2.1 Ozone Observations

As stated above, the discovery in the mid-1980s of the dramatic, seasonal drop in ozone over Antarctica during the spring took atmospheric scientists by surprise. Previously, they had thought that they understood rather well the physical and photochemical processes controlling ozone production and loss. As late as 1983, work by Rolando Garcia and Susan Soloman produced computer model simulations of ozone production, transport, and loss processes that closely resembled the observed data. Two years later, in 1985, Joesph Farman and his colleagues discovered that measured amounts of Antarctic total ozone fell a stunning 50% in early spring between 1975 and 1984. The large losses were primarily confined to the spring season (September-October). Farman's measurements came from ground-based Dobson spectrophotometer data at Halley Bay (76°S, 27°W). Figure 11.01 displays a plot of these same Halley Bay observations which have been extended from the original 1957-1984 to include 1985 to 1995.

Farman and his research colleagues found that spring total ozone values had decreased from 300+ Dobson Units (DU) in the late 1950s and early 1960s to around 200 DU in the early 1980s. The original Farman et al. data only included points from 1957 to 1984, and did not include any satellite data. In was immediately clear that a large depletion of ozone was occurring over Halley Bay, Antarctica, reaching lowest values during the early October period. Farman suggested that these large depletions were a result of the uniquely cold temperatures over Antarctica combined with the increasing burden of chlorine in the stratosphere. While the suggestion that the ozone hole results from the cold temperatures and the increasing chlorine was eventually shown to be correct, Farman's proposed mechanism was not. Nevertheless, this seminal work opened up serious questions about the characteristics and physics of these Antarctic losses, radically altering our perception of the capability of models to represent ozone photochemical processes.

Subsequent analysis of TOMS satellite data by Richard Stolarski and his colleagues in 1986 showed that the depletion of ozone during southern hemisphere spring occurred over much of Antarctica, centered over the South Pole. Thus, the TOMS data confirmed Farman's observations. It is important to research scientists to compare satellite measurements of total ozone with ground measurements of total ozone. In this way, the satellite instrument is "ground truthed" and gives the research scientist more confidence in the reliability of the satellite instrument.

Figure 11.02 displays false color images of the Antarctic ozone hole derived from the BUV instrument (1970, 1971, and 1972), the Nimbus-7 TOMS (1979, and 1992), the Meteor-3 TOMS (1993, and 1994), and the Earth Probe TOMS in 1996 (the minimum values of the October averages from these satellite instruments are included in Figure 11.01, showing excellent agreement with the Halley Bay ground based measurements). The "ozone hole" is the large region of reduced ozone over Antarctica seen in the 1992-1995 period in comparison to the values observed in earlier years. The spring low of total ozone over Antarctica in the southern spring prior to 1980 (seen in the top four images) is a natural phenomenon. It was first noted in the by Dobson in 1966. This ozone minimum results from normal winter Antarctic circulation patterns, and is not what is referred to as the ozone "hole."

The largest ozone loss is observed over Antarctica, but ther loss of ozone at subpolar latitudes is also substantial. Figure 11.03 displays the average of the October 1970, 1971, 1972, 1979 (top left), and the average of October 1992, 1993, 1994, 1995 (top right). The percent difference of these two October averages is shown in the bottom panel of Figure 11.03.

As is clear from the figure, largest ozone losses are confined to the Antarctic region. However, ozone losses are also seen in region of the high ozone collar towards the midlatitudes. This "subpolar" ozone loss was first noted by Schoeberl et al. in 1986 using TOMS data.

Figure 11.04 displays a longitudinally averaged (commonly referred to as a zonal mean) plot of each year of October data displayed in Figure 11.02. Note the precipitous drop in total ozone amount from over 300 DU between 60° and 90°S in the 1970s to total ozone values under 200 DU in the mid 1990s. Note also the decrease and equatorward shift of the subpolar ozone maximum between the early 1970s and the mid 1990s. It is now recognized that almost all of the ozone loss is generally confined to the Antarctic polar vortex.

It is important to note that even without the ozone hole, there is a natural minimum in the total ozone field over Antarctica duringthe southern hemisphere spring. Even in the early 1970s, ozone amounts were low over Antarctica in the month of October, as you can see in the top panels of Figure 11.02 illustrating pre-ozone-hole years.

Figure 11.05 is adapted from Dobson's 1966 work. It shows the difference between the northern polar and southern polar values. Dobson showed that Antarctic ozone amounts were anomalously low in comparison to springtime polar measurements in the northern hemisphere. This low is a natural phenomena during the southern spring. It arises due to differences in atmospheric circulation between the northern and southern polar latitudes.

The black line in Figure 11.05 displays Arctic latitude total ozone amounts observed in the 1960s, while the blue line illustrates Antarctic latitude total ozone amounts during 1964 and 1965. The much smaller values in the southern hemisphere are a result of the circulation differences between the northern and southern hemispheres. Superimposed are 1994 observations at Halley Bay. These observations show that springtime total ozone levels over Antarctica are significantly lower in 1994 than during the mid-1960s. This is consistent with the data shown in Fig. 11.01. During the 1960s, ozone amounts were rather constant around 300 DU between May and October over Antarctica. During the 1980s and into the early 1990s, ozone amounts dramatically decreased in the August to November, with the minimum occurring in October each year.

Other ground based observations have shown the presence of the ozone hole. These include measurements from Syowa station in Antarctica (Chubachi, 1984). Figure 11.06 displays ozone values measured by a Dobson spectrophotometer at Syowa station from February 1982 through January 1983, adapted from Chubachi (1984). These observations show very low amounts during the October period, consistent with the Halley Bay data in 1994 shown in Fig. 11.05. The ozone decrease occurred during August to October. Ozone values dramatically recovered on October 28, 1982 as the ozone hole moved away from Syowa station. (The hole is "mobile" in the sense that the ozone-depleted polar vortex region is moved around by the dynamics of motion in the stratosphere.)

Figure 11.07 shows September-October 1982 false color images of the Nimbus-7 TOMS total ozone fields for individual days during the September through October period with Syowa's location denoted by the white dot on the edge of Antarctica. As is clear from these images, ozone decreased over Antarctica during September, with Syowa well inside the hole (i.e. inside the poalr vortex region). Fluctuations in ozone occur at Syowa as high ozone amounts in the warmer midlatitudes are swept back and forth across the Antarctic coast in response to the weather systems and associated jet stream undulations (see Chapter 2). Only at the end of the winter season (in what is called the "final warming") do much higher ozone amounts from the mid-latitdues move irreversibly southward over the continent. Figure 11.06 shows this "final warming" as ozone readings increase in November and December.

In both figures 11.06 and 11.07, ozone decreased over Antarctica during September, 1982 as the ozone hole developed. Thus, the 1984 Chubachi paper was an important contribution, since it showed that the Halley Bay observations were consistent with the Syowa observations. Evidence of the ozone hole comes from both locations. However, Chubachi did not point out that these October 1982 measured amounts were a decrease over October measured amounts from previous years, thus leaving out the critical "trend" element. Without it, one would conclude that ozone amounts fall dramatically every October!

The Chubachi paper also showed the first vertical ozone profiles of the ozone hole, as shown in Figure 11.08. These profiles showed an ozone decrease occurring between 15-24 km altiitude region (i.e. in the lower stratosphere) during the August-September, 1982 period. Figure 11.08 covers the period February 1982 through January 1983, The altitude range is from the surface to about 32 km (20 miles). Balloon flight dates are indicated by the black arrows at the bottom of the figure. Ozone measurements are in partial pressure . This refers to the air pressure due to ozone alone, so that higher ozone density is reflected in increased partial pressure. The period of decreasing ozone is easily seen during the August-September 1982 period (bracketed by the white lines).

2.2 Theories

Having established the existence of a seasonal ozone hole in the satellite and ground-based record over Antarctica, and that this hole did not exist as recently as the early 1970s, Farman et al. attempted to explain why the hole should suddenly have developed. What could have changed in the atmosphere over Antarctica to produce such a large, sudden loss of ozone on a seasonal basis?

Farman and his colleagues in the same 1985 paper proposed a possible mechanism for the sudden loss in ozone each October. They began by pointing out that there was: (1) no apparent shift in meteorological parameters (e.g., temperature and winds); (2) weak transport effects at the altitude of the hole; (3) an apparent increase in halocarbon (see Glossary) amounts since the 1960s; and (4) extremely cold temperature in the lower stratosphere above Antarctica. They postulated that increasing chlorine concentrations with the very cold temperatures over Antarctica were enhancing ozone loss.

The ozone loss rates computed by McElroy et al. (1986) and Solomon et al. (1986) using Farman's mechanism were much too small to explain the large ozone losses seen during September. Their computer models showed that amounts of free oxgyen atoms (necessary for the catalytic destruction of ozone) were too low in the 15-24 km altitudes range where most of the ozone loss was taking place. Hence, the Farman theory was found to be incorrect.

Soon thereafter, three theories emerged to explain the Antarctic ozone hole. As explained succinctly in Figure 11.09, these included the dynamical theory, the nitrogen oxide theory, and the heterogeneous chemistry theory. Additional research lead scientists in the field to conclude that the heterogeneous chemistry theory was correct (most consistent with observations). The dynamics and nitrogen oxide theories were found to be incorrect.

2.2.1 Dynamical Theory -- The dynamical theory of the ozone hole proposed that the Antarctic circulation had changed. It had long been recognized that the dominant circulation of the lower stratosphere in winter involves the poleward and downward motion of ozone-rich air from the middle and upper stratosphere of the tropics, where ozone is photochemically created all year long. This is the Brewer-Dobson circulation (see Chapter 6, section 3.0). Recall from the discussion of the Brewer-Dobson circulation in Chapter 6 that this poleward and downward motion of ozone-rich air leads to a build up of ozone in the mid- to high latitudes during the winter. The dynamical theory proposed that this normal pattern was changing, and that ozone-poor air from the troposphere was being transported into the lower stratosphere, instead of the ozone-rich air (Tung, 1986).

If ozone-poor air from the troposphere was indeed being transported into the lower stratosphere, then other long-lived trace gases should also be measurably increasing in the lower stratosphere. An example of such a tracer is nitrous oxide (N2O), which is produced in the troposphere by biological processes, and is destroyed in the stratosphere by either ultraviolet (UV) radiation photolysis or by a reaction with excited O atoms (WMO, 1985). The loss of N2O takes place in the upper stratosphere, since O atoms are generally produced by the photolysis of O2, which requires UV wavelengths under 240 nm (see Chapter 1, Section 3.0). Such energetic UV radiation cannot penetrate into the troposphere because of the screening by ozone molecules. Hence, N2O has fairly high amounts in the troposphere (between 300-310 ppbv) and low amounts in the upper stratosphere. This general profile of N2O has been confirmed by satellite, balloon, and aircraft observations.

The dynamical theory predicts that Antarctic N2O amounts should be high if the air was transported upward from the troposphere into the lower stratosphere where ozone was low. Figure 11.10 displays a plot of N2O measured during the Airborne Antarctic Ozone Experiment (AAOE) on September 9, 1987, over Antarctica.

The observations show that N2O is substantially lower than amounts characteristic of the troposphere (300-310 ppbv) in the region inside the polar vortex where the Antarctic ozone hole appears in the spring. These N2O observations ( as well as observations of other long-lived trace gases) demonstrate that air inside the lower stratospheric Antarctic polar vortex had indeed descended from the middle and upper stratosphere, otherwise N2O amounts would've been much higher. Furthermore, we can conclude that the air ought to contain higher ozone amounts, since we know that the air was brought down by the Brewer-Dobson circulation from higher altitudes and lower latitudes (e.g., see Loewenstein, et al., 1989). The dynamical theory for ozone loss is thus incorrect.

2.2.2 Nitrogen Oxide Theory -- The nitrogen oxide theory of the ozone hole was proposed by Callis and Natarajan (1986), and it proposed that large amounts of NOx compounds were being produced as a result of the peak in sunspot activity in 1979. Sunspots follow an 11-year cycle and the peak period is called a solar maximum. This NOx would be photochemically produced as a result of increased energetic UV light in the mid- to upper stratosphere of the tropics and transported into the polar lower stratosphere by the Brewer-Dobson circulation. The loss process would occur catalytically as

NO+O3 --> NO2 +O2

NO2 +O --> NO+O2


O3 +O --> 2 O2

Measurements of NO2 (nitrogen dioxide) at high latitudes indicate very low concentrations during the spring period of the ozone hole (e.g. Farmer et al., 1986; Coffey and Mankin, 1989).

Figure 11.11 displays column abundances of NO and NO2 measured from the NASA DC-8 on September 21, 1987 (Toon et al., 1989), during the same AAOE field research campaign. The data show that NOx decreases between the midlatitudes and the interior of the polar vortex. This is the opposite of what the nitrogen oxide theory predicts. In addition, the satellite observations of total ozone reveal no solar cycle variation in the depth of the ozone hole, as would be expected if NOx amounts were varying with the solar cycle (McElroy et al. (1986) and Solomon et al. (1986)).

2.2.3 Heterogeneous Chemistry Theory -- The third theory of the ozone hole involves heterogeneous chemical reactions on the surfaces of solid (frozen) particles that formed in the cold lower stratosphere of the Antarctic vortex. These reactions would free chlorine from "inactive" forms into "reactive" forms, where the chlorine could destroy ozone in the catalytic cycles discussed in Chapter 5. Originally proposed by Crutzen et al. (1986), McElroy et al. (1986), and Solomon et al. (1986), this theory proposed that reactions which normally do not occur in gas phase might be greatly enhanced if chlorine-containing compounds such as ClONO2 (chlorine nitrate) and HCl (hydrochloric acid) could collect on the surfaces of these particles and then react to release the chlorine into a reactive form that could cause large ozone losses. This is called the heterogenous chemistry theory , and it turned out to be the correct one for explaining Antarctic ozone losses. It is discussed in more detail in Section 5.0 "Why the Antarctic Ozone Hole Exists: Ozone Hole Theory" of this current chapter.