Photochemical Modeling of the Earth's Stratosphere

Citation

Froidevaux, Lucien (1984) Photochemical Modeling of the Earth's Stratosphere. Dissertation (Ph.D.), California Institute of Technology. doi:10.7907/hzdt-5z21. https://resolver.caltech.edu/CaltechTHESIS:10012018-103419193

Abstract

We have helped develop a one-dimensional photochemical model of the Earth's stratosphere, in order to provide an up-to-date comparison with mid-latitude observations. This work focuses on the present state of the stratosphere, and includes studies of the radiation field (absorption and scattering), the important partitioning and vertical distribution of halo-carbons and their products, as well as certain intriguing discrepancies related to light and heavy ozone.

We briefly comment on the detection by J. R. Herman and J. E. Mentall of a 10% ratio of total scattered flux to direct solar flux at a wavelength of about 200 nm and an altitude of 40 km. This ratio is over a factor of two higher than our theoretical results and cannot be explained without the existence of a scattering component not included in the model. We also explicitly demonstrate the first-order effects of the inclusion of sphericity (spherical shell atmosphere) on the stratospheric photochemistry at solar zenith angles close to 90°. The resulting changes in model concentrations for short-lived radicals such as O, OH, ClO, and NO are largest in the lower stratosphere, but relatively small compared to current observational uncertainties.

We propose that a significant overestimate of the molecular oxygen absorption cross sections in the important spectral window from about 200 to 220 nm is in large part responsible for the discrepancy between observed and modeled vertical profiles of some halocarbons (CFCl₃ in particular), as well as for the long-standing problem of simultaneously fitting N₂O, CH₄, CF₂Cl₂, and CFCl₃ profiles with a single eddy diffusion model. Recent measurements of atmospheric transmission by J. R. Herman and coworkers seem to support this idea. The use of their proposed reduction in O₂ cross sections leads to significant decreases in the CFCl₃ concentration above about 20 km, with smaller reductions in N₂O, CF₂Cl₂ and HNO₃. The concentrations of CH₄, H₂, and CO are not significantly altered. Changes in other gases (including ozone) are also discussed, as well as the effect on eddy diffusion coefficients obtained from measurements of N₂O or CH₄ profiles in the stratosphere. Accurate determinations of these small O₂ absorption cross sections are needed, since they affect the vertical distribution of halo-carbons in the stratosphere, and the lifetime of these species has an impact on ozone depletion estimates.

In terms of the halocarbon decomposition products in the stratosphere, our model vertical distribution of ClO is shown to provide a reasonably good fit to the mean of available observations. As discussed by others, changes in certain rate constants affecting HO_x in the lower stratosphere have led to decreases in model ClO concentrations by over a factor of three in the lower stratosphere, thus improving the shape of the vertical profile. In addition, the amount of upper stratospheric ClO has increased due to recent changes in the kinetics (reactions O + HO₂, O + ClO, and possibly OH + HCl). The diurnal variation of ClO observed from the ground (microwave emission) by P. Solomon and coworkers is consistent with our model results in terms of the maximum day-to-night decrease in column abundance above about 30 km. However, the observed mid-morning increase is slower than theoretical values, while the predicted afternoon decrease might be too slow, even if one considers the uncertainties in photochemical data. This could indicate the existence of missing chemistry in the models. Although the different observations show somewhat contradictory results. Other observations (balloon-borne microwave spectroscopy and infrared laser radiometry) are also discussed in relation to our model. To first-order, indirect evidence for the breathing cycle between ClO and ClONO₂ seems to have been established. The mean observed HCl mixing ratio profile decreases somewhat faster towards the lower stratosphere than model profiles, a discrepancy which has previously been noted, particularly at high latitudes. Measurements of ethane in the lower stratosphere seemed to indicate that the atomic chlorine concentration was three to five times lower than predicted, but more recent data do not show such a discrepancy.

The fluorine products consist mostly of HF and COF₂. We show that the main uncertainty for this system is the value of the quantum yield (as a function of wavelength) for COF₂ photodissociation, which translates into a factor of three or more uncertainty in the ratio of HF to COF₂ concentrations in the upper stratosphere. If this quantum yield has an average value close to 0.25, a better model fit to observations of HF and [HF]/[HCl] is obtained than if the value is close to unity. Simultaneous stratospheric measurements of COF₂ and HF, as well as ClO and HCl, would greatly enhance our ability to test photochemical models of these halocarbon products.

Finally, we stress that, although generally good agreement is found between our model and observations of HO_x, NO_x, and ClO_x species (involved in catalytic cycles destroying ozone), the mean observed mid-latitude ozone abundance from about 35 to 50 km is up to 50 or 60% greater than current model results. Certain observations of a 10 to 15% daytime increase in ozone concentration in the 30 to 40 km region are also puzzling, if real. We explore the model sensitivity to various input parameters and point out that, given the present uncertainties in photochemical laboratory data, no reasonable change in one or even three or four of these parameters can eliminate the ozone discrepancy. There might well be some missing chemistry in relation to the effectiveness of the loss processes for odd oxygen, or a (less likely) unknown significant O₃ source. We have to understand the present upper stratospheric ozone distribution, before estimates of possible future ozone depletion can be made with confidence. We also discuss our understanding of heavy ozone photochemistry, which might be related to a light ozone photochemical source. Fast isotopic exchange processes between O and O₂ will dominate the heavy odd oxygen chemistry, and we do not find any significant heavy ozone enhancement possibilities in the stratosphere, unless unusually large fractionation processes exist. The in situ mass spectrometer observations of a 40% enhancement in ¹⁸O³²O₂ near 30 km by K. Mauersberger remain a mystery, and further data collection -- possibly via infrared or microwave spectroscopy as well -- should be undertaken if this potentially significant discrepancy is to be understood.

Thesis Files