To text describing this figure   To MLS home page    To MLS scientific results page

Zonal mean ozone from MLS from October 1991 through March 1994
from Froidevaux et al. J. Atmos. Sci. vol. 51, 2846-2876, 1994


time series of zonal mean ozone; there are 3 plots, one at 2hPa, 5hPa, and 10hPa
time series of zonal mean ozone; there are 3 plots, one at 22hPa, 46hPa, and 100hPa

The figures above show stratospheric zonal mean ozone mixing ratios (ppmv), based on MLS level 3AL data. Contour plots are shown as a function of latitude and time for 1 October through 14 March 1994. Panels are labeled with the surface pressure of the retrieved field (starting from top: 2, 5, 10, 22, 46, 100 hPa).

There are a number of general points which one can make, based on the above figures, mostly as a check of previous results on stratospheric ozone. For example, largest ozone mixing ratios are observed at equatorial latitudes in the mid-stratosphere (see the 10 hPa plots), and largest variability is observed at high latitudes in winter, as expected. Note that small temporal oscillations can be seen that are tied to the UARS yaw period of approximately 36 days. These are known artifacts in MLS data, and investigations continue regarding these effects, which are present at the few percent level. The lower stratospheric ozone maxima are observed during February/March in the northern hemisphere mid-to-high latitudes, with a similar peak during August/September in the southern hemisphere (see the 46 hPa plots); this spring maximum is caused by transport of high ozone air from the tropics, where the primary production occurs, with a subsequent decrease induced by increasing photochemical destruction (Perliski et al. 1989). The ozone hole-related decrease in ozone during August and September is observed at the highest southern latitudes, in the 46 hPa plots (not so much in the 100 hPa plots, interestingly); a synoptic view of early results from MLS in the 1992 winter over these regions was given by Waters et al. (1993b).

In the mid-stratosphere (see the 10 hPa plots) at mid-to high latitudes, the annual cycle dominates, as a consequence of significant photochemical production which maximizes in the summer. This annual variation is observed, with the expected 6-month shift between hemispheres. A well-known semi-annual oscillation (SAO) dominates at low latitudes in the mid-to upper stratosphere, as observed in these data as well (see also Eluszkiewicz et al. 1994). The SAO and associated vertical motions are believed to play a role in producing features like the double-peak structures in pressure/latitude cross-sections of SAMS N2O and CH4 fields (Gray and Pyle, 1987; Choi and Holton, 1991). In the upper stratosphere, temperature-dependent ozone destruction cycles play a dominant role, and the 2 hPa MLS plots appear to follow such trends (e.g.,spring mid-latitude decreases when temperatures are increasing). These and other features related to the annual and semi-annual ozone variations are in general agreement with previous analyses by Perliski and London (1989) and Perliski et al. (1989). Further correlative studies would be useful, however, for quantitative conclusions on these variations. Ray et al. (1994) provide a more detailed analysis of the SAO, based on MLS data, and comment on the amplitude characteristics versus pressure/latitude.

An interesting tropical feature is the existence of low ozone abundances at 46 hPa from Oct. 1991 to mid-1992. The low equatorial values rise sharply during June-July 1992 and level off until the Sep. 1993. From late 1993 on, low values are again observed near the equator. Fig. 2 further shows that this low ozone tropical feature splits into two somewhat weaker sub-tropical lows after the summer of 1992. The low ozone values may be linked to upward motion in connection with the quasi-biennial oscillation (QBO). Indeed, the sub-tropical lows occur during a period of equatorial westerlies (near 20 hPa), a period traditionally associated with enhanced downward motion at the equator, with upwelling in the sub-tropics as a result of the return arms of the induced circulation (e.g. Gray and Pyle, 1989). Observations of tropical H2O from MLS also show variations possibly associated with the QBO in the lower stratosphere (Carr et al. 1994). Observations of the sub-tropical ozone QBO, as deduced from TOMS data by Bowman (1989), can place constraints on details of the mechanism for the QBO and the spread of related anomalies to other latitudes. We simply note here that the sub-tropical low MLS ozone values are reasonably symmetric about the equator - Bowman (1989) and Lait et al. (1989) point out that the TOMS analyses show a more symmetric QBO behavior about the equator than previous analyses of ground-based or Nimbus-4 BUV data. However, uplift effects and circulation changes arising from post-Pinatubo aerosol heating would also have to be considered as an explanation for low tropical ozone (Kinne et al. 1992, Schoeberl et al. 1993, Grant et al. 1992, 1994, Pitari 1993).

For further discussing of these data and analyses see Froidevaux et al. J. Atmos. Sci. vol. 51, 2846-2876, 1994.

Back to top of this figure   To MLS home page    To MLS scientific results page