A global analysis of the ozone deficit in the upper stratosphere and lower mesosphere

Abstract

The global measurements of temperature, ozone, water vapor, and nitrogen dioxide acquired by the Limb Infrared Monitor of the Stratosphere (LIMS), supplemented by a precompiled distribution of chlorine monoxide, are used to test the balance between odd oxygen production and loss in the upper stratosphere and lower mesosphere. An efficient photochemical equilibrium model, whose validity is ascertained by comparison with the results from a fully time‐dependent one‐dimensional model at selected latitudes, is used in the calculations. The computed ozone abundances are systematically lower than observations for May 1–7, 1979, which suggests, contrary to the conclusions of other recent studies, a real problem in model simulations of stratospheric ozone. The “ozone deficit” at 30°N is smaller than previous analyses of LIMS data have indicated. In the stratosphere, this reduction in the deficit is due to the fact that ClO abundances for the 1979 period utilized in this study are much lower than in earlier work, mainly as a result of lower Cly concentrations. In the mesosphere, a correlation of the ozone deficit with the distribution of water vapor is indicated. The ozone deficit in the stratosphere can be eliminated by modifying only one model reaction rate: either by decreasing the rate of odd oxygen loss (e.g., by decreasing the abundances of atomic oxygen through an increase in the rate of the reaction O + O2 + M→ O3 + M within the recommended uncertainties) or by increasing the rate of odd oxygen production (by increasing the photodissociation rate of molecular oxygen primarily in the Herzberg continuum and/or invoking photolysis of vibrationally excited molecular oxygen). With the ozone abundances thus increased, a small residual deficit in the lower mesophere can be eliminated by reducing, within the recommended kinetic uncertainties, the efficiency of odd hydrogen‐catalyzed odd oxygen loss. With the adjusted model, the calculated ozone abundances for the week of January 1–7, 1979, outside of winter latitudes, also agree with the LIMS observations to within 10%.