GEOPHYSICAL RESEARCH LETTERS,VOL. 13, NO. 12, PAGES 1206-1209, NOVEMBER SUPPLEMENT 1986 OCTOBER ANTARCTIC TEMPERATURE AND TOTAL OZONE TRENDS FROM 1979-1985 Paul A. Newman Applied Research Corporation Mark R. Schoeberl NASA/Goddard Space Flight Center Abstract. The morphology of temperature trends in the south polar stratosphere from 1979-1985 is detailed. A significant cooling has occurred in the stratosphere over this period. South polar October mean temperatures have declined 18øK at 24 km over the seven year period. A large fraction of this decline occurs as a result of including 1979 temperatures in the trend. The years 1979 and 1982 differ from other years in that they include large mid- October wave events which produce significantly warmer polar temperatures. Nevertheless, a significant temperature decline has still occurred in the stratosphere even when the dynamically active years are not included. Comparing two relatively dynamically quiet years, 1980 and 1985, the largest temperature decrease in the 30 mb temperature field occurs not at the pole but in the sub-polar temperature maximumcentered near 60øS. The changes in temperature of both the maximum and the polar low regions correlate fairly well with total ozone changes observed by TOMS. Introduction Considerable attention has been focused on the secular decrease in total ozone over Antarctica during October from 1979 to 1985 reported by Farman et al. (1985) and Stolarski et al. (1986). Stolarski et al. also found that the Antarctic total ozone distribution is spatially well correlated with lower strato- spheric temperatures on a day to day basis. This result is not surprising as total ozone - lower stratosphere temperature correlations were noted almost as soon as total ozone observations were begun (Reed, 1950; for example). The total ozone-temperature linkage, however, suggests that the observed total ozone decline may be a result of (or may be contributing to) a secular decline in stratospheric temperatures. In the scenarios suggested by Mahlman and Fels (1986) and Tung et al. (1986), for example, a mechanism external to the stratosphere, such as a climatic shift in the troposphere or volcanic aerosol injection, might be acting to systematically reduce lower stratospheric temperatures, which results in the observed decrease in total ozone. In this paper, two topics will be addressed: the October temperature trend from 1979 to 1985, and the relationship of this trend to total ozone. The first topic will be assessed with seven years (1979 to 1985) of Copyright 1986 by the American Geophysical Union. Paper number 6L6321. 0094-8276 / 86 / 006L-6321 / $03. O0 Southern Hemisphere temperatures, while the second topic will be analyzed with Total Ozone Mapping Spectrophotometer (TOMS) data and discussed with respect to the lower stratospheric reverse circulation proposed by Tung et al. (1986) and Mahlman and Fels (1986). Data The temperature data analyzed here are produced by the Climate Analysis Center (CAC) at the National Meteorological Center (NMC). A detailed description of this data set can be found in Geller et al. (1983). Since the ozone reduction has primarily occurred in the lower stratosphere for which radiosonde data are available, this analysis will concentrate on that region. Radiosonde information has been adapted from monthly averaged station temperatures available in "Monthly Climate Data for the World." A comparison of NMCand radiosonde temperatures, which will be discussed in the next section, is important since NMC data are operational analyses utilizing a variety of data sources (both conventional and satellite) over the seven year span. Results The monthly averaged 30 mb radiosonde station temperatures and collocated 30 mb temperatures from the NMC gridded data are •hown in Table 1. Comparisons of these station data to NMC data indicate general agreement with the exception of 1979. In October 1979, NMC temperatures tend to underestimate the radiosonde reports. These 1979 NMC under- estimates of the radiosondes are located in the sub-polar temperature maximum region. Thus, the radiosonde determined trend is larger than the NMC determined trend. Since radiosonde stations show a downward temperature trend, and NMC data are in good agreement with these radiosondes (albeit with some underestimates), then NMC data provide a sound base for stratospheric temperature trend analyses at the levels of 30mb and lower. Figure 1 displays zonal mean temperatures for September 1979 (la), October 1979 (lb), and November 1979 (lc). The salient latitudinal features of the lower stratosphere during these months are a cold pole, warm sub-polar region (the warm ring) and a cold equatorial region. The polar lower stratospheric temperatures warm- up from a cold extreme during early spring until the sub-polar-to-pole temperature gradient reverses during October (Newman, 1986). This gradient reversal occurs on October 15, 1979, at the 30 mb level during a large wave event. Figure ld displays the temperature difference 1206