JOURNAL OF GEOPHYSICAL RESEARCH, VOL. 96, NO. D3, PAGES 5055-5071, MARCH 20, 1991 Three-Dimensional Simulations of Wintertime Ozone Variability in the Lower Stratosphere RICHARD B. ROOD, ANNE R. DOUGLASS, AND JACK A. KAYE Atmospheric Chemistry and Dynamics Branch, NASA Goddard Space Flight Center, Greenbelt, Maryland MARVIN A. GELLER AND CHI YUECHEN Institute for Terrestrial and Planetary Atmospheres,State University of New York at Stony Brook DALE J. ALLEN, EDMUND M. LARSON, ERIC R. NASH, AND J. ERIC NIELSEN Applied Research Corporation, Landover, Maryland The evolution of ozone has been calculated for the winters of 1979 and 1989 using winds derived from our stratosphericdata assimilationsystem (STRATAN). The ozone fields calculated using this technique are found to comparewell with satellite-measured fields for simulations of 2-3 months. Here we present comparisonsof model fields with both satellite and sonde measurementsto verify that stratospheric transport processes are properly represented by this modeling technique. Attention is focussedon the northern hemispheremiddle and high latitudes at the 10-hPa level and below, where transport processes are most important to the ozone distribution. First-order quantities and derived budgets from both the model and satellite data are presented. By sampling the model with a limb-viewing satellite and then Kalman filtering the "observations" of the model, it is shown that transient subplanetary-scale features that are essential to the ozone budget are missedby the satellite system. INTRODUCTION To understandthe behavior of ozone in the stratosphere, it is necessary to evaluate both the transport and photochem- ical terms of the ozone continuity equation. The transport terms can be calculated usingwind fields derived from either general circulation models (GCMs) or data. Models offer the advantage of providing a completely consistent data set for winds and temperatures, but have the disadvantage that simulations of the stratosphere are not quantitatively accu- rate. General circulation models of the middle atmosphereare plagued by a series of persistent problems [Mahlman and Umscheid, 1984; Geller, 1984]. These include the cold pole problem (i.e., simulated stratospherictemperatures at the winter pole being much colder than observations) and lack of interannual variability. In a series of experiments with the SKYHI model, Mahlman and Umscheid [1987] showed that the cold pole problem became systematicallyless severe as the horizontal resolution was increased. This was attributed to the resolution of gravity waves in the GCM. Hayashi et al. [1989] confirmed that an important part of the momentum budget in the model is in fact associated with gravity waves. They also point out that the observations by Fritts and Vincent [ 1987] indicate that gravity waves with spatial scales too small to be resolved even by the highest resolution presented by Mahlman and Umscheid [1987] play a crucial role in the middle atmosphere momentum budgets. From this modeling experience, it has been concluded that the behavior of both planetary scales and the mean flow are Copyright 1991 by the American GeophysicalUnion. Paper number 90JD02537. 0148-0227/91/90JD-02537505.00 profoundly affected by small-scaleprocesses.Proper inclu- sion of gravity wave effects is necessary for an accurate simulation of climate. Two strategies have evolved to address small-scale pro- cesses in stratospheric GCMs, both of which impact constit- uent modeling. The strategy of Mahlman and Umscheid [1987] is to resolve gravity waves explicitly (with Smagor- insky [1963] nonlinear diffusion for subscale momentum closure in the model). However, even with the 1 ø version of the SHYHI model, significant differences exist between model and observed monthly mean wind and temperature fields [Mahlman and Umscheid, 1987]. Furthermore, to calculate the behavior of reactive constituents at such high resolution is prohibitively expensive. The second strategy is to parameterize gravity wave processesand add them to coarser resolution models [e.g., Rind et al., 1988]. However, at the present time there is insufficient knowledge of gravity wave sources and gravity wave interactions with the mean flow, and therefore gravity wave parameterizations are commonly treated as adjustable parameters. Thus, while the incorporation of subscale pa- rameterizations can improve the comparison of model tem- perature and wind fields with data, it is not clear how such parameterizations affect constituent transport [see Mied and Lindemann, 1984, Figure 4]. The alternative to using model wind fields to understand constituent transport is to use data. However, there are very few direct wind data for the stratosphere,and wind estimates must generally be derived from geopotential heights. Calcu- lations of constituent variability using geostrophic winds derived from height fields are accurate only in a time- averaged qualitative sense. Due either to inaccuracies of the geostrophicapproximation or the presence of noise, calcu- 5055