Carbon Monoxide in the Atmospheres of the Terrestrial Planets

Citation

Clancy, Robert Todd (1983) Carbon Monoxide in the Atmospheres of the Terrestrial Planets. Dissertation (Ph.D.), California Institute of Technology. doi:10.7907/mf7e-1581. https://resolver.caltech.edu/CaltechTHESIS:01252013-152959933

Abstract

Microwave spectra of carbon monoxide (¹²CO) in the mesosphere of Venus were measured in December of 1978; May and December of 1980; and January, September, and November of 1982. These spectra are analyzed to provide mixing profiles of CO in the Venus mesosphere and best constrain the mixing profile of CO between ~ 100 and 80 kilometers altitude. From the January 1982 measurement (which, of all our spectra, best constrains the abundance of CO below 80 km altitude) we find an upper limit for the CO mixing ratio below 80 kilometers altitude that is 2-3 times smaller than the stratospheric ( ~ 65 km) value of 4.5 ± 1.0 x 10^-5 determined by Connes et al. (1968) in 1967, indicating a possible long-term change in the lower atmospheric concentration of CO.

Intercomparison among the individual CO profiles derived from our spectra indicates considerable short-term temporal and/or spatial variation in the profile of CO mixing in the Venus mesosphere above 80 kilometers. A more complete comparison with previously published CO microwave spectra from a number of authors (Kakar et al., 1976; Gulkis et al., 1977: Schloerb et al., 1980; Wilson and Klein, 1981; Schloerb et al., 1981) specifies the basic diurnal nature of mesospheric CO variability. CO abundance above ~ 95 kilometers in the Venus atmosphere shows approximately a factor of 2-4 enhancement on the nightside relative to the dayside of Venus. The magnitude of this nightside CO bulge is in good agreement with the dynamical modeling results of Dickinson and Ridley (1977), indicating that subsolar to antisolar circulation proposed for the thermosphere of Venus by Dickinson and Ridley extends to below 100 km altitude in the Venus mesosphere. Furthermore, peak nightside CO abundance above ~ 95 kilometers occurs very near to the antisolar point on Venus (local time of peak CO abundance above ~ 95 kilometers occurs at 0.6^+0.7_-0.6 hours after midnight on Venus), strongly suggesting that retrograde zonal flow is substantially reduced at an altitude of 100 kilometers in the Venus mesosphere.

By contrast, CO abundances between 80 and 90 kilometers altitude show a maximum that is shifted from the antisolar point towards the morning side of Venus (local time of peak CO abundance between 80 and 90 kilometers occurs at 8.5 ± 1.0 hours past midnight on Venus). The magnitude of the diurnal variation of CO abundance between 80 and 90 kilometers is again, approximately a factor of 2-4. Given the recombination of CO strongly peaks in this altitude region (Yung and DeMore, 1982), we investigated the possible effects of diurnal photochemistry as a driving force for diurnal CO variations between 70 and 90 kilometers from model calculations. We find that the likely magnitude of diurnal CO variability due to photochemistry and vertical eddy diffusion is smaller than that indicated by the microwave data, and that such variations cannot predict the observed phase behavior of the diurnal variations. Photochemical models invariably predict peak CO abundances in the afternoon rather than morning hours on Venus. However, a simple model for the circulation of the Venus mesosphere is presented to explain the observed diurnal variations of CO both above 90 kilometers altitude and between 80 and 90 kilometers altitude. We propose that the subsolar to antisolar circulation of the Venus thermosphere (and the resulting nightside enhancement of CO) persists down to altitudes of ~ 80 kilometers. Above ~ 90 kilometers zonal flow is small and the nightside CO bulge remains centered near the antisolar point on Venus. Below ~ 90 kilometers altitude retrograde zonal wind velocities increase abruptly to several tens of meters/sec displacing the nightside enhancement of CO towards the morningside of Venus.

We also present a J = 1 → 2 spectrum of ¹³CO absorption in the mesosphere of Venus. This ¹³CO spectrum was measured at the same time as our high quality, January 1982 ¹²CO spectrum. Radiative transfer models employing a single pressure-temperature model of the Venus mesosphere are fit to both the ¹³CO and ¹²CO spectra. The ¹²CO spectrum is used to specify the altitude distribution of CO. Subsequently, we solve for the ratio ¹²CO/¹³CO = 185 in order to best fit the ¹³CO spectrum. Based on an extensive error analysis we believe that the standard deviation of this value is ± 69. This result applies only to the mesosphere of Venus, i.e. from 80 to 110 km. Values of the ¹²CO/¹³CO ratio measured deeper in the Venus atmosphere are closer to the terrestrial value of 89. We suggest several fractionation mechanisms in order to account for the difference between our result and the terrestrial value. However, as yet, none of these mechanisms is known to produce significant fractionation of CO isotopes in the upper atmosphere of Venus.

In January of 1982 we measured a microwave spectrum of CO in the Martian atmosphere utilizing the rotational J = 1 → 2 transition of CO. We have analyzed our data and reanalyzed the microwave spectra of Kakar et al. (1967, measured in 1975) and Good and Schloerb (1981, measured in 1980) in order to constrain estimates of the temporal variability of CO abundance in the Martian atmosphere. Long-term (≳ 1 year) variations in CO abundance have been predicted on the basis of possible variations in eddy diffusion (McElroy and Donahue, 1972) and/or condensible H_xO_y compounds (Hunten, 1974) in the Martian atmosphere. Our values of CO column density from the data of Kakar et al., Good and Schloerb, and our own are 1.7 ± 0.9 x 10²⁰ cm^-2, 3.0 ± 1.0 x 10²⁰ cm^-2, and 4.6 ± 2.0 x 10²⁰ cm^-2, respectively. The most recent estimate of CO column density from the 1967 infrared spectra of Connes et al. (1969) is 2.0 ± 0.8 x 10²⁰ cm^-2 (Young and Young, 1977). The large uncertainties given for the microwave measurements are due primarily to uncertainty in the difference between the continuum brightness temperature and atmospheric temperatures of Mars. We have accurately calculated the variation among the observations of the continuum (surface) brightness temperature of Mars which is primarily a function of the observed aspect of Mars. A more difficult problem to consider is variability of global atmospheric temperatures among the observations, particularly the effects of global dust storms and the ellipticity of the orbit of Mars. The large error bars accompanying our estimates of CO column density from the three sets of microwave measurements are primarily caused by an assumed uncertainty of ± 10 K in our atmospheric temperature model due to possible dust in the atmosphere. A qualitative consideration of seasonal variability of global atmospheric temperatures among the measurements suggests that there is not strong evidence for variability of the column abundance of CO on Mars, although variability of 0-100% over a time scale of several years is allowed by the data set. The implication for the variability of Mars O₂ (which is directly tied to photodissociation of CO₂) is, crudely, a factor of two less. We find that the altitude distribution of C) in the atmosphere of Mars is not well constrained by any of the spectra, although our spectrum is marginally better fit by an altitude increasing profile of CO mixing ratios.

Finally, we consider variations in the CO content of the terrestrial mesosphere. The Earth's mesospheric carbon monoxide was observed in absorption against the Moon in early December of 1979 and late January of 1982 at a wavelength of 1.3 mm, and in early December of 1980 at a wavelength of 2.6 mm. The January 1982 spectrum was also measured in emission with very high signal-to-noise ratios. The observed wavelengths correspond to the respective rotational transitions of CO, J = 1 → 2 and J = 0 → 1. No significant change in the column density of CO above ~ 65 km is found between the 1979 and 1980 observations, but the January 1982 measurement indicates an ~ 30% reduction in column density relative to the December observations. Inversion of the spectra did not provide unique CO mixing ratio profiles for a direct quantitative comparison of December 1979 and 1980 and January 1982 profiles, due to limited signal-to-noise ratios for the 1979, 1980 observations. One of the best constrained mixing profiles published to date is presented for the very high signal-to-noise January 1982 emission spectrum. Comparison with other published spectra of mesospheric CO suggests a large seasonal variation (~ a factor of 2-4) in the column density of CO above 65 km, with a maximum in winter and a minimum in summer. The phase of this seasonal variation in CO abundance is opposite to the phase of seasonal variation in insolation suggesting that a hemispheric pattern of circulation is responsible for seasonal variations in the Earth's mesosphere.

We summarize by noting the very different time scales for variations of CO in the upper atmospheres of Venus, the Earth, and Mars. The long diurnal period of Venus produces a very strong diurnal variation in mesospheric CO which is driven primarily by subsolar to antisolar circulation. CO in the terrestrial mesosphere shows strong seasonal variation which is apparently produced by seasonally driven meridional circulation. By contrast, if atmospheric CO does vary on Mars, it is most likely controlled by long-term changes in the chemistry and/or vertical mixing in the Martian atmosphere.

Thesis Files