5 -- WHY THE ANTARCTIC OZONE HOLE EXISTS: OZONE HOLE THEORY

The cause of the ozone hole became a topic of intense scientific research interest in the 1980s. An alphabet soup of studies together with accompanying instrumentation were launched beginning in 1986 and continuing through the present. These include ground based field campaigns like NOZE I and II (National Ozone Expedition I and II); aircraft missions such as AAOE (Airborne Antarctic Ozone Experiment) and ASHOE/MAESA (Airborne Southern Hemisphere Ozone Experiment/Measurements for Assessing the Effects of Stratospheric Aircraft); and satellite studies such as TOMS (Total Ozone Mapping Spectrometer), SAGE I and II (Stratospheric Gas and Aerosol Experiment I and II), and the UARS (Upper Atmospheric Research Satellite) mission, which included an array of remote sensing devices designed to measure trace gases in the atmosphere.

Comprehensive findings from these many research efforts are found in the World Meteorological Organization (WMO) 1991 and 1994 global ozone assessments. These studies have shown that stratospheric ozone loss is caused by chlorine and bromine catalytic reaction (as detailed in Chapter 5). The source of stratospheric chlorine ultimately comes from photochemical breakdown of manmade chlorofluorocarbons released in the troposphere from human activities. Subsequent observations from aircraft and satellites, together with laboratory and modeling studies, have given better understanding to details of chemical ozone losses inside the ozone hole, as discussed in the 1994 WMO assessment.

In this section, we will review how heterogeneous processes lead to stratospheric ozone loss. In the first part, we will discuss the morphology of polar stratospheric clouds, and briefly touch the properties of these clouds. In the second part, we will look at the heterogeneous reactions themselves that release chlorine into active forms. The third part will look at the direct loss process by which ozone is destroyed. The fourth and final part will pull all of these threads together to illustrate the behavior of trace gases including ozone over the entire winter-spring period.

5.1 Polar Stratospheric Clouds

Polar stratospheric clouds (PSCs) are clouds that form in the extremely cold and dry conditions that exist in the polar night regions of the stratosphere. PSCs are fundamental to an understanding of the Antarctic ozone hole. We've already discussed a little about PSCs and their role in ozone loss above in Section 3 and also in Chapters 1 and 5. PSCs are important because they convert chlorine species that do not destroy ozone (like hydrochloric acid, HCl, and chlorine nitrate, ClONO2) into forms that can destroy ozone. These reactions take place on the solid surfaces of the PSC cloud particles. Observations of these clouds have a long and detailed history. These clouds have been visually observed at high latitudes (e.g. Scandinavia) for centuries. There are two types of PSCs, referred to as Type I and Type II PSC. The properties of each are listed in Figure 11.42, which also displays a picture of a Type II PSC in the northern hemisphere taken in January 1989.

5.1.1 Comparing Type I and Type II PSCs -- While the composition of the Type II PSC is known to be water ice, the composition of the Type I PSC is rather poorly understood. Originally, Crutzen and Arnold (1986) and Toon et al. (1986) proposed that Type I PSCs were formed as nitric acid trihydrates (abbreviated NAT). The equilibrium temperature of NAT (i.e. temperature at which the surrounding air is just saturated with respect to NAT ice crystals) was thought to be consistent with the formation temperatures of Type I PSC (Poole & McCormick, 1988). Recent aircraft observations (Dye et al., 1990; Del Negro et al., 1997) using extremely precise temperature measurements have shown that the formation temperature of Type I PSC is inconsistent with the NAT equilibrium temperature. Current theories suggest that some Type I PSC cloud particles are composed of mixtures of nitric acid (HNO3), water vapor (H2O), and sulfuric acid (H2SO4). These are known as liquid ternary solutions. Droplets grow as the temperature decreases, achieving sizes providing sufficient surface areas to enable heterogeneous chemical reactions.

5.1.2 Denitrification and Dehydration of the Stratosphere -- While the details of growth, composition, and size of Type I PSCs is a subject of intense research, the presence of such particles is easily detected by both aircraft and satellite observations. Because of the small size of Type I PSCs (~1 micron), they have very small sedimentation velocities (~10 meters per day). (Sedimentation refers to the settling out of small particles in the atmosphere.) In the case of Type II PSCs, the particles are quite large (greater than 10 microns), and can settle out of the stratosphere relatively rapidly (~1.5 km/day). This sedimentation transports reactive nitrogen and water that are on the particles out of the lower stratosphere. The removal of reactive nitrogen is known as denitrification and the removal of water vapor is known as dehydration. These denitrification and dehydration processes occur inside the Antarctic polar vortex over the course of the very cold winter. Because of this nitrogen loss, there is no NO2 to react with ClO to form the reservoir ClONO2, hence denitrification results in a maintenance of high levels of active chlorine that can continuously destroy ozone.

5.1.3 PSC Formation: Hemispheric Differences in Temperature -- The presence of PSCs is fundamentally different for the northern and southern hemispheres. Because of the extremely cold temperatures in the southern hemisphere, PSCs form early in the winter, and persist late into the spring. Temperatures fall below the critical threshold for PSC formation (known as the frost point) through large volumes of the Antarctic stratosphere and for long periods of time. The frost point temperature for PSC formation is 195 K or -78°C for Type I and 188 K or -85°C for Type II. As a result, the southern polar vortex is both dehydrated and denitrified by late winter. This is not usually the case in the northern hemisphere, where temperatures are typically much warmer.

The threshold formation and growth of PSC particles have been observed by balloon-borne measurements over Antarctica (Hofmann et al., 1989; Hofmann and Deshler, 1989), and by aircraft measurements (Dye et al., 1990; Del Negro et al., 1997). Satellite observations from the POAM II instrument shown in Figure 11.43 provide evidence of PSCs inside the Antarctic polar vortex in winter, primarily in the intensely cold lower stratosphere during the depth of winter. The data shown is for the southern hemisphere winter season (May through November) for 1994-1996.

Note the correspondence between the PSC probability in Fig. 11.43 and the polar temperature evolution in Figure 11.17. Note also the much higher frequency of PSCs in the southern hemisphere than in the northern hemisphere, due to the colder southern hemisphere temperatures. While there are usually about 2.5 months in the southern hemisphere winter during which time temperatures are below the frost point required for Type II PSC formation, the northern hemisphere typically experiences only a few days with temperatures below this frost point. Aircraft observations of particle surface area made just inside the polar vortex on 28 July 1994 by the NCAR MASP instrument show the relationship of aerosol particle size (volume) to temperature in Figure 11.44 (Del Negro et al., 1997; Kawa et al., 1997).

The temperatures required for formation of the two types of PSCs are quite low, 195K (- 78°C) for Type I and 188K (-85°C) for Type II. In Figure 11.45 we have a cross-sectional plot of northern hemisphere and southern hemisphere wintertime temperature at three different altitudes, 100 hPa, 50 hPa, and 30 hPa inside the polar vortex region. There are two lines on each plot. The top one corresponds to the temperature required for Type I PSC formation and the bottom to the temperature required for Type II PSC formation.

As we see in this figure, temperatures fall below the Type I PSC frost point in May, and then fall below the Type II PSC frost point in mid June. The time period over which temperatures are below the frost point can extend to as much as four months in the southern hemisphere, whereas the frost point time period in the northern hemisphere is only a few days. Typically, air passes through these very cold temperatures such that nearly the entire volume of air inside the polar vortex has been exposed to polar stratospheric clouds within a couple of weeks of the first appearance of PSCs. Since temperatures are so cool in the southern hemisphere, the period over which large particles can form is long, while this period in the northern hemisphere is quite short.

5.2 Heterogeneous Reactions

Heterogeneous chemistry is fundamental to an understanding of the Antarctic ozone hole. Recall our discussion of heterogeneous chemistry theory from Sections 1.2 and 2.2.3. This theory for Antarctic ozone loss proposed that reactions were occurring on the surfaces of tiny cloud particles that form in the extremely cold conditions of the Antarctic winter stratosphere. It turned out to be the correct theory. We've already discussed the sorts of reactions associated with both stratospheric ozone photochemical production and loss and the catalytic reactions associated with the ozone hole phenomenon (see Chapter 5).

5.2.1 Heterogeneous Versus Homogeneous Chemical Reactions -- These heterogeneous chemical processes are extremely important because they convert reservoir species such as HCl (hydrochloric acid) and ClONO2 into active chlorine species such as ClO that can destroy ozone. By heterogeneous, we mean that the chemical reactions occur on the solid surfaces of particles, principally PSCs. Heterogeneous proccesses are different from homogeneous processes. Homogeneous processes involve chemical reactions in the gas phase only.

5.2.2 Principal Heterogeneous Reactions -- The principal heterogeneous reactions under discussion are listed below in (A) through (E).

Also produced is HNO3 (nitric acid), which is retained on the PSC cloud particle surfaces. Meanwhile, these heterogeneous reactions simultaneously act to remove nitrogen from gas phase compounds such as N2O5 (dinitrogen pentoxide) and ClONO2, and sequester (lock up) reactive nitrogen as HNO3, a reservoir species for nitrogen. This process of sequestering nitrogen is known as denoxification (as opposed to dentrification which carries reactive nitrogen out of the stratosphere). Without reactive nitrogen compounds to act as a brake on reactive chlorine compounds, the chlorine compounds are free to destroy ozone. HNO3 must be photolyzed by sunlight to form reactive nitrogen compounds. There is no sunlight during in Antarctic during the wintertime polar night. The lifetime of HNO3 therefore becomes very long. By mid-October, the photolysis time scale for HNO3 in the lower stratosphere is over a month, making HNO3 a relatively inert trace gas during the ozone hole period. Furthermore, since PSC particles slowly settle, the particle denitrification process will carry HNO3 to lower altitudes. In short, the reactive chlorine compounds are able to destroy ozone without reacting with reactive nitrogen, which is locked up in HNO3 form.

Reaction A of the above heteogeneous processes is illustrated in Figure 11.46 where HCl and ClONO2 react to form Cl2 and HNO3.

5.2.4 In Situ Measurements of Reactive Chlorine Inside the Polar Vortex -- Proof that such heterogeneous reactions are occurring in the Antarctic stratosphere in winter come from aircraft measurements into southern hemisphere polar vortex. Measurements of the ratios of ClOx/Cly and also HCl/Cly reveal the extent of reactive chlorine in the Antarctic stratosphere. Figure 11.47, adapted from the work of Kawa et al, (1997), displays an example of the effect of heterogeneous reactions that free reactive chlorine. The plot shows ClOx/Cly and HCl/Cly ratios measured on July 28, 1994 by the NASA ER-2 aircraft flight into the southern hemisphere polar vortex.

In Figure 11.47, we see how much total reactive chlorine in the form of ClOx is present in the polar vortex. Note the large concentrations of (reactive) ClO and the low concentrations of (non-reactive) HCl at temperatures below 195 K (the frost point for PSC Type I clouds). Kawa et al. (1997) and Del Negro et al. (1997) concluded that the concentrations of reactive chlorine were consistent with heterogeneous reactions on liquid ternary solutions, such as are believed to make up Type I PSC cloud particles. These aircraft observations conclusively demonstrate the relationship of cold temperatures and PSC heterogeneous chemical effects, well before the onset of large polar loss processes.

5.3 Catalytic Loss of Ozone

As we saw in Chapter 5, the principal loss processes for ozone during the ozone hole period involve ClO (chlorine monoxide) and BrO (bromine monoxide). We will examine the catalytic reactions associated with both compounds.

5.3.1 Catalytic Loss of Ozone: ClO-ClO Reaction -- ClO is formed via the photolysis of Cl2 (molecular chlorine) as the sun rises over Antarctica. Cl2, a byproduct of heterogeneous reactions A and B shown above in section 5.2.2, has a large cross-section at visible wavelengths (280-420 nm). That is, only weak sunlight is required to break up Cl2 into two free chlorine atoms. As a result, Cl2 is almost immediately broken up as the sun rises over Antarctica. The two free Cl atoms are then free to engage in catalytic reactions, where each Cl destroys an ozone molecule, and then reforms itself. The ClO-ClO reaction ( see the work of Molina and Molina, 1987) is

The quantity hf represents a photon of solar ultraviolet radiation (where f is the frequency) and M represents a third-body molecule. This reaction is illustrated schematically in Figure 11.48.

There are two key parts to this reaction worth noting: (1) the reaction is limited by the amount of ClO, and proceeds as the square of the ClO concentration; and (2) the reaction does not require free oxygen atoms (O) in order to destroy ozone, as do most conventional catalytic loss processes (see Chapter 8, Section 4).

Mixing ratios of ClO in the Antarctic vortex have been measured from aircraft (Anderson et al., 1989), the ground (Solomon et al., 1987), and from satellite (Waters et al., 1993). The in-situ aircraft observations show ClO concentrations in excess of 1 ppbv (parts per billion by volume), an amount that accounts for a substantial fraction of the inorganic chlorine found in the stratosphere. Figure 11.49 displays the observations of ClO from two aircraft flights during the Antarctic Airborne Ozone Experiment (AAOE) mission. The first panel shows mixing ratios for ozone (in red) and ClO (in yellow) for 23 August 1987 and the second panel shows the same two ratios for 16 September 1987. This corresponds to the time of the initial onset of ozone depletion and to the time of peak ozone loss.

Comparing the two dates, Figure 11.49 shows that ozone mixing ratios fall as ClO concentrations rise. On 16 September 1987, during the peak ozone loss period, ClO concentrations are in excess of 1 ppbv at 70°S. Anderson et al. (1989) calculated that the ClO-ClO reaction was responsible for approximately 40% of the ozone loss during the 1987 ozone hole period. Similar calculations by Salawitch et al. (1993) show that in the Arctic polar vortex (see section 6.0 below), about 50% of the January 1992 northern polar ozone losses resulted from this same reaction.

5.3.2 Catalytic Loss of Ozone: BrO-ClO Reaction -- A second reaction of nearly comparable effect to the ClO-ClO reaction is the BrO-ClO reaction (McElroy et al., 1986). This catalytic reaction consists of the following set of reactions.

Again, as with the ClO-ClO reaction, this BrO-ClO reaction does not require free oxygen atoms. Anderson et al. (1989) calculated that this reaction accounts for a loss rate that is about half that of the ClO-ClO reaction. Similar ozone loss rates are calculated by Salawitch et al. (1993) for the northern polar winter of 1992. While BrO concentrations are much smaller than ClO concentrations (nearly two orders of magnitude less or about 1/100th, the BrO-ClO reaction rate is very large, yielding ozone loss rates only somewhat smaller (about 1/2).

5.3.3 Catalytic Loss of Ozone: ClO-O Reaction -- A third loss process is the conventional catalytic loss process of ClO with free oxygen atoms via the following two reactions:

Calculations by Salawitch et al. (1993) indicate that this process accounts for less than 15% of the observed ozone loss. Anderson et al. (1989) indicates that the ClO-O reaction is much less important that the ClO-ClO and ClO-BrO reactions, accounting for only 3% of the observed ozone loss.

5.4 Evolution of the Ozone Hole Through a Season

The ozone hole phenomenon follows a regular pattern each year. This temporal evolution of the hole from onset to end is described in this section.

5.4.1 Chlorine Activation and Nitrogen Denoxification -- The first set in the formation of the ozone hole occurs during the Antarctic polar night. The extremely cold temperatures that are found there (due to the near-total isolation of air inside the Antarctic polar vortex) allow for the formation of polar stratospheric clouds (PSCs). Heterogeneous chemistry on the surface of the PSCs transforms inactive chlorine compounds into active forms. These compounds initially exist as inactive, so-called reservoir species, such as chlorine nitrate, ClONO2, and hydrochloric acid, HCl. By way of a heterogeneous reactions on the surface of PSCs, these inactive chlorine species are converted into active (i.e. reactive) chlorine species (e.g. molecular chlorine, Cl2, chlorine monoxide, ClO, and the chlorine monoxide dimer, Cl2O2). These reactions occur during the polar night. By middle winter, most of the chlorine inside the southern lower stratospheric vortex is in the form of Cl2.

These same reactions also remove nitrogen from the stratosphere in a process known as denoxification. Reactive NOx compounds are sequestered into nonreactive forms, specifically, nitric acid, HNO3.

The heterogeneous reactions are again listed

The key point here is the conversion of relatively benign reservoir species ClONO2 and HCl into Cl2, which can then be photolyzed by visible and ultraviolet wavelengths. Also formed is HClO, another reactive chlorine compound that can be photolyzed by ultraviolet wavelengths. These processes, which determine the location and timing of the ozone hole, are illustrated schematically in Figure 11.50, which is adapted from Webster et al., (1993).

5.4.2 Return of Sunlight and More Heterogeneous Chemistry -- The next step occurs when sunlight first returns to the Antarctic stratosphere in early spring. Figure 11.51 illustrates the return of sunlight which photolyzes hypochlorous acid, HOCl, and marks the end of cold temperatures.

Figure 11.52 displays the chemical evolution of an air parcel just inside the Antarctic polar vortex from 17 August 1992 to 17 September 1992. This evolution is computed from a chemical trajectory model (adapted from Schoeberl et al., 1996).

Over this 30 day period, the parcel loses approximately 30% of its ozone (top left panel). The parcel initially encounters temperatures below the nitric acid trihydrate (NAT) equilibrium temperature (top right panel), leading to the formation of PSCs. Heterogeneous reaction (A) occurs, causing a rapid decrease in HCl (middle left panel) as the reaction converts HCl into Clx (ClO + 2*Cl2O2 , bottom left panel). HNO3 (bottom right panel ) is also formed in this reaction, but it is photolyzed into OH and NO2, and the NO2 reacts with ClO to reform ClONO2. Heterogeneous reaction (C) between ClONO2 and H2O is initiated on day 244 as the temperature falls below the frost point. This heterogeneous reaction converts all of the ClONO2 into nitric acid, HNO3, and HClO.

Differences exist in how an air parcel evolves if it is near the edge of the polar vortex or deep within the vortex. For parcels deep within the vortex: (1) most of the HNO3 has been removed through denitrification as large PSCs settle out of the stratosphere, carrying nitrogen with them; (2) most of the chlorine is in the form of Clx , since there is no HNO3 to photolyze to reform ClONO2; and (3) ozone losses are larger, since Clx is quite large (more than 1ppbv).

5.4.3 UARS Observations of Antarctic Vortex Trace Gas Concentrations, September 17, 1992 -- The evolution of chemical constituents in the ozone hole is illustrated using the global observations of the UARS satellite. As discussed in Section 3 on the polar vortex, downward transport of air from the upper and middle stratosphere will tend to increase values of O3, HNO3, and ClONO2 inside the polar vortex. Figure 11.40 showed how ozone concentrations in the southern hemisphere actually increase in the fall and early winter (April-June). This is followed by a decrease that begins in late winter (August) and becomes quite dramatic by early spring (late September-early October).

Figure 11.53 shows simultaneous UARS satellite observations of ClO, O3, HNO3, and ClONO2 on 17 September 1992. The ClO and O3 concentrations were measured with the Microwave Limb Sounder (MLS) instrument while the HNO3 and ClONO2 concentrations were measured with the Cryogen Limb Array Etalon Spectrometer (CLAES) instrument, both of which are carried onboard the UARS. These show strongly enhanced levels of ClO inside the southern hemisphere polar vortex with concomitant low levels of ClONO2. The spatial patterns of ClO are also clearly mirrored in the low ozone values. The Antarctic polar vortex is also strongly denitrified, with virtually no HNO3 (Douglass et al., 1995).

5.4.4 UARS Observations of Trace Gases Inside the Antarctic Vortex for Selected Days in 1992 -- The next three figures are based on measurements of trace gases inside the Antarctic polar vortex made by the MLS and CLAES instruments aboard the UARS satellite. The observations are for eight selected days in 1992. These days are selected from each season, and thus give an idea of what is occurring throughout the course of the year. The trace gases include chlorine monoxide, chlorine nitrate, and nitric acid.

(i) ClO Observations -- Figure 11.54 presents UARS-MLS observations of chlorine monoxide, ClO, for eight selected days in 1992 in the southern hemisphere, including the Antarctic polar vortex region. Measurements are made on the 500 K isentropic surface, corresponding to an altitude of about 20 km in the Antarctic stratosphere. The figure shows that peak amounts of ClO occur in the August-September (late winter-early spring) period during the time of rapid ozone loss. ClO amounts decrease to background levels by early November. Midwinter values of ClO do not achieve the high amounts found in since most of the chlorine is found in species such as Cl2 , HCl, and Cl2O2. As discussed previously, the source of the ClO is from the reservoir species such as HCl and ClONO2. The enhanced amounts of ClO shown in Figure 11.54 indicate that chlorine was being released from its reservoir species on those days in 1992 as a result of heterogeneous reactions on the surfaces of PSCs. This led to the development of an ozone hole as spring sunlight photolyzed the resulting Cl2 and allowed the ClO catalytic process of ozone loss to occur.

(ii) ClONO2 Observations -- The evolution of chlorine nitrate, ClONO2, in the southern hemisphere on the 500 K surface for the same eight days in 1992 as in the previous figure is shown in Figure 11.55. These measurements were made by the UARS-CLAES instrument. Note that values of ClONO2 are small in midsummer, and begin to increase in the fall. As temperature decrease inside the polar vortex to the necessary threshold, PSCs form, allowing heterogeneous reactions to convert the chlorine in ClONO2 into the reactive forms. Thus, large ClONO2 deficits are observed inside the polar vortex during the July, August, and September periods. The ClONO2 distributions also display large variability inside the polar vortex. An example of such variability is the large maximum of ClONO2 near the Antarctic peninsula on 17 August 1992. This maximum results when HNO3 is photolyzed and the photolysis product NO2 reacts with ClO to reform ClONO2 as HNO3 Such behavior is evident in the trajectory calculation of Figure 11.52.

(iii) HNO3 Observations -- Nitric acid, HNO3, is formed by heterogeneous reactions that convert nitrogen from reservoir species such as N2O5 and ClONO2 in the presence of PSCs. HNO3 remains on the solid (frozen) surfaces of the PSC cloud particles, and subsequently sediments out of the stratosphere. Figure 11.56 displays UARS-CLAES observations of the evolution of HNO3 for the same 8 days in 1992. Nitric acid amounts are small in midsummer, and begin to increase in the fall because of the downward transport of higher concentrations of HNO3 into the lower stratosphere. As temperatures decrease inside the polar vortex and PSCs eventually form, the nitrogen in ClONO2 and N2O5 is converted by heterogeneous reactions into HNO3. This is the denoxification process by which reactive nitrogen is sequestered inside nonreactive compounds like nitric acid. The HNO3 remains on the solid PSC surfaces. As the temperatures continue to cool, large ice particles form and fall out of the stratosphere, removing the HNO3. This is known as the denitrification process. By midwinter, a substantial fraction of the HNO3 has been removed from the polar vortex. Just outside the vortex in the collar region, there remains relatively high concentrations of nitric acid, since PSCs do not form here.

As noted in the previous paragraph, the photolysis of HNO3 into OH and NO2 will decrease the concentrations of ClO. The maximum of ClONO2 near the Antarctic peninsula on 17 August 1992 in Figure 11.55 is related to the maximum of HNO3 in Figure 11.56. The HNO3 photolysis product NO2 reacts with the high amounts of ClO inside the polar vortex creating the ClONO2 maximum in Figure 11.55. The amounts of HNO3 inside the polar vortex remain low into November.