As a result of such complexities, the magnitude of trends in ozone and even the direction of the changes do not always agree, particularly when segregated by altitude and latitude (e.g., see Rusch et al., 1994; McPeters et al., 1994; Reinsel et al., f994a,b; and Miller et al., 1995). For example, Rusch et al. (1994) report that in the high-latitude upper stratosphere, SBUV data give a negative trend in O-, from 1979 to 1991, in the range -0.4 to -1.0% per year; however, the trend inferred from SAGE data is zero to slightly positive, 0 to 0.5% per year. They also showed that while data from the TOMS and SBUV instruments were in good agreement, those from the SAGE were not.
Similarly, Miller et al. (1995) compared ozone trends derived from Umkehr and balloon ozonesondes. Figure 13.10 shows that while both predict the same trend qualitatively in the lower stratosphere, they do not agree quantitatively. In addition, from ~25 to 35 km, even the direction of the changes is not in agreement. Harris et al. (1997) report that at altitudes of f 6-17 km in midlatitudes, the ozone trend measured using SAGE is as much as — 20 + 8% per decade, whereas ozonesonde measurements in the Northern Hemisphere give an average of — 7 + 3% per decade.
In addition, as seen in Fig. 13.9, changes in tropo-spheric composition can also impact the results. For example, the trend in total ozone over 20 years measured in Belgium was -1.38 + 0.50% per decade but after correcting for changes in tropospheric S02, no significant trend in ozone could be discerned (De Muer and De Backer, 1992). There is also a concern regarding the effects of trends in tropospheric aerosols on the derived trends in 03.
Because of such uncertainties, trend analysis is an active area of current research. However, it is clear that many studies have shown qualitatively a decreasing
Was this article helpful?