The basic observations on the Antarctic ozone hole have been covered above in Section 2.0. Refer back to it for historical background on observations of and early hypotheses for the existence of the ozone hole. We've also touched upon the structure of the ozone hole. In this section, we will review this structure in much greater detail. We begin with a discussion of the ozone hole's horizontal and vertical structure. This includes discussions of (1) the day-to-day variability of the ozone hole; (2) the growth of the ozone hole during the August-September time period; and (3) the decay of the ozone hole in the November-December time period.
There are at least three aspects to the horiztonal structure of the ozone hole as it appears in October. First, the ozone hole is almost symmetric about the South Pole, which is usually located near its midpoint. Second, there is an underlying "wave one" structure in the ozone field that becomes evident when the zonal mean field is subtracted out. By wave one, we mean that ozone has one high and one low value along a circle of constant latitude. Thirdly, the ozone hole is quite mobile, sometimes rotating along an elongated axis towards the east. Each of these three points and their implications are discussed below.
4.1.1 Almost Symmetric -- The ozone hole is almost symmetric about the South Pole. Figures 11.26a and 11.26b display TOMS total ozone October averages for 1970 through 1990. The early years (1970-1982) are shown in 11.26a, and the later years (1983-1990) shown in 11.26b. Figure 11.26a displays low total ozone over Antarctica inside the collar region that is just slightly offset towards the South Atlantic Ocean. Figure 11.26b shows a similar structure of low total ozone over Antarctica that is also slightly offset in the same direction. The difference between the two periods is that ozone amounts are much lower over Antarctica in the the later years, corresponding to the annual appearance of the ozone hole. So the ozone hole has gotten more dramatic over the years, but the approximate geographical configuration has not really changed.
4.1.2 Wave One Structure -- Also note in Figures 11.26 a and b that there appears to be a north-south wave structure in the total ozone field. That is, ozone is almost always higher at 60°S than it is over Antarctica (which extends from roughly 80°S to the pole). If we average the ozone amount for each available longitude (every 2°) on a circle of constant latitude, we obtain a zonal mean. That is, we take a longitudinal average for each available latitude (every 1°). This result is plotted in Figure 11.27.
If we now subtract this zonal mean field (Fig.11.27) from the total ozone field (Fig.11.26), we can see the wave structure in the total field. This is shown in Figure 11.28.
Note that the October fields are dominated by a single low in the South Atlantic region, and a single high near 150°W. This high/low structure is known as a "wave one" pattern, since we encounter only one maximum and one minimum as we travel (longitudinally) around the pole on a constant latitude circle. If we encountered two maxima and two minima, we would have a wave two pattern. The wave one pattern has its maximum amplitude near 60°S. It falls off to a near zero amplitude near 40°S
We already know that global total ozone amounts are low over Antarctica, have a midlatitude maximum, and are low in the tropics. Prior to 1980, the October average amounts in the polar region were greater than 280 DU. These amounts have decreased in the late 1980's and 1990's to about 120 DU.
4.1.3 Highly Mobile; Eastward Rotation -- The daily structure of the ozone hole is illustrated in Figure 11.29 for an 8-day period in October 1996 using Earth Probe TOMS (EP-TOMS) satellite data.
As is evident from this sequence of daily false color images, the ozone hole tends to be highly mobile (also evident in the images of total ozone over Antarctica, including the Syowa station, in 1983, shown in Figure 11.07). A typical pattern that develops is the elongation of the ozone hole that slowly rotates eastward. The Figure 11.29 panel for 10 October 1996 shows an ozone hole that is generally centered on the pole, but is elongated towards the tip of South America. On 11 October 1996, this elongation axis is oriented towards the South Atlantic. The hole continues to rotate in a clockwise sense over this 8-day period. Such elongations result in quasi-periodic passages of extremely low ozone amounts over sites on the edge of the hole, such as the Antarctic Peninsula. For example, note the ozone difference between 12 and 16 October at Cape Horn. Such large variability in overhead ozone is not typical of sites such as the South Pole, which is located usually near the center of the ozone hole.
In this section, we explore a little bit about the vertical structure of ozone. We examine how ozone varies with height by the different ways it is measured. This includes ozone mixing ratio (the fraction of ozone molecules to the total number of air molecules), ozone density (the number of ozone molecules per unit volume of air), and partial pressure (the fraction of air pressure due to the pressure of ozone molecules alone). Partial pressure is proportional to ozone density, which you can get by dividing the partial pressure by the local temperature. The results from these different ways of measuring ozone allow us to see important features in the vertical distribution of stratospheric ozone.
4.2.1 Mixing Ratio and Density -- Ozone is produced in the stratosphere via the photolysis of oxygen molecules by energetic UV light (see Chapter 8). As a result, ozone mixing ratios are largest in the tropical middle stratosphere, since it is here that there is the most solar radiation. The Brewer-Dobson circulation transports these high ozone concentrations through the winter hemisphere, towards the pole, descending from the upper to the middle and lower stratosphere (see Chapter 6). Figure 11.30 displays a zonal mean image of ozone observations from the Nimbus-7 SBUV instrument in October 1987.
The left panel of Figure 11.30 displays ozone mixing ratio, while the right panel displays ozone density. The right panel shows that most of the ozone is contained in the lower stratosphere between about 70 hPa to 20 hPa (18 and 28 km). Over Antarctica, both the ozone density and mixing ratios were extremely low during 1987. The important features of these Nimbus-7 SBUV images are: (1) low ozone amounts in the lower stratosphere over the Antarctic region resulting from chemical loss processes; (2) high ozone amounts in the midlatitudes resulting from the poleward and downward circulation near the edge of the polar vortex; (3) low amounts of ozone in the tropics resulting from the upward lifting of ozone-poor tropospheric air; and (4) increasing ozone amounts with altitude caused by production of ozone via oxygen molecule photolysis.
It should be noted that the SBUV observations are unable to resolve the vertical structure of the lower stratosphere, and hence the SBUV instrument is not an adequate monitor of the ozone hole loss processes which predominantly occur in the lower stratosphere. For this reason, we turn to vertical profiles provided by ozonesondes.
4.2.2 Ozonesonde Vertical (Partial Pressure) Profiles -- Vertical ozone profiles observed with ozonesondes at South Pole station provide a clear picture of the evolution of the ozone hole as the sun rises over Antarctica during the August-September period. Figure 11.31 shows a sequence of profiles acquired at South Pole station in 1994 (courtesy of NOAA/CMDL).
This image displays ozone partial pressure rather than mixing ratio or density as displayed in Figure 11.30. The figure shows two key features of the ozone hole: (1) the ozone hole is largely confined to the 14-22 km region over Antarctica, and (2) astonishingly, virtually 100% of the ozone is destroyed in this 14-22 km region between early August and late September. (The ozone hole is an apt name in this region!)
The annual cycle of total ozone over the whole globe, including Antarctica, is displayed in Figure 11.32. This figure shows Nimbus-7 TOMS total ozone as a function of time and latitude. The data are averaged over both time (1979-92) and longitude. The large white bites taken out in the polar regions during the winter months result from the inability of TOMS to make measurements during polar night, since the TOMS observations require solar UV light for its ozone measurement technique.
Figure 11.32 shows ozone amounts over the course of an entire year for all latitudes outside of the polar night. As it is based on 14 years worth of data, we can inquire as to whether the average amount of ozone changes over the course of the year at different latitudes. From the figure, we can see that ozone amounts vary little over the course of the year in the tropics, while they vary considerably over both polar regions.
During October, ozone amounts are extremely low over Antarctica, with a collar of high ozone just north of there in the 40°-70°S region, and relatively low amounts throughout the tropics. (Figs. 11.26a-b show a similar picture.) The southern hemisphere ozone high collar region is almost always present, though the amounts decrease in the southern summer. It reaches its highest amounts in late October as a result of the continual accumulation of ozone in the lower stratosphere that is driven by the poleward and downward transport of the Brewer-Dobson cell.
4.3.1 Evolution of Ozone Hole in 1992 -- The evolution of the ozone hole from its onset in August to its disappearance by December is illustrated in Figure 11.33 with a set of Nimbus-7 TOMS total ozone images taken on eight selected days from August to December 1992. As in the previous figure, regions of polar night are not observable by TOMS. These show up as black areas over Antarctica in August and September.
The images show a rapid deepening of the ozone hole from mid-August to early October, followed by its disappearance after early October. This is also shown in Figure 11.32 at the bottom of the image between October and November. Meanwhile, the high ozone collar region increases in amount during the August-September period, and is quite strong by early October.
4.3.2 Minimum Ozone Amounts: 1992 Versus 1996 -- From Figure 11.33, we see that the minimum value of ozone over Antarctica in the 8 days shown occurs on 4 October 1992. Column ozone drops to 126 DU at 81°S, 1°W. This raises an important question: are the minimum amounts inside the ozone hole changing from year to year? More to the point, are the minimums getting smaller each year?
We can answer this question by plotting the actual daily TOMS ozone amounts for different years and compare the results. In Figure 11.34, we have the TOMS minimum ozone amounts for each day in 1992 plotted as a solid line and for each day in 1996 plotted as dots. The daily "minimum" amounts are based on the fact that ozone exhibits a diurnal cycle. The values displayed in Figure 11.34 are based on the zonal mean averages that are then averaged over the 40° to 90°S latitude band. Thus, the amounts are averaged both longitudinally and latitudinally averaged. The 1992 readings are plotted as the solid black line and the 1996 readings are plotted as dots. The solid white line represents the average of daily minimum amounts for each date based on 1979-1994 TOMS measurements. The gray shading shows the full range of minimum amounts for each day in the 1979-1994 period. Thus, the 1996 readings are not included in the white line average or the gray shading range.
The gray shading shows that while both 1992 and 1996 had extremely low total ozone values, the record low observations occurred during other years, since neither the black line nor the black dots coincide with the bottom of the gray range. It turns out that the all-time minimum readings were made in 1993 and 1994. Ozone amounts during the time of the ozone hole (September-October) appear to be slightly lower in 1996 than in 1992. We also note that the recovery in ozone begins in October and is complete by December. This shows up in all the curves and the gray shaded region on the image.
4.3.3 Determining the Size of the Ozone Hole From the TOMS Record, 1979-1996 -- Another way of measuring the severity of the ozone hole is to measure the total surface area of the ozone hole. The "hole" is defined as an area where total column ozone amounts are under 220 DU. The black line sketched around the ozone hole in the those panels of Figure 11.33 where the hole is present indicates the area within which total ozone amounts are less than 220 DU.
(i) Why the 220 DU contour? -- The 220 DU contour is a good representation of the ozone hole, since: (1) it cleanly separates the low total ozone from the high total ozone; (2) it is an amount of total ozone that was not observed over Antarctica prior to 1979, and hence represents a region of real ozone loss with respect to the historic record; and (3) it is relatively insensitive to variations in absolute instrument calibration. By this, we mean that 220 DU typically exists within a sufficiently tight gradient of total ozone that the effects of calibration errors in instrument measurements (which are typically on the order of 5 DU), which then produce errors in our areal size estimate, are minimized. This is not true of a higher ozone amount, such as the 300 DU level, since the gradient is fairly small around the 300 DU contour, and a 5 DU calibration error can easily produce large errors in our estimate of area inside the 300 DU contour.
(ii) Annual cycle in ozone hole size -- In early August, there is only a small region near the Antarctic peninsula at the edge of polar night where ozone amounts are below 220 DU. By mid-September, the area of the ozone hole has greatly expanded to well over 20 million square kilometers. The area contained within the 220 DU contour has been plotted in Figure 11.35. The black line represents the area inside the 220 DU ozone contour for 1992 and the black dots represents the area inside the 220 DU ozone contour for 1996. The average ozone hole size for 1979-1994 is plotted as the thick white line, while the gray shading displays the range of area values observed on each day over the same 16 year period. The 1992 data is thus included in the 16 year average, while the 1996 data is not. As in Figure 11.34, such a comparison allows us to see whether or not the areal size of the hole has changed between 1992 and 1996, and to compare both to a longer record.
From the figure, we see that the ozone hole is not well resolved in August because of the polar night. During this time, the TOMS instrument cannot measure ozone. By early September, the hole is clearly observable, rapidly growing in size. The areal extent of the ozone hole reaches a maximum by late-September. It begins to contract in early October, falling quickly in late October and November. The hole has vanished by mid-December.
(iii) Average size of the ozone hole, 1979-1996 -- The average size of the ozone hole can be determined by averaging the daily amounts displayed in Figure 11.35 between early September and mid-October. This average is shown in Figure 11.36 for all of the years 1979 to 1996, except 1995, since no TOMS instrument was in orbit this year. The average amounts are displayed as the heavy yellow dots, while the vertical red lines show the range of observations for a particular year.
Figure 11.36 shows us that in 1992, the average size of the ozone hole was slightly less than 22 million square kilometers, but at its maximum (represented by the top of the vertical red line) in late September, the hole exceeded 24 million square kilometers. The average size in 1996 was nearly the same as in 1992, but the maximum extent of the the hole expanded to over 26 million square kilometers. The maximum size of the ozone hole is approximately the same area as the North American continent (about 24 million square kilometers).
(iv) Vertical range and severity of ozone hole
1. POAM II satellite data -- The ozone hole develops quite rapidly over a broad altitude range as the sun rises over Antarctica in the spring. Figure 11.37 shows ozone number density observations from the POAM satellite between 15 and 35 km from November 1, 1993 to November 1, 1995. In this data, we can see the vertical depth and severity of the ozone hole as it quickly develops. Examining Figure 11.37, we see that at 20 km, the ozone number densities fall from peak concentrations of 5x10-12 /cm3 to nearly zero over the August-September period in the altitude range from 16 km (the lowest altitude that the POAM II satellite instrument can "see" into the atmosphere) to 22 km. The data is in excellent agreement with the South Pole balloon observations plotted in Figure 11.31. Recall that balloon observations there showed a nearly 100% percent loss in ozone between 14 and 22 km in the same August-September period. POAM II data show that ozone number densities recover in the November-December period. The decrease of ozone observed by the POAM II satellite is best seen by observing how ozone number density concentrations varies at the 20 km altitude level, as shown in Figure 11.38. We see in Figure 11.38 how ozone amounts first decrease slowly in July and August, followed by a rapid drop during September to near zero by late-September. This is followed by a steady recovery in the October-December period.
2. McMurdo Base, Antarctica balloon data -- Other data similarly show this severe drop in ozone through a significant vertical range of the lower and middle stratosphere during the August-September period. Observations of column ozone amount between 12 and 20 km by ozonesonde balloons launched from McMurdo Base, Antarctica show ozone decreases of over 100 DU between mid-August and early October. This is shown in Figure 11.39.
(v) Hemispheric perspective of the ozone hole using MLS satellite data -- While the development of the ozone hole is discernible from column ozone observations (TOMS) over the hemisphere, this instrument provides little information on the vertical structure of the ozone hole. The POAM II satellite profiles and the McMurdo balloon provide excellent vertical information on the ozone hole, but do not give us a hemispheric perspective on the hole. The MLS instrument on the UARS satellite provides information on both the vertical and horizontal structure of the ozone hole. Figure 11.40 shows a series of MLS false color images of ozone on the 500 K isentropic surface (approximately 22 km).
Based on the plots in Figure 11.40, we see that January (early summer) shows a relatively flat ozone field. Ozone amounts steadily increase through the fall and winter, peaking by July, corresponding to southern mid-winter. The decrease begins in mid-August, as ozone amounts begin to decrease inside the polar vortex (denoted the black lines on the plots). By mid-September, the ozone amounts are nearly zero over Antarctica. Ozone amounts begin to climb as the polar vortex breaks down when sunlight returns and warms the stratosphere, though amounts are still relatively low in early November. The ozone hole completely disappears by December.
These images reinforce several key points about the ozone hole: (1) the hole develops in the August-September period; (2) ozone concentrations build up in the lower stratosphere by downward advection (via the Brewer-Dobson winter time circulation) of high ozone mixing ratios from the upper stratosphere; (3) the region of ozone depletion inside the polar vortex is highly contained; and (4) the recovery of ozone occurs in November and December (mid-to-late spring) as the polar vortex region breaks down when the stratosphere over Antarctica warms.
Following the August-September ozone hole period, ozone amounts within the Antarctic polar votex where the hole is found begin to rise to more normal levels. The is recovery is apparent in Figures 11.33 (ozone hole maps), 11.34 (minimum ozone values over Antarctica), 11.35 (ozone hole areal coverage), 11.37 ( vertical distribution of ozone), and 11.38 (ozone concentrations at 20 km). The recovery is driven largely by the breakdown of the polar vortex, when sunlight reappears and warms the Antarctic stratosphere, and the consequent horizontal transport and mixing of higher ozone concentrations from the midlatitudes into the south polar region.
The polar vortex breaks down via the steady decrease of wind speeds during the spring period, which in turn is due to the warming influence of the Sun, with the tight, pole-equator temperature gradient becoming smaller. As the wind decreases, the barrier to transport between the midlatitudes and the polar region weakens, and ozone-rich, midlatitude air is swept over Antarctica. The breakup of the vortex begins first at higher altitudes (around 36 km in August), and gradually moves downward, with the breakup at 16 km occurring in early December. The period of ozone recovery first appears at the higher altitudes, and gradually descends to lower altitudes. This recovery is evident in Figure 11.37, where the low ozone amounts inside the hole (purple) first increase at the higher altitudes, with later recovery at lower altitudes.
The steady filling of the depleted region from the top downward manifests itself in the column concentrations as a gradual rise in the minimum total ozone concentration, as shown in Figures 11.33 (ozone hole maps) and 11.34 (minimum ozone amounts over Antarctica), and a steady decline of the ozone hole area, as shown in Figure 11.35 (ozone hole areal coverage). The recovery of ozone at a single station is quite dramatic, as high ozone concentrations are swept over given locations in Antarctica, as shown in balloon ozone measurements in Figure 11.05 (Halley Bay) and Figure 11.06 (Syowa).
The ozone hole observations are summarized in Figure 11.41. First, the hole is a seasonal phenomena that develops in August and September, has reached its largest extent and deepest values in early October, and breaks up in early December. Second, the ozone hole was weakly observable as far back as the mid-1970s, and became easily observable by the early 1980s as it grew in severity. Third, the ozone hole constitutes a 60% reduction in total column ozone concentrations, and a 100% local loss in the 12-20 km layer. Fourth, the ozone hole is associated with the Antarctic polar vortex and extremely cold winter temperatures that occur inside the vortex.