situation. At night, the chemical processesat lower levels are the only responsiblemechanisms for the 02 loss,and in this case, downward transport by molecular diffusion oc-
curs.It is worthwhilenotingthat our modelpredictsthat 02
1oo
• •- H202 0 5o
lO
1O0 1000
104
105
106
107
108
109
101ø 101•
and CO concentrationsbecome equal at about 90 km, and Concentration (cm-3) this agreeswith Kong and McElroy's [1977a]computations and with estimations from Viking 1 and 2 measurements Fig. 12. Concentration profiles (30-200 km) for the hydrogen compounds.Solid curves, noon; dashedcurves, midnight. [Nier and McElroy, 1977]. The CO volume mixing ratio obtained in this model in the
lower atmosphereis equal to 7 x 10-4 which is more consistent with near infraredobservations (8 _+3) x 10-4
Despite the great influencethat H compoundsexert in the Martian photochemistry,the abundanceof thesecompounds wave spectraanalysis(3.2 _+1.1) x 10-3 [Goodand Schlo- hasnot been experimentallymeasuredwith the exceptionof erb, 1981].At 100km, the modelgivesfor CO mixingratio a the total water vapour content and some estimations of the valueof 2.1 x 10-3 whichtotallyagrees withthatof (2 _+1) atomichydrogenflux in the higherthermasphere.Neverthex 10-3 obtained by Clancyet al. [1983].Ourresultof 4.4 x less, the results obtained can be globally analyzed from a 10-3 for the CO mixingratioat 125km is withinthelimits qualitativepoint of view in order to evaluatethe validity of deducedby Nier and McElroy [1977] from spectroscopic the model. In general, the altitude distribution of the H measurementsby Viking 1 and 2. The value.for the volume compoundsmainly dependson the photochemicalprocesses mixing ratio of CO at the ionosphericmaximum coincides and on the turbulent mixing processes. Therefore the obwith that for atomic oxygen volume mixing ratio estimated served differences between our results and other theoretical by Strickland et al. [1973], and it agreesvery well with the modelscan be attributed to differencesin the eddy diffusion predictions of the model which gives values for the O and coefficient profile and chemical reaction rates as well as to CO volumemixingratio of 5 x 10-3 and 4.8 x 10-3 , the different treatment given to the photodissociationpro[Cannes et al., 1969; Kaplan et al., 1969] than with micro-
respectively.
cesses.
The presenceof a highturbulencein the lower atmosphere Molecular hydrogen is the dominant constituent among of Mars, togetherwith the recombinationprocessesof CO the hydrogen compounds. It is formed in the lower atmoand O to regenerateCO2, in the presenceof odd-hydrogen sphereas a result of the interactionbetweendifferentH20 compoundsacting as catalysts,lead to CO volume mixing and CO2 photodissociationproducts,and it is transportedto ratios smaller than 1% in accordance with the observations. higheraltitudeswhere a part is dissociatedby reactionswith At altitudes above 100 km, however, higher values of the ionic constituents (mainly CO•) and by reactionswith eddy diffusion coefficientproduce lower abundancesof CO O(1D).The balance of theseprocesses resultsin a volume in this region, down to valuesincompatiblewith the exper- mixingratioof---10-5 at 120km in agreement withestimaimental results. We found that a downward vertical flux for tions by Anderson [1974]. The model gives an upward
CO, of the orderof 109cm-2 s-l at 160km, is neededto balancethe net chemicalloss in the lower atmosphere,and this is also in agreementwith other theoretical results [see, e.g., Kong and McElroy, 1977a].
verticalfluxfor molecular hydrogen of theorderof 108cm-2
s-l at 70-80kmanda H2 mixingratioof 7.5 x 10-6. H2 concentration is uniformly distributed with a constant mixing ratio up to about 100 km. At higher altitudes, the volume
mixingratioincreases untilreachinga valueof 40 x 10-6 at 3.1.
Hydrogen Compounds
The altitude concentration profiles obtained for the H compoundsbetween 30 and 200 km at noon and midnightare shown in Figure 12. The profiles show a noticeable dayto-nightvariationsin the casesof activecomponentsOH, H, and HO2. Atomic hydrogen and hydroxyl molecule almost disappearbelow 40 km at nighttime. H202 concentration remains practically constantduring a day, and only a small day-to-nightvariation is apparentin the lower atmosphere. For the rest of long-lived hydrogen compounds,H2 and H20, both duringthe day and at night, a uniform distribution has been obtained with an approximately constant scale height in the homosphereand accordingto their individual scale heights at higher altitudes.
200 km, whereas the upward flux decreases until values
closeto 107cm-2 s-l . In thealtituderangeconsidered here, an upwardvertical flux for H 2 exists. This flux is coupledto a downward vertical flux for atomic hydrogen up to ---130 km, this being greater in absolute value. Above this altitude, the flux for H becomesin the upward directionand thus part of the atomic hydrogenescapesto space.These resultsare in agreement with those obtained by Kong and McElroy [1977a], who determineda value for the numberdensity and
thefluxfor H2 of 7.5 x 104cm-3 and1.2 x 107cm-2 s-1, respectively at 200 km. Furthermore, our results are in accordance with those deduced to adjust the Lyman a
emissiondata obtained by Mariner 6 and 7 by Anderson [1974],givinga mixingratio of H 2 at altitudeshigherthan 80
kmof---20x 10-6 anda fluxfor H2 of 3 x 107cm-2 s-1,
RODRIGO ET AL.' M^RS' NEUTRAL ATMOSPHERIC COMPOSITION BETWEEN 30 AND 200 KM
108
106 f
30
I
I
I
I
I
I
--'
1,5
18
14,805
I
I:
3O
107
105• ' ......'bS'
.....
104
100 I 1011 0
I 5
1 ooo
6
9
i2
i5
18
21
24
5
6
Local Time (hours)
9
12
21
24
Local Time (hours)
lOlo
lO•o
lO9
?E109
•.108
.............iO0
?
1=107
3O
•-,-.. 108
.;,lO 6 o
150
o
"
":•
50 30
/
........•
; ........... .............. ............. ..-"i.........
'•105
200
104
lO7
8O
1 ooo
o
3
6
9
12
15
18
21
24
LocalTime(hours)
-=1
I
0
3
6
I
I
9
12
15
18
21
24
LocalTime(hours)
Fig. 13. Diurnal variation of different compoundsat selected altitudes.
althoughoverestimated CO• concentrations andeddydif- provides the escape of hydrogen and oxygen atoms. On the fusion coefficient were used in this computation. Lindner [1988] obtained, in his analysisof ozone variation, an upper
other hand, the vertical distribution of water vapor in the middle atmosphereis affected by turbulent processes.The limitof50x 10-6fortheH2mixing ratio.In thesamesense, use of lower values of the eddy diffusion coefficient leads to Liu and Donahue [ 1976] obtained a volume mixing ratio for a drier thermosphere,sincethe upward transport of H20 is H2 equalto 11x 10-6 anda valuefortheH2 concentrationaffected by the values of K in the 80-110 km altitude range. of 1.3x 105cm-3 at200km.Ourvaluefor[H2]of 1.5x 105 The OH and HO2 diurnal profiles are very similar at cm-3 is verycloseandit doesnotexceedthe upperlimit altitudes above 40 km, and both profiles present the maxireported by Moos [1974], who evaluated a maximum value mum density at the lower boundary of the model. The OH of 2.7 x 105cm-3 at 250 km from the absence of H2 concentrationdecreasesrapidly below ---45 km at nighttime ultraviolet emission. because of the absence of photodissociationprocesses. It There are no direct in situ or remote sensing measure- follows from the absence of the hydroxyl OH Meinel bands ments of the vertical distribution of H20 in the Mars' in the Martian atmosphere [Krasnopolsky and Krysko, 1976] atmosphere. However, Viking observations with different and from the production mechanism (R10) of vibrationally angularconditionsabove the Viking lander site were usedby excited OH that midnight OH concentration has an upper Davies [1979] to derive a vertical profile of H20 relative to limitof 105cm-3, andthisis accomplished by ourmodel. After sunset, the concentration of some active comthat of dust, although the method cannot give absolute values of H20 concentration. This notwithstanding,it is pounds,as O(3p), OH, H, and O(•D), dramatically deknown that the H20 content in the Mars atmospherede- creases,becomingpractically zero at the lower regions. As it creasesvery rapidly with height in the altitude region near is theoretically predicted, a considerable decrease of the the surface.The presentmodelgivesan H20 volumemixing HO2 concentration occurs below 40 km. This decrease is ratioequalto 1.1 x 10-8at 30km,andit remains practically less abrupt at higher altitudes, and the HO2 concentration constant up to about 100 km. These results are compatible remains practically constant at any altitude higher than 60 with the non-steady theoretical results of Krasitskii [1978]. km. The diurnal OH variation, at altitudes below 40 km (see From the analysis of the profiles obtained in our model, it can be deduced that diurnal variation of water vapour is Figure 13a), shows similar characteristics to those exhibited negligiblein the middle and upper atmosphere.The equilib- by HO2 and H variations, with a sharpincrease(decrease)at rium situation for the water vapor concentration is main- sunset(sunrise). At higher altitudes, the day-to-night variatained as a result of the balance between the direct loss by tion is smaller, and maximum values for the OH concentraphotodissociation andby reactionwith O(•D) andits pro- tion are obtained before sunrise. After that time, a gradual duction throughout odd-hydrogen compounds ((R13) and decreaseof the OH concentration takes place until a mini(R15) in Table 1). A small deficit exists in this balance which
mum value after sunset is reached.
14,806
The measurements
RODRIGO ET AL..' MARS' NEUTRAL ATMOSPHERIC COMPOSITION BETWEEN 30 AND 200 KM
carried out with different instrumenta-
tion on board Mariner 6, 7, and 9 [Anderson and Hord, 1971; Anderson, 1974], Mars 3 [Dostovalov and Chuvakhin, 1973], and Mars 5 [Bertaux et al., 1975] give an H concentration in
therangeof values(1 - 3.3) x 104cm-3 at altitudes higher than 200 km. Anderson and Hord [1971] deduced a value for
for O and CO) in this region, down to values incompatible with the experimental results. 3.2.
Oxygen Species
Figure 11 shows the altitude concentration profiles ob-
the H concentration equalto (3 -+ 1) x 104cm-3 at the tained for the O compoundsbetween 30 and 200 km at noon and midnight. The existence of a secondary maximum of
critical level of atomic hydrogen escape (at around 250 km) if an exospherictemperature of --•350-+ 100 K is assumed.A detailed study of the atomic hydrogen diffusion and escape problem in the thermosphere was made by Anderson [1974]. The main source of atomic hydrogen at altitudes above --•130 km is upward flux for H2 followed by reactions of H2 with
tions [McElroy et al., 1977; Krasnopolsky, 1986]. The atomic oxygen is the dominant constituent among the odd-oxygencompoundsduring the day in the altitude range
ioniccompounds ((R24)and(R25))witha netresult:CO•- +
considered in themodel.We haveobtained a daYt•ime peak
H 2 + e -• CO2 + 2H. Part of this atomic hydrogen is downward transported, where it recombinates to produce H 2 via reactionswith other odd-hydrogencompounds.The rest of atomic hydrogenis transportedupwardsand finally a fraction of its concentration escapes from the atmosphere with a mean flux of the order of 108 cm-2 s-• . This value
agrees verywellwiththeescape fluxof 1.4x 108cm-2 s-1 calculated from Mariner data [Anderson, 1974]. As a result of the processesconsidered, the daytime H profile shows a secondary maximum near 50 km. A secondary maximum is also reached at this altitude at nighttime but is smaller in absolute value by a factor of about 2.
It is well known that the escapeflux of H stronglydepends on the exospheric temperature, and therefore variations in To• yield variation in this flux and thus in the atomic hydrogen concentration at the higher levels of the atmosphere. This dependence was studied by Izakov and Krasitskii [ 1977], who obtained a reduction of a factor of 5 in the H concentration if an increase of 155 K is produced in To•. This influence of To• is also capable of explaining the day-to-night variation obtained in the present model at the top of the atmosphere. Figure 13b shows the diurnal variation of the atomic hydrogen concentration at selected altitudes. The maximum values of--•3 x 104 cm -3 at 200 km
nearmidnight andtheminimum valuesof--•3.5x 103cm-3 at the same altitude near noon, correspondto the lowest and highest values of the exospheric temperature, respectively. At altitudes below 140 km, the day-to-night variation can be considered as negligible down to 60 km. In the lower atmosphere, at 30 km, the atomic hydrogen almost disappears during night because of the absence of its main production mechanisms. At 40 km, the diurnal variation is smaller, and [H] reaches a maximum value at about 0700 LT. In this region, the turbulent processesbegin to be competitive with the photochemical processesand both are responsible for the odd-hydrogen distribution. Atomic and molecular hydrogen are very affected by turbulent diffusion, and therefore their concentrationsare very sensitiveto the eddy diffusioncoefficientprofile used in the model. Thus H 2 and H concentrations decrease as a consequence of smaller values for K in the lower atmosphere, and this decrease is more dramatic at the upper boundary of the model. In the thermosphere, however, as K increases, the total abundancesof H2 and H are seriouslyaffectedat altitudesabove
odd-oxygen(O + 03) at about 40 km is readily apparentfrom Figure 11, and this is in accordance with previous estima-
of atomic oxygen centered at 48 km, with a value of 7.4 x
109cm-3. At nighttime, thealtitudeof thepeakis higher,at 56km,anditsmagnitude is smaller,witha valueof 6.2 x 109 cm-3 . Below40km, atomicoxygenconcentration decreases rapidly during night. This behavior is similar to that displayed in the terrestrial atmosphere at equivalent densities [see, e.g., Rodrigo et al., 1986, 1989] and predicted by Shimazaki [1981] in his analysis of the temporal variations in ozone density in the Martian atmosphere. The concentrations obtained in the present model are smaller than the
upperlimitof 10l• cm-3 at 80 km requiredto explainthe oxygen nightglow measurementsat 557.7 and 630 nm carried out by the Mars 5 orbiter [Krasnopolsky and Krysko, 1976]. The diurnal variation of the atomic oxygen concentration at selected
altitudes
between
30 and 200 km is shown in
Figure 13c. The most outstanding feature is the sharp variation
in O concentrations
at altitudes
below 40 km after
sunsetand before sunrise, as a consequenceof the presence (absence) of the photodissociationprocesseswhich are the main production mechanismsof atomic oxygen. At 40 km, dynamical processesbegin to be competitive with the photochemical processes, and in this way, O concentration shows a less dramatic change in the day-to-night transitions. At higher altitudes, this variation is still smaller, and a maximum O concentration is reached at sunset and a minimum at sunrise as a result of the contributions of the
photochemical processesduring daylight and of the downward transport. At 100 km and above, atomic oxygen remains practically constant throughout the day sincetransport phenomena are predominant. Carbon monoxide and atomic oxygen flow downward with values for the rate of
fluxof about108cm-2 s- • at 200km, andthisis in excellent agreement with the values reported by Liu and Donahue [1976] and Krasnopolsky and Parshev [1979]. This flow becomesgreater at lower altitudes because of CO2 photodissociation. Thus downward transport of atomic oxygen occurswith valuesfor the flux of 2.6 x 10TMand 1.8 x 10•ø
cm-2 s-• at 60and100kin, respectively. Thisfluxbalances the important loss of atomic oxygen by recombinationat the lower levels. On the other hand, atomic oxygen is one of the compoundsmost affected by turbulent mixing in the middle atmosphere, and its concentration increases in the 70-110 km region up to 1 order of magnitude, when a value 5 times
greaterfor K is useddownto 100km. If a valueof 4 x 108 cm2 s-1 is adoptedfor K in the 40-70 km region,O
90 km, the H2 concentrationincreasingup to almost2 orders
concentrations
of magnitude at 200 km if K increases in a factor of 5. In general, higher values of K at altitudes above 100 km produce lower abundancesof H (and the same is also valid
magnitude in the same altitude range, and these values are not compatible with observations. At altitudes above 100 kin, the use of greater values for K yields lower O abun-
would
increase
in more than 2 orders of
RODRIGO ETAL.'MARS' NEUTRAL ATMOSPHERIC COMPOSITION BETWEEN 30AND200KM dances at these altitudes as turbulent diffusion becomes
14,807
is consideredand then if the atmospherictemperatureis as
forthe03 mixing ratioof4 x 10-8 morecompetitivewith moleculardiffusion,and the values lowas140K. Ourvalue is consistent with more recent values of 5 x 10-8 calculated obtainedfor [O] are againincompatiblewith the experimental results.
The diurnal variation of atomic oxygen obtained in this
by Krasnopolsky[1986]. The secondarymaximumandits diurnalvariationare very
modelis in goodagreementwith that deducedfrom airglow measurements at 130.4nm by Mariner 6, 7, and 9 basedon the variabilityof the emissionprofilesnearlimb. The analysisof thesedata showsthat the atomicoxygenconcentrationis greaternearsunsetthanduringthe earlymorning(in someoccasionsup to a factor of 3) [Stricklandet al., 1972, 1973;Kondratyevand Hunt, 1982;Alexanderet al., 1989].
wellreproduced in thismodel(seeFigure13d).At 50kmand
between 1.5 and 6%, were deduced from Mars 3 spectro-
to highervaluesof the 03 concentration.
above, after the considerabledecreaseat sunrise,the concentrationincreasesduringthe day until sunset.This result is in accordance with the predictionsof Shimazaki[1981].At these altitudes the different photochemicaland turbulent
processes compete.The varyingbalancebetweenthem,as the dayprogresses, mayaccountfor the changes shownin are very affectedby the From the same measurements,a value for the atomic oxygen Figure13d.The 03 concentrations mixing ratio of---1% at the ionosphericpeak has been value used for the eddy diffusioncoefficient.In the lower a lowerdensityof turbulentmixingcorr,esponds inferred. Values somewhathigher at the ionosphericpeak, atmosphere, From the absenceof OH Meinel bandsin the Martian scopicobservations [Dementyevaet al., 1972]as well as Krasnopolsky and Krysko[1976]evaluatedan from the derivedCO• numberdensities[Krasnopolsky,atmosphere, upper limit for the [HI x [03] product of"•1015 cm-6 at60 1975]. Stewart [1972]found an O mixing ratio of 2% by comparing the measuredand calculatedintensities of the km. The Valueobtainedat midnightin the modelfor this is of 1.3x 10TM cm-6, belowtheupperlimit.By CO• bands.Atomicoxygenwasnot directlymeasured by product andlossmechanisms of the Viking massspectrometers. However, a valuefor its takingintoaccountthe production vibrationally excited OH proposed by L6pez-Moreno et al. mixingratioof 7.2 x 10-3 at 140kmcanbededuced. Onthe of the compounds calculated other hand, Hanson et al. [1977],by usingion data, estab- [1987]andthe concentrations of the OH Meinel lishedthat an O mixingratio of 1.25%at 130 km matched here,it may be deducedthe weakness verywelltheabsolute concentrations ofbothO• andCO•. systemin the Martianatmosphe[e. The resultsobtainedin the presentmodel,with a value for
[O]of 108cm-3 at 135kmandanatomicoxygenmixingratio of 7.4 x 10-3 at 140 km, are in completeagreementwith thosededucedfor the interpretationof the dayglowat 130.4 nm and those inferred from Viking 1 and 2 measurements.
4.
CONCLUDING
REMARKS
A new nonsteadyone-dimensional theoreticalmodelof the Martianatmospherehasbeendeveloped.This allowsus
Atomicoxygen electronically excited at state1Dis pro- to determine the vertical distribution of different atmospheric compounds asa functionof thelocaltimedespite the duced byphotodissociation processes. O(1D)concentration
profile at noonis shown in Figure11.Although theO(1D) limitationsimposedby the unavailabilityof data on the number density is extremely low, this specie plays an impo•ant role in the photochemistry of the Martianatmosphere.Becauseof the absenceof the productionmechanisms,this specierapidly disappearsat night. Ozone is the only active compoundwhich has been experimentally measured in the loweratmosphere. 03 total content was first determinedby means of the ultraviolet spectroscopy experimenton boardMariner 7 [Barthand Hord, 1971;Barth et al., 1971]and by Mariner 9 [Barth et
nighttime structure of the atmosphere. We havecarriedout a studyof the possibletemporalvariationsof the temperature as well as of differentphotochemical processes taking placein the differentatmospheric regions.Turbulentdiffusion, which plays a primary role in the atmosphere,is parameterized by means of aneddydiffusion coefficient K(z) capableof explaining the observedabundance of carbon
bution of ozone concentration obtained in the model is
of the differentaeronomicalprocessestaking place in the
shownin Figure 11.A secondarymaximumis apparentat 40 km. The valuesof the 03 concentration increaseat night, mainlydue to the absenceof 03 photodissociation. These resultsagreewell with thoseobtainedby Krasnopolsky et al. [1977, 1979] from limb ultraviolet photometric measure-
middle atmosphereof Mars.
monoxidein the Martian atmosphere.This eddy diffusion
coefficient provides uswithinformation aboutthedynamical andallowsus, togetherwith the al., 1973;Lane et .al., 1973],and a great seasonal and structureof the atmosphere schemeproposed, a reasonable description latitudinal variation has been detected. The altitude distri- photochemical
ments on board the Mars 5 orbiter. These observations
revealan ozonelayer centeredat --•40km and with a width of--•10 km. They reportedan ozone concentrationin the
The resultsshowthat turbulenceplays a primary role in the distributionof the atmosphericcompoundsat altitudes below 130-140km, where the turbopausecan be placed. Vertical mixing is strongerthan in the terrestrial atmo-
sphere,reachingK(z) valuesup to 2 ordersof magnitude greaterat levelsof the sameatmospheric densities. Values
ashighas4 x 108cm2s-1 areobtained inthetherrnosphere, evening (atabout1800LT) of (3 -+1.5)x 109cm-3 at the which are compatiblewith other estimations.At altitudes layermaximum and(10-+1.5)x 109cm-3 onthemorningabove 160 km, moleculardiffusionpredominates,and the constituents accomplish sidelimb(at about0930LT). Thisleadsto an 03 mixingratio verticalprofilesof the atmospheric of 2 x 10-6. Thesevaluesare greater,in almost1 orderof diffusive equilibrium. The concentrationof the odd hydrogen compounds,H, magnitude,than thosegenerallyaccepted.In order to exday-to-night variationas plain the discrepancy, theseauthorsinvokedthe possible OH, andHO2, showsa noticeable of the changein the physicaland chemical influenceof heterogeneous reactionswith the surface.In the a consequence accounting for theirverticaldistribution. sameway, Shimazaki[1981]pointedoutthatit is possible to processes Atomic oxygen and ozonepresenta maximumconcentrareproduce thesehighvaluesif an extremelydry atmosphere
14,808
RODRIGO ETAL' MARS' NEUTRAL ATMOSPHERIC COMPOSITION BETWEEN 30AND 200KM
tionaround40 km. At loweraltitudes, theatomicoxygen theglobaltemperature structure of the thermosphere of Venus andMars:A comparison, Eos Trans.AGU, 66, 323, 1985. concentrationrapidly decreasesat night whereasozone S. W., R. E. Dickinson, R. G. Roble,andE. C. Ridley, concentration increases in all thealtituderangeaftersunset. Bougher, In general, theresultsof themodelarein goodagreement Mars thermosphericgeneralcirculationmodel: Calculationsfor thearrivalofPhobos atMars,Geophys. Res.Lett.,15,1511-1514, with the scarceavailabledata. In orderto improveour 1988. knowledge on the structure andcomposition of theMartian Carleton, N. P.,andW.B.Traub,Detection ofmolecular oxygen on atmosphere, it isnecessary to carryoutbothnewlaboratory Mars, Science,177, 988-992, 1972. S., andT. G. Cowling, TheMathematical Theory of measurementsof the absorption cross sections and rate Chapman, Non-Uniform Gases,3rded., Cambridge UniversityPress,New coefficients for the temperature andpressureconditions of York, 1970.
Mars and new in situ experimentalmeasurements.In this
sense,thenewMarsobserver mission willprovidea greater data basisand will allow a better and more accuratedeter-
Clancy, R. T., D. U. Muhleman, andB. M. Jakosky, Variability of carbonmonoxide in theMarsatmosphere, Icarus,55, 282-301, 1983.
F. D., W. B. Hanson, andF. S.Johnson, Eddydiffusion minationof someimportant atmospheric parameters aswell Colegrove, andoxygentransport in the lowerthermosphere J. Geophys. as their spacialand temporalvariations. Res., 70, 4931-4941, 1965.
Acknowledgments. We thanktwo refereesfor their helpful
suggestions. This researchwas supported by the Comisi6nInterministerial de Cienciay Tecnologfa undercontractESP88-0566 and by the Junta de Andalucfa.
Colegrove, F. D., F. S. Johnson, andW. B. Hanson, Atmospheric composition in the lowerthermosphere, J. Geophys. Res., 71, 2227-2236, 1966.
Connes, J., P. Connes, andJ.P. Maillard,Atlasdesspectres infrarouges de Venus,Mars,Jupiteret Saturne,CentreNational de la RechercheScientifique,Paris, 1969.
Davies,D. W., Theverticaldistribution of Marswatervapor,J. REFERENCES
Geophys.Res., 84, 2875-2879, 1979.
Dementyeva, N. N., V. G. Kurt,A. S. Smirnov,L. G. Titarchuk,
Ackerman, M., Ultraviolet solarradiation relatedto mesospheric andS. D. Chuvakhin,Preliminaryresultsof measurements of UV processes, in Mesospheric Modelsand RelatedExperiments, emissions scattered in theMartianupperatmosphere, Icarus,17, editedby G. Fiocco,pp. 149-159,D. Reidel,Hingham, Mass., 475-483, 1972. 1971. DeMore,W. B., andH. Patapoff, Temperature andpressure depenAlexander, M. J., A. I. F. Stewart,C. G. Fesen,R. R. Meier,L. J. denceof CO2 extinctioncoefficients, J. Geophys. Res., 77,
Paxton,and S. W. Bougher,Atomic oxygenin the martian 6291-6293, 1972. thermosphere, Bull.Am. Astron.Soc.,21,990, 1989. DeMore,W. B., J. J. Margitan,M. J. Molina,R. T. Watson,D. H. Anderson, D. E., Mariner6, 7 and 9 ultravioletspectrometer Golden,R. F. Hampson, M. J. Kurylo,C. J. Howard,andA. R. experiment: Analysis ofhydrogen Lymanalphadata,J. Geophys. Ravishankara, Chemicalkineticsandphotochemical datafor use Res., 79, 1513-1518, 1974. in stratospheric modelling, Evaluation no. 7, JPL Publ.,85-37, Anderson,D. E., and C. W. Hord, Mariner6 and 7 ultraviolet 217 pp., 1985. spectrometer experiment:Analysisof hydrogenLymana data,J. Geophys.Res., 76, 6666-6673,1971.
Barker, E. S., Detectionof molecularoxygenin the martian atmosphere,Nature, 238, 447-448, 1972.
Dostovalov,S. B., and S. D. Chuvakhin,Distributionof neutral
hydrogen in the Martianupperatmosphere, CosmicRes.,11,
687-692, 1973.
A., Eddydiffusion modelsfor the mesosphere and lower Barth,C. A., Thephotochemistry oftheatmosphere ofMars,in The Ebel, thermosphere, J. Atmos.Terr.Phys.,42, 617-628,1980. Photochemistry of Atmospheres: Earth, the OtherPlanets,and Fjeldbo, G., A. Kliore, andB. Seidel,The Mariner 1969occultation Comets,editedby J. S. Levine,pp. 337-392,Academic,San Diego, Calif., 1985.
Barth,C. A., andC. W. Hord,Mariner6 and7 ultraviolet experiment:Topography andpolarcap,Science,173, 197-201,1971.
measurements of the upperatmosphere of Mars, Radio Sci., 5, 381-386, 1970.
Fox,J. L., andA. Dalgarno, Theproduction of nitrogen atomson Marsandtheirescape,Planet.SpaceSci., 28, 41-46, 1980.
Barth,C. A., C. W. Hord,J. B. Pearce,K. K. Kelly, G. P. Anderson, andA. I. Stewart, Mariner6 and7 ultraviolet spec- Good,J. C., and F. P. Schloerb,Martian CO abundancefrom theJ = 1 -->0 rotationaltransition: Evidencefor temporalvariations, trometer experiment: Upperatmosphere data,J. Geophys. Res., Icarus, 47, 166-172, 1981. 76, 2213-2227, 1971.
datasheetsfor Barth,C. A., A. I. Stewart,C. W. Hord,andA. L. Lane,Mariner Hampson,R. I., Chemicalkineticandphotochemical atmospheric reactions, Rep.FAA-EE-80-17, U.S. Dep.of Trans9 ultraviolet spectrometer: Marsairglowspectroscopy andvaria-
tionsin Lymanalpha,Icarus,17, 457-468,1972.
portation, Washington,D.C., 1980.
variationof ozoneon Mars,Science,179,795-796,1973.
ers, J. Geophys.Res., 82, 4351-4363, 1977.
Barth,C. A., C. W. Hord,A. I. Stewart,A. L. Lane,M. L. Dick, Hanson,W. B., S. Sanatani,and D. R. Zuccaro,The martian asobserved by theVikingretarding potential analyzandG. P. Anderson, Mariner9 ultraviolet experiment: Seasonal ionosphere
Bates,D. R., andM. Nicolet,Thephotochemistry of atmosphericHedin, A. E., A revised thermosphericmodel based on mass water vapour, J. Geophys.Res., 55, 301-327, 1950.
spectrometer andincoherent scatterdata:MSIS-83,J. Geophys.
Battaner, E., Transport phenomena andthechemical composition Res., 88, 10,170-10,188, 1983. in the mesosphere and lower thermosphere, J. Atmos.Terr. Hedin,A. E., MSIS-86thermospheric model,J. Geophys. Res.,92, Phys., 37, 1155-1165, 1975.
4649-4662, 1987.
Battaner, E., andR. Rodrigo, Mesospheric 03, H andH20 at high Hedin,A. E., H. B. Niemann,W. T. Kasprzak,andA. Seiff,Global latitudes' A theoretical model,Planet.SpaceSci.,22, 819-823, empirical modelof the Venusthermosphere, J. Geophys. Res., 1981. 88, 73-83, 1983.
Bertaux, J. L., J. Blamont, S. I. Babichenko, N. N. Dementyeva,Huguenin, R. L., Chemical weathering and the Vikingbiology A. V. Dyachkov, V. G. Kurt, V. A. Sklyankin, A. S. Smirnov, experiments on Mars,J. Geophys. Res.,87, 10,069-10,082, 1982. andS. D. Chuvakhin, Measurements of intensity andspectral Hunten,D. M., Aeronomyof Marsloweratmosphere, Rev. Geoproperties of Lyman-ain theupperatmosphere of Mars,Cosmic
Res., 13, 35-40, 1975.
Bittner,H., andK. H. Fricke,Dayside temperatures of theMartian upperatmosphere, J. Geophys.Res., 92, 12,045-12,055, 1987.
phys., 12,529-535, 1974.
Inn, E. C. Y., K. Watanabe,andM. Zelikoff,Absorption coeffi-
cientsofgases in thevacuum ultraviolet, partIII, CO2,J. Chem.
Phys., 21, 1648--1650,1953.
Bougher,S. W., and R. E. Dickinson,Mars mesosphere and Izakov,M. N., The Martianupperatmosphere structurefromthe thermosphere, 1,Globalmeanheatbudget andthermal structure, Vikingspacecraft experiment, Icarus,36, 189-197,1978. J. Geophys.Res., 93, 7325-7337, 1988. Izakov,M. N., andO. P. Krasitskii, A modelforthecomposition of
Bougher,S. W., R. E. Dickinson,andR. G. Roble,Calculations of
theMartianatmosphere, CosmicRes., 15, 391-404,1977.
RODRIGO ET AL.: MARS' NEUTRAL ATMOSPHERICCOMPOSITION BETWEEN 30 AND 200 KM
Izakov, M. N., and S. K. Morozov, Structure and dynamics of the equatorial thermosphere of Mars, Cosmic Res., 14,476-478, 1976. Kaplan, L. D., J. Connes, and P. Connes, Carbon monoxide in the Martian atmosphere, Astrophys. J., 157, LI87-LI92, 1969. Kasting, J. F., S. C. Liu, and T. M. Donahue, Oxygen levels in the prebiologicalatmosphere,J. Geophys.Res., 84, 3097-3107, 1979. Kerzhanovich, V. V., Mars 6: Improved analysis of the descent module measurements, Icarus, 30, 1-25, 1977. Kliore, A. I., G. Fjeldbo, and B. L. Seidel, Summary of Mariner 6 and 7 radio occultation results on the atmosphere of Mars, Space Res., XI, 165-175, 1971. Kliore, A. I., D. L. Cain, G. Fjeldbo, B. L. Seidel, M. J. Sykes, and S. I. Rasool, The atmosphere of Mars from Mariner 9 radio occultation measurements, Icarus, 17, 484-516, 1972. Kockarts, G., Absorption and photodissociation in the SchumannRunge bands of molecular oxygen in the terrestrial atmosphere, Planet. Space Sci., 24, 589-604, 1976. Kondratyev, V. N., and G. E. Hunt, Weather and Climate on Planets, Pergamon, New York, 1982. Kong, T. Y., and M. B. McElroy, Photochemistry of the Martian atmosphere, Icarus, 32, 168-189, 1977a. Kong, T. Y., and M. B. McElroy, The global distributionof 0 3 on Mars, Planet. Space Sci., 25,839-857, 1977b. Krasitskii, O. P., A model for the diurnal variation of the composition of the Martian atmosphere, Cosmic Res., 16,350--356, 1978. Krasnopolsky, V. A., On the structure of Mars' atmosphere at 120--220 km, Icarus, 24, 28-35, 1975. Krasnopolsky, V. A., Photochemistry of the Atmospheres of Mars and Venus, Springer-Verlag, New York, 1986. Krasnopolsky, V. A., and A. A. Krysko, On the night airglow of the Martian atmosphere, Space Res., XVI, 1005-1012, 1976. Krasnopolsky, V. A., and V. A. Parshev, Ozone and photochemistry of the Martian lower atmosphere, Planet. Space Sci., 27, 113-120, 1979. Krasnopolsky, V. A., A. A. Krysko, and V. N. Rogachev, Ultraviolet photometry of Mars by the satellite Mars 5, Cosmic Res., 15,214-218, 1977. Krasnopolsky, V. A., V. A. Parshev, A. A. Krysko, and V. N. Rogachev, Structure of Mars' lower and middle atmosphere,Rep. 483, Acad. of Sci., Space Res. Inst., Moscow, 1979. Lane, A. L., C. A. Barth, C. W. Hord, and A. I. Stewart, Mariner 9 ultraviolet spectrometer experiment: Observations of ozone on Mars, Icarus, 18, 102-108, 1973. Leovy, C. B., J. E. Tillman, W. R. Guest, and J. Barnes, Interannual variability of Martian weather, in Recent Advances in Planetary Meteorology, edited by G. E. Hunt, pp. 69-84, Cambridge University Press, New York, 1985. Lettau, H., Diffusion in the upper atmosphere, in Compendium of Meteorology, edited by T. F. Malone, pp. 320--333, American Meteorological Society, New York, 1951. Lewis, J. S., and J. G. Prinn, Planets and Their Atmospheres: Origin and Evolution, Academic, San Diego, Calif., 1984. Lindner, B. L., Ozone on Mars: The effect of clouds and airborne dust, Planet. Space Sci., 36, 125-144, 1988. Lindner, B. L., and B. M. Jakosky, Martian atmospheric photochemistry and composition during periods of low obliquity, J. Geophys. Res., 90, 3435-3440, 1985. Liu, S.C., and T. M. Donahue, The regulation of hydrogen and oxygen escape from Mars, Icarus, 28, 231-246, 1976.
14,809
Composition and structure of the Martian upper atmosphere. Analysis of results from Viking, Science, 194, 1295-1298, 1976. McElroy, M. B., T. Y. Kong, and Y. L. Yung, Photochemistry and evolution of Mars' atmosphere: A Viking perspective, J. Geophys. Res., 82, 4379-4388, 1977. Molina, L. T., S. D. Schinke, and M. J. Molina, Ultraviolet absorption spectrum of hydrogen peroxide vapor, Geophys. Res. Lett., 4, 580--582, 1977.
Moos, H. W., An upper limit on H2 ultraviolet emissionsfrom the Martian exosphere, J. Geophys. Res., 79, 2887-2891, 1974. Mount, G. H., and G. J. Rottman, The solar absolute spectral
irradiance1150--3173 ]k: May 17, 1982,J. Geophys. Res., 88, 5403-5410, 1983.
Nicolet, M., The solar spectral irradiance and its action in the atmosphericphotodissociationprocesses,Planet. Space Sci., 29, 951-974, 1981a.
Nicolet, M., The photodissociation of water vapor in the mesosphere, J. Geophys. Res., 86, 5203-5208, 1981b. Nicolet, M., Photodissociationof molecular oxygen in the terrestrial atmosphere:Simplified numerical relations for the spectral range of the Schumann-Rungebands, J. Geophys. Res., 89, 2573-2582, 1984.
Nicolet, M., Aeronomical aspects of mesospheric photodissociation: Processesresultingfrom the solar Lyman-alpha line, Planet. Space Sci., 33, 69-80, 1985.
Nicolet, M., and W. Peetermans,Atmosphericabsorptionin the O2 Schumann-Rungeband spectral range and photodissociationrates in the stratosphere and mesosphere, Planet. Space Sci., 28, 85-103, 1980.
Nier, A. O., and M. B. McElroy, Structure of the neutral upper atmosphereof Mars: Resultsfrom Viking 1 and Viking 2, Science, 194, 1298-1300, 1976. Nier, A. O., and M. B. McElroy, Composition and structure of Mars' upper atmosphere:Resultsfrom the neutral mass spectrometers on Viking I and 2, J. Geophys. Res., 82, 4341--4349, 1977. Nier, A. O., W. B. Hanson, A. Seiff, M. B. McElroy, N. W. Spencer, R. J. Duckett, T. C. D. Knight, and W. S. Cook, Composition and structure of the Martian atmosphere: Preliminary results from Viking 1, Science, 193,786-788, 1976. Owen, T., The compositionof the Martian atmosphere, Adv. Space Res., 2, 75-80, 1982. Owen, T., K. Biemann, D. R. Rushneck, J. E. Biller, D. W.
Howarch, and A. L. Lafleur, The compositionof the atmosphere at the surface of Mars, J. Geophys. Res., 82, 4635--4639, 1977. Parisot, J., and J. Zucconi, Temperature effect on the photodissociation rates in the atmospheresof Mars and Venus, Icarus, 60,
327-331, 1984. Parkinson, T. D., and D. M. Hunten, Spectroscopy and aeronomy of 02 on Mars, J. Atmos. Sci., 29, 1380--1390, 1972. Paukert, T. T., and H. S. Johnston, Spectra and kinetics of the hydroperoxyl free radicals in the gas phase, J. Chem. Phys., 56, 2824-2838, 1972. Rodrigo, R., J. S. Nisbet, and E. Battaner, The effect of horizontal winds upon the chemical composition of the lower thermosphere at high latitude, J. Geophys. Res., 86, 3501-3508, 1981. Rodrigo, R., J. J. Lrpez-Moreno, M. L6pez-Puertas, F. Moreno, and A. Molina, Neutral atmospheric compositionbetween 60 and 220 km: A theoretical model for mid-latitudes, Planet. Space Sci., L6pez-Gonzfilez, M. J., J. J. L6pez-Moreno,E. Garcia-/dvarez, 34, 723-743, 1986. Rodrigo, R., J. J. L6pez-Moreno, M. J. L6pez-Gonz•ilez, and E. and R. Rodrigo, Behaviour of the 02 infrared atmospheric(0-0) Garcia-Alvarez, Atomic oxygen concentrationsfrom OH and O2 band in the middle atmosphereduring evening twilight and night, nightglow measurements,Planet. Space Sci., 37, 49-60, 1989. Planet. Space Sci., 37, 61-72, 1989. Rodrigo, R., E. Garcia-Alvarez, M. J. L6pez-GonzSlez, and M. A. L6pez-Moreno, J. J., R. Rodrigo, F. Moreno, M. L6pez-Puertas, L6pez-Valverde, Estimates of eddy diffusion coefficient in the and A. Molina, Altitude distribution of vibrationally excited states Mars' atmosphere, Atmosfera, 3, 31--43, 1990. of atmospherichydroxyl at levels v = 2-7, Planet. Space Sci., 35, 1029-1038, 1987. Sch0rgers, M., andK. H. Welge,Absorptioonskoeffizient vonH202 und N2H4 zwischen 1200 und 2000 A, Z. Naturforsch., 23, L6pez-Moreno, J. J., R. Rodrigo, F. Moreno, M. L6pez-Puertas, 1508-1510, 1968. and A. Molina, Rocket measurementsof 02 infrared atmospheric Seiff, A., and D. B. Kirk, Structure of the atmosphere of Mars in system in the nightglow, Planet. Space Sci., 36, 459-467, 1988. Marov, M. Y., V. V. Kerzhanovich, V. S. Avduevsky, and M. K. summer at midlatitudes, J. Geophys. Res., 82, 4364--4378, 1977. Rozhdestvensky, Measurements with Mars-6 lander and a model Shemansky, D. E., CO2 extinction coefficient 1700-3000 •, J. of the Martian atmosphere, Space Res., XVI, 1019-1027, 1976. Chem. Phys., 56, 1582-1587, 1972. McElroy, M. B., and T. M. Donahue, Stability of the Martian Shimazaki, T., A model of temporal variations in ozone density in the Martian atmosphere, Planet. Space Sci., 29, 21-33, 1981. atmosphere, Science, 177, 986-988, 1972. Shimazaki, T., and M. Shimizu, The seasonal variation of ozone McElroy, M. B., T. Y. Kong, Y. L. Yung, and A. O. Nier,
14,810
RODRIGOET AL.: MARS' NEUTRAL ATMOSPHERIC COMPOSITIONBETWEEN 30 AND 200 KM
density in the Martian atmosphere, J. Geophys. Res., 84, 12691276, 1979.
Singhall, R. P., and R. C. Whitten, Thermal structure of the ionosphere of Mars: Simulations with one- and two-dimensional models, Icarus, 74, 357-364, 1988.
Slanger,T. G., B. J. Wood,andG. Black,Kineticsof O(3p) + CO + M recombination, J. Chern. Phys., 57, 233-238, 1972. Stewart, A. I., Mariner 6 and 7 ultraviolet spectrometer experiment:
Implications of the CO•-, CO andO airglow,J. Geophys. Res., 77, 54-68, 1972. Stewart, A. I., and W. B. Hanson, Mars upper atmosphere: Mean and variations, Adv. Space Res., 2, 87-101, 1982. Stewart, A. I., C. A. Barth, C. W. Hord, and A. L. Lane, Mariner 9 ultraviolet spectrometerexperiment: Structure of Mars' upper atmosphere, Icarus, 17, 469-474, 1972. Strickland, D. J., G. E. Thomas, and P. R. Sparks, Mariner 6 and 7 ultraviolet spectrometerexperiment: Analysis of the OI 1304 and
1356.• emissions, J. Geophys. Res.,77,4052-4068, 1972. Strickland, D. J., A. I. Stewart, C. A. Barth, C. W. Hord, and A. L.
Lane,Marsatomicoxygen1304• emission, J. Geophys. Res., 78, 4547-4559, 1973.
Stubbe, P., The thermosphereand the F-region: A reconciliationof theory with observations,Sci. Rep. 418, Ionosphere Res. Lab., Pa. State Univ., University Park, 1973. Thompson, B. A., P. Harteck, and P. R. Reeves, Ultraviolet
absorption coefficients ofCO2o, CO,02, N20, NH3,NO, SO2and CH4 between 1850 and 4000 A, J. Geophys.Res., 68, 6431-6436,
Trauger, J. T., and J. I. Lunine, Spectroscopyof molecular oxygen in the atmospheresof Venus and Mars, Icarus, 55,272-281, 1983. von Zahn, U., and D. M. Hunten, Aeronomy of Mars' lower atmosphere: Status and prospects, in The Planet Mars, Eur. Space Agency ESA Sci. Publ., 185, 37-46, 1982. von Zahn, U., K. H. Fricke, D. M. Hunten, D. Krankowsky, K. Mauersberger, and A. O. Nier, The upper atmosphere of Venus during morning conditions, J. Geophys. Res., 85, 7829-7840, 1980.
Wayne, R. P., Chemistry ofAtrnospheres, Clarendon Press, Oxford, 1985.
Yung, Y. L., and W. B. DeMore, Photochemistryof the structureof Venus: Implications for atmospheric evolution, Icarus, 51, 199247, 1982.
Yung, Y. L., D. F. Strobel, T. Y. Kong, and M. B. McElroy, Photochemistryof nitrogen in the martian atmosphere,Icarus, 30, 26-41, 1977. Zurek, R. W., Diurnal tide in the Martian atmosphere, J. Atrnos. Sci., 33, 321-337, 1976. Zurek, R. W., Solar heating of the Martian dusty atmosphere, Icarus, 35, 196-208, 1978. Zurek, R. W., Martian great dust storms: An update, Icarus, 50, 288-310, 1982.
E. Garc/a-Alvarez, M. J. L6pez-Gonz•ilez, J. J. L6pez-Moreno, and R. Rodrigo, Instituto de Astoffsica de Andalucfa, Apartado de Correos 2144, 18080 Granada, Spain.
1963.
Traub, W. A., N. P. Carleton, P. Connes, and J. F. Noxon, The
latitude variation of O2 dayglow and 03 abundanceon Mars, Astrophys. J., 229, 846-850, 1979.
(Received June 15, 1989; revised December 5, 1989; accepted December 13, 1989.)
