A model of stratospheric chemistry and transport on an isentropic surface

Abstract

First results from a new type of photochemical model are presented. The model has an Eulerian grid with latitude and longitude coordinates on a single isentropic surface, which in this work is specified at a potential temperature of 846.6 K. The photolysis rates in the model are determined using a specified distribution of the total ozone column above the surface. The model contains 107 reactions and 40 species whose concentrations are solved for explicitly with an algorithm that fully accounts for diurnal variations. Species are advected implicitly using a semi‐Lagrangian technique in which 12 hour air particle trajectories are computed and then the particles are reinitialized back onto the Eulerian grid. The photochemical model is integrated for 12 days with winds and temperatures supplied from a three‐dimensional model integration of an idealized wavenumber one disturbance. The results for the long‐lived model tracers such as N2O show excellent correlation with the potential vorticity distribution, providing confidence in the transport scheme. The distribution of the tracers and the night‐time distribution of the reservoir species show the effect of the wave disturbance in enhancing gradients at the edge of the vortex region and in low to mid‐latitudes where a tongue of low potential vorticity air is advected. The effect of the photochemistry is to produce ringed structures surrounding the vortex in reservoir species such as HNO3 and ClONO2. For the radical species, the differences in the steepness of the gradients at the nighttime terminator and the daytime terminator are clearly apparent. Steep high‐latitude gradients in NO2 (the Noxon cliff) are also seen in the results. The results demonstrate the highly complex nature of the photochemistry resulting from the movement in the dynamical vortex and its interaction with the diurnal variation near the terminator. The model provides an interesting way of viewing the full complexity of the photo‐chemistry in the presence of large‐scale advection. Calculations with zonally averaged wind and temperature fields reveal discrepancies in the calculation of the zonal mean of less than 10% for O3 and HNO3 compared with the zonal mean of the previous results. However, the discrepancies for NO2 and N2O5 were much larger, especially in high latitudes. This suggests the need to exercise caution in the interpretation of the results of some species calculated by latitude‐height models.

On the explicit versus family solution of the fully diurnal photochemical equations of the stratosphere

“…Thus these species may be parameterized in terms of a dynamical tracer to good accuracy. Austin and Holton [1990], for example, showed excellent correlation between N20 and potential vorticity in their middle stratosphere integration. To set up an appropriate parameterization, perhaps the easiest approach is to set up a tracer with monotonically changing values between North and South Poles.…”

“…The model versions described below were each integrated for 12 days for an isentropic surface in the middle stratosphere (potential temperature, 0 = 846.6 K) using the dynamical and ozone fields of Austin and Holton [ 1990]. For this situation, the initial state was zonally symmetric, with zonal winds corresponding to typical mid-winter conditions and a wave number 1 geopotential height disturbance developed during the course of the integration.…”

“…We use a two-dimensional latitude-longitude hemispheric model on a single isentropic surface as described by Austin and Holton [1990], where further details may be obtained. The model requires specification of total ozone above the isentropic surface and winds and temperatures on the surface.…”

“…In this paper, the family approach is studied in detail by integrating a hierarchy of versions of a two-dimensional (latitude-longitude) model of stratospheric chemistry and transport on an isentropic surface. The standard model, as described by Austin and Holton [1990], solves the chemical transport equations explicitly. A set of versions of the model has also been developed using the family method consisting of 26, 20, 16, 14, 12, 10, and 8 members.…”

Stratospheric Chemistry

“…Transport improvements include assimilation of stratospheric meteorological data into a 3‐D model to accurately predict tracer evolution for 2 months [ Rood et al ., 1989] and new advection algorithms that minimize numerical diffusion [ Prather et al, 1986; Shia et al, 1990]. New coupled models (chemistry, transport, and in some cases radiation) that are tested against existing data sets, mostly from satellites, have been used to understand the effects that dynamics and photochemistry have on each other [ Brasseur and DeRudder, 1987; Brasseur et al , 1988; Jackman et al ., 1989b; Schneider and Ko, 1990; Gray and Chipperfield, 1990; Austin and Holton, 1990]. Numerous other studies [ WMO, 1990] have coupled tracer transport, chemistry, and radiation in two and three dimensional models.…”