LABORATORY NEEDS FOR PLANETARY ATMOSPHERES
by
Darrell F. Strobel
Johns Hopkins University/Observatoire de Paris - Meudon
Abstract: The laboratory needs for planetary atmospheres including the Earth's
mesosphere and thermosphere will be reviewed. A representative group of
current problems in planetary atmospheres will be discussed and the role that
laboratory measurements can make toward their solutions will be indicated.
This group includes 1) the transmission of radiation through N2 atmospheres of
the Earth, Titan, Triton, (and Pluto) in the 80-100 nm region, 2) the chemistry
of CNN radicals in Triton's atmosphere, 3) the potential role of heterogeneous
chemistry in the atmospheres of the outer solar system, 4) illustrations of
critical low temperature, chemical kinetic reaction rates, 5) inference of
atmospheric SO2 abundances from UV electronic bands, 6) remote sensing of
atomic oxygen in the Earth's mesopause region.
I am been asked to represent both the space science ITM
(ionosphere-thermosphere-mesosphere) community and the planetary science
community at this workshop. Given the broad nature of these two communities, I
cannot in good conscience suggest that the topics selected for this
presentation represent a priority list for laboratory research. Instead the
topics discussed below represent topics worthy of laboratory effort, which
might potentially emerge high on a priority list.
One of the most important constituents in the Earth's thermosphere is
atomic oxygen which plays an important role in the chemistry of the mesopause
and lower thermosphere region and is the dominant specie in the upper
thermosphere. Atomic oxygen is notoriously difficult to measure accurately in
the upper atmosphere and when global measurements are required, the only
available techniques are remote sensing. The three airglow emissions which can
yield information on the atomic oxygen density distribution in the mesopause
regions are radiation from 1) O(1S) green line, 2) O2(1[EQN "Sigma"]) A-band,
and 3) OH(v") Meinel bands. For the O(1S) green line, the generally accepted,
currently believed (as opposed to rigorously demonstrated in the laboratory)
process is the indirect two-step Barth mechanism (Barth, 1961) illustrated in
Fig. 1. Similarly a two-step mechanism is also believed to produce the O2(1
[EQN "Sigma"]) A-band emission (Witt, 1979) as shown in Fig. 1. The OH(v")
Meinel band emission is initiated by the chemical reaction of H and O3 (Bates
and Nicolet, 1950; Meinel, 1950) and the reaction of O and HO2, as also shown
in Fig. 1. All three of these emissions were observed with the HRDI instrument
on the UARS satellite and some observed characteristics of these emissions are
given in Fig. 2, from the paper of Yee et al. (J. Geophys. Res., in press)
where the other references may also be found. As discussed in Yee et al. these
emissions yield relatively consistent information on the atomic oxygen density
distribution in the mesopause region. They also illustrated the sensitivity of
the inferred densities to sources of uncertainty in the input quantities in the
expressions in Fig. 1 that relate volume emission rates ([EQN "eta"]) to atomic
oxygen density ([O]). The TIMED Mission will be using similar methods to
measure atomic oxygen densities (and winds) as on UARS and one would like to
see the mechanisms generating these three airglow emission be put on a rigorous
basis from laboratory measurements with accurately determined rate coefficients
such as the ones given in Fig. 1.
In some cases laboratory measurements are available with, from the point
of view of laboratory scientists, sufficient accuracy that no further
measurements are warranted. Yet the utilization of these laboratory
measurements in chemical models of the middle atmosphere yield inconsistent
densities in comparison with atmospheric measurements. A classical
illustration of this situation is shown in Figs. 3-5, where in Fig. 3 the
MAHRSI OH inferred densities from the CHRISTA/MAHRSI shuttle flight are
compared with standard JPL 94 chemistry in a chemical model of the middle
atmosphere. Summers et al. (Science, 277, 1967, 1997) find that reducing the
rate coefficient for the reaction of O and HO2 by 50% (Model B) brings theory
and measurement into agreement, just as Clancy et al. (J. Geophys. Res., 99,
5465, 1994) required to bring their measurements of O3 and HO2 into agreement
with theory as shown in Figs. 4 and 5. This adjustment in the reaction rate of
O and HO2 has little impact on the chemistry of the lower stratosphere as shown
in Fig. 4. If indeed this is the solution to the reconciliation of atmospheric
measurements with atmospheric models, it is unlikely that one additional
laboratory measurement (yielding the reduced rate) will be universally
accepted, unless a clear flaw can be demonstrated in all previous
measurements. From the perspective of a laboratory scientist, this is high
risk research in the sense that obtaining another rate identical to the
previous values is not careering advancing research.
There are currently four known atmospheres in the solar system where N2 is
the dominant constituent: Earth, Titan, Triton, and Pluto. Voyager spacecrafts
have performed solar occultation measurements of differential absorption of UV
sunlight with their Ultraviolet Spectrometers (UVS) to infer composition,
density and temperature profiles. In the wavelength region 700-1000 A and in
particular 800-1000 A, the absence of detailed laboratory N2 band structure
measurements of rotational line profiles/widths and line positions has
prevented the analysis of UVS data for Titan (Figs. 7-8) and Triton (Fig. 9).
Figure 8 from the Ph. D. thesis of Ron Verack shows large differences between
data and model spectra in this wavelength region. The situation is much more
critical on Triton (Fig. 9) as it is the only Voyager data with information on
atmospheric structure between 50-150 km and on the N2 density profile from the
ground up to 400 km. So far the UVS Triton 800-1000 A has not been analyzed
due to lack of laboratory data.
One of the key issues in the study of Triton's upper atmosphere are the
sources of energy. Some scientists (the Caltech group - Yung and Lyons)
believe that only solar energy is important, where as others (Strobel, Summers,
Stevens, and Krasnopolsky) maintain that approximately equal contributions from
solar EUV radiation and magnetospheric electrons are required. There are three
key sets of observational data that any theory must explain: 1) the inferred
upper thermospheric temperature of 102 [EQN "+-"] 3 K from UVS data, 2) the
atomic nitrogen density profile and escape rate of (10 [EQN "+-"]) 3 x 1024
cm-2 s-1 (Fig. 9), and 3) the observed electron density profiles from the
Voyager radio occultation measurement (Fig. 10). Whereas Strobel, Summers,
Stevens, and Krasnopolsky have shown that the solar plus magnetospheric energy
can explain the observed temperature profiles, no viable solar-only model has
been published. Model electron density profiles depend critically on the key
reaction rate of N2+ + C -> C+ + N2 , for which no laboratory measurements are
available (see Fig. 10). If the rate is in the range of 10-10 - 10-9 cm3 s-1,
then Lyons and Yung have shown that a solar-only model is adequate to explain
the observed electron density profiles. Whereas if the rate is equal to or
less than the upper limit of 10-11 cm3 s-1, for the similar reaction of N2+ + H
-> H+ + N2, then there is general agreement that magnetospheric energy is
needed to sustain an ionosphere at observed densities. For the model
calculations of the atomic nitrogen density profile, the CNN radical plays a
critical role as an intermediate participant in the carbon atom catalytic
conversion of N atoms back to N2 (see Fig. 10).
C + N2 + N2 -> CNN + N2 (1)
CNN + N -> CN + N2 (2)
CN + N -> C + N2 (3)
----------------------------
net N + N -> N2
Only reaction (1) has been measured, and then only at room temperature. If the
reaction rates of (2) and (3) are of order 10-10 cm3 s-1, then large N atom
recombination occurs in Triton's atmosphere (and also large carbon atom
densities for additional reasons) and the corresponding N density profile is
significantly lower than derived from UV data, if only solar EUV energy is
available. Whereas if the reaction rates of (2) and (3) are of order 10-11 cm3
s-1, then smaller N atom recombination occurs with smaller carbon atom
densities and solar plus magnetospheric energy sources can easily account for
the observed N density profile and thermal N atom escape rate.
Photochemical models applied to atmospheres of the outer planets are
generally one dimensional, due to the absence of detailed knowledge on actual
atmospheric dynamics. In 1D models dynamics below the homopause is lumped into
a single transport parameter, the eddy diffusion coefficient, which is assumed
to be identically for all species. To derive an eddy diffusion coefficient
profile, a chemical tracer is needed which is sensitive to atmospheric
transport and for which the chemistry is accurately known. In Fig. 11, the
recent detection of CH3 radicals in Saturn's atmosphere is an excellent example
of a potentially good chemical tracer. The CH3 radical is a product of CH4
photolysis near the homopause and then undergoes downward diffusion to higher
pressures where three body recombination to C2H6 is the predominant chemical
loss process. But its chemistry is still poorly constrained due to the lack of
an accurate rate coefficient for CH3 recombination at the low temperatures ~
140 K in Saturn's atmosphere. The two available reaction rates for this
termolecular reaction were measured at room temperature and higher and must be
extrapolated to 140 K, where their extrapolated values differ by a factor of
15. Thus the usefulness of CH3 as a tracer awaits accurate low temperature
measurements of CH3 recombination.
I do note that laboratory techniques are available for making low
temperature measurements as the recent paper by Picard and Canosa (1998)
demonstrates for the reaction rate of CH + N2, which they measured at a
temperature of 53 K (Fig. 12). But considerable efforts in many laboratories
are needed to make measurements of the large number of reactions that are
important at low temperatures.
It has been recognized for over a decade that ozone chemistry in the
Earth's stratosphere and the Antarctic ozone hole in particular cannot be
understood with gas phase chemistry only and that heterogeneous reactions
occurring on particles of polar stratospheric clouds are an essential
ingredient in ozone chemistry. The atmospheres of the giant planets and the
atmospheres of Titan and Triton contain significant amounts haze particles
which can be potentially important surfaces for heterogeneous reactions. Yet
almost all photochemical models and supporting laboratory research focus on gas
phase chemistry exclusively. There is a critical need to address the role of
heterogeneous reactions in the laboratory and models.
The paper by Romani (Icarus, 122, 233, 1996; see Figs. 13-15) cautions
against the expectation of a quick fix to any outstanding problem by one key
rate constant measurement. The outstanding problem addressed in this paper is
the chronic underprediction by models of the observed [C2H6]/[C2H2] ratio.
What Romani shows is that while a new laboratory measurement offered great hope
in solving this long term problem, multiple new laboratory measurements have
conspired in a way to sustain underprediction of the observed [C2H6]/[C2H2]
ratio.
My final topic in this presentation concerns the atmosphere of Io. While
the Galileo spacecraft has been very successful in the study of the Galilean
satellites, the atmosphere of Io is one subject that the Galileo Mission is
only making a minimal impact. Multiple radio occultation measurements by
Galileo spacecraft have substantially increased our knowledge of the
ionosphere, but the Io's ionospheric structure is controlled by the complicated
electrodynamic interaction of Io torus plasma with the atmosphere and
ionosphere and extraction of neutral density profiles from measured electron
density profiles is not straightforward. Currently there are only two ways to
observe the atmosphere directly: 1) at millimeter wavelengths (Lellouch and
colleagues) and 2) in the UV with HST (Ballester, McGrath, Trafton, and
colleagues; see Fig. 16). Right now Jupiter is low in the sky and the airmass
factor is sufficiently large to render millimeter observations not worthy of
radio telescope time. HST provides the best way currently to remotely sense
Io's atmosphere, but interpretation of the data is hindered by the absence of
sufficiently high resolution laboratory measurements in the critical 2000-3000
A region, and for low column densities, the 2000-2150 A region, to determine
the location and width of rotational lines in the electronic band system of SO2
([EQN "C tilde super 1 B sub 2 #3 <- #3 X tilde super 1 A sub 1"]). These
lines are most likely Doppler, but predissociation may alter the line profile,
especially in the wings and all of this information is needed to properly model
the data.
Once the required laboratory measurements have been made there is the
issue of dissemination of the data. While there is a centralized database at
infrared wavelengths (the HITRAN database) for the terrestrial atmosphere,
there are no equivalent databases at visible and ultraviolet wavelengths with
evaluation and recommendation procedures by a panel/group. With rapidly
expanding data acquisition world-wide, it is in the best interests of NASA to
either initiate or, in conjunction with other government agencies, sponsor
centralized databases of evaluated data.
With a new start in the FY99 budget for an Europa Mission there is a need
for laboratory studies of space weathering of ices, salts, and organic
materials in the harsh environment of Io plasma torus. Topics would include
production of gas molecules by sputtering and decomposition, changes in
reflectivity, particle-induced chemistry on surfaces. Given the mission
objectives, identification of decomposition products generated by energetic ion
bombardment of pre-biotic materials is a key issue.
While this talk has focused on the outer planets (and Earth) there are
also laboratory needs for CO2 atmospheres as well. Some of the highest
priority laboratory kinetic needs are: the ClCO equilibrium constant at
temperatures 150-200 K, the net yield of O2(1 [EQN "Delta"]) from the reaction
of 2O + CO2, and the radiative lifetime of O2(1 [EQN "Delta"]) to an 1-sigma
accuracy of 3%. Whether this is possible is another issue.
The detection of x-ray emission from Jupiter's atmosphere and also from
comets brings the need for laboratory data at wavelengths and energies not
commonly encountered by planetary scientists. Of great importance in
understanding the generation of x-rays are charge transfer, electron stripping,
and associated emission cross sections of energetic oxygen and sulfur ions
interacting with atomic and molecular hydrogen.
Peter L. Smith
5/15/1998