Ozone loss is especially strong over the frozen Antarctic continent because the winter circumpolar stratospheric vortex prevents extensive air exchange with mid-latitudes. This produces very low temperatures (below -80°C) which favour the generation of polar stratospheric clouds (PSC) of ice particles. Normally, chlorine and bromine are 'locked' into stable reservoir compounds (such as ClONO2 BrONO2 and HCL) The ice particles attract water vapour and absorb nitrogen compounds, then fall with them to lower levels of the atmosphere dehydrating and denitrifying the air in the stratosphere. With the return of sunlight in the early spring, these reservoir compounds are converted to active chlorine and bromine species on the surface of the PSCs. These substances can break apart ozone molecules with amazing efficiency (Figure 6 and Figure 7).

Figure 6 -- Schematic description of ozone destruction over Antarctica

Figure 7 -- Schematic sequence of the destruction of ozone by active Cl released from a CFC-12 molecule

In October 1987, ozone concentrations over Antarctica fell to half their normal (1957- 1978) levels, and the hole spread across an area the size of Europe. Since then, the decline has accelerated and extremes reached during the past three years include:

The ozone destruction is strongest in the lower stratosphere. During late September and October of the last three years the ozone in Antarctica was practically annihilated between 13 and 20 km as shown in Figure 8. The ozone decline in spring, when the Antarctic stratosphere is isolated and extremely cold, is many times greater than the decline in the summer season (Figure 9).

Figure 8 -- Vertical ozone profiles at Syowa (69°S) during October 1992 illustrating the complete ozone destruction in the lower stratosphere. Average ozone profile from pre-ozone-hole years (1967- 1980) is also shown (Ito, JMA, Tokyo)

Figure 9 -- Ozone seasonal deviations from pre-ozone-hole (1957-1978) averages in the Antarctic. Summary for stations Faraday, Syowa, Halley Bay, South Pole (WMO Bulletin 1994)

The ozone hole forms only over Antarctica because of the unique combination of weather conditions favouring ozone- destructive reactions upon the appearance of the sunlight in spring. Figure 10 shows the extent of the ozone hole on 17 October 1994 spreading over the southern tip of South America.

Figure 10 -- Ozone contour map based on TOMS on Meteor 3 data shows the elongated ozone hole reaching over the southern cone of South America on 17 October 1994
(NASA -- Goddard TOMS Processing Team)

Since 1988, extensive measurements have been taken in the northern polar region, including many from aircraft, balloons and specialized satellites. These measurements and those from ground-based international expeditions revealed that during winter-spring the Arctic stratosphere has the same type of disturbed chemical composition, with high concentrations of destructive chlorine and bromine compounds, that causes the problems in the Antarctic. However, ozone destruction over the Arctic is not so strong for two reasons: the stratospheric temperatures are seldom below -80°C due to the frequent extensive exchange of air masses with the middle latitudes; and the Arctic vortex normally dissipates in late winter before sunlight can cause large-scale ozone destruction.


QBO-Alternation of easterly and westerly wind regimes in the stratosphere in equatorial latitudes with a periodicity of roughly 24 to 30 months. The alternation has substantial effects on atmospheric transport. When the stratospheric winds are westerly. a 6-8% ozone deficiency is observed in mid-polar latitudes. When they are easterly, a similar surplus is usually recorded.

In parallel with the evidence of polar ozone decline, scientists also stepped up their search for global erosion of the ozone layer. During 1987-88, the International Ozone Trends Panel scrutinized recent studies and measurements made by satellites and by ground-based instruments around the world. In WMO Ozone Report No. 18 (1988) the verdict was:

In the Ozone Assessment Report released in 1991 the news was even worse: ozone values had dropped significantly not only in winter- spring but also in summer. Since people spend far more time outdoors and UV-B is highest during the summer. ozone loss at that time of the year poses a much greater threat to human health.

Ozone decline (in per cent per decade ±2s) using GO30S data, January 1964 to March 1994, with linear trend fit for 1979-1994

  Region                 Dec., Jan.,     May, June,     Sept., Oct.,      Year 
                         Feb., Mar.      July, Aug.     Nov. 

  35°-65°N                5.8±1.7         2.6±1.5        2.5±1.0         3.8±1.2
  Northern hemisphere     4.0±1.1         1.9±1.1        1.6±0.9         2.6±0.9
  Southern hemisphere     2.7±1.0         3.4±0.8        6.6±1.5         3.9±0.8
  35°-65°S                3.6±1.2         4.9±1.3        7.3±2.0         5.0±1.0

The estimated error limits are given as plus or minus twice the standard deviation. In fact the rates of decline are significant to more than three times the standard deviation. This means that the results would occur by chance less than once in a hundred years. Similar rates of ozone destruction are reported in the 1994 ozone assessment.

The continuing decline in total ozone since the 1970s is statistically significant all the year round everywhere except over the equatorial belt. The GO3OS quality- controlled data, including satellite data, show that the cumulative ozone decline over the middle and polar latitudes is close to 10%. Taking into account known natural variability the decline in both hemispheres, it is especially strong during the winter- spring (over 6-7% per decade). It is half that during the summer- autumn seasons. Detailed studies show a statistically significant increase in the rate of ozone decline by approximately 1.5-2.0% in the period 1981-1991, as compared to 1970-1980. Numerical expressions of ozone trends over the middle latitudes and the entire northern and southern hemispheres are shown in the table above. These figures provide further strong confirmation of the global decline in ozone.

The greatest winter ozone declines seen in the northern hemisphere occurred in 1992- 1993 and in 1995. Ozone levels between 9 and 20% below normal were recorded in middle and high latitudes. Natural long-term variability is greatest between December and March. In that period, a deviation of more than about 13%, (i.e twice the standard deviation) is unlikely to occur more than once every 20 years. The decline of 20% which occurred in February and March 1993 and 1995 were, therefore, extreme cases.

    Scientific evidence points to the fact that chlorine and bromine released by CFCs and halons are to blame for part of the ozone decline. For example, the NASA Upper Atmosphere Research Satellite measured high concentration of ClO in air masses moving southward from the Arctic into the sunlit 45°-65°N latitudes. Several other factors contributed to the extremely low ozone levels:

  • The lower stratosphere was relatively cold, stimulating PSC generation and chemical ozone destruction on their surfaces;
  • The quasi-biennial oscillation (see box above) was in its westerly phase both in 1993 and in 1995, affecting stratospheric circulation. reducing ozone by 6 to 8%;
  • In 1993 the upwelling air motions of a 'blocking' anticyclone situated over the North Atlantic and Europe for several weeks transported ozone- deficient air from the troposphere in the sub-tropics to the polar region;
  • Remnants of volcanic aerosols from the Mount Pinatubo eruption in June 1991 could have been responsible for a further reduction of I to 2% (mainly on the surface of sulphates) in 1992-93.

    In January-March 1995 the extremely strong ozone deficiency of 15-25% occurred over mid-latitudes from eastern Europe to the Far East. It was especially strong over Siberia (<35%) without the presence of any volcanic aerosols but with obvious abundance of ClO and a strong westerly QBO. However, researchers are not yet fully confident that they know exactly the mechanism behind the ozone drop. Because of the high natural variability in ozone over the Arctic, it is difficult to determine the exact proportion of the ozone destruction attributable to human action.

deviations from long-term mean Figure 11 -- In March 1993, total ozone deviations from the long- term mean show considerable deficiencies over northern middle latitudes. The deficiency is relatively smaller over the polar cap than over the sunnier regions. Note that 13% represents a fall of more than twice the standard deviation. Based on near- real- time information provided by GO3OS. (WMO Bulletin, July 1993)

An ozone decline of nearly 10% is clearly seen on the plot of long-term ozone values over Europe and North America (Figure 12). The principal interannual fluctuations are related mainly to stratospheric air transport variations, related to the phase changes of QBO (see first box) in the equatorial stratosphere, however, the overall decline is in concurrence with the chlorine- and bromine- initiated ozone destruction predicted by models.

Figure 12 -- Total ozone deviations from the 1964- 1980 average (smoothed by 12-month running means) for Europe and North America show a major decline since the early 1970s. The quasi- biennial- oscillations are related to stratospheric dynamics

During the last ten years ( 1984-1993). the overall global ozone average level has fallen to 297 m atm cm from 306 in 1964-1980 (about 3%) (Figure 13). However, if the equatorial belt, where there are no significant ozone changes, is excluded, the decline over the middle and polar latitudes is more than twice as large . Some continental- scale regions have even greater cumulative deficiencies as shown in Figure 12. Figure 14 shows the differences throughout the year between mean ozone concentrations in 1964-1980 and those in the 1984-1993 period. The plot clearly shows the variation from pole to pole. It demonstrates the drastic decline (up to 35% in October) that has occurred over the southern polar region. In the northern middle and polar latitudes the major decline occurs during the winter- spring months when the difference between the 1964-1980 and 1984-1993 periods is close to 7%. There has not been any significant change in the equatorial belt.

[monthly global average ozone]

Figure 13 -- The monthly global average ozone values show substantial decline in the 1984-1993 period especially strong in September- January compared with the 1964-1980 level. Since the global averaging includes the huge surface of the equatorial belt where there are no significant changes, the actual decline in extratropical latitudes is much larger

Figure 14 -- Difference (per cent) between total ozone values for the two periods (1964-1980 and 1984-1993). Southward of 60°S from September through November, the difference is more than 15% with a maximum of 35% poleward of 75° in October

Vertically, ozone decline is strongest in the lower stratosphere. The ozone soundings at Hohenpeissenberg show that the ozone partial pressure in the 19-21 km layer has declined by about 30nb representing about 20% in the past 25 years (Figure 15).

[Ozone above Hohenpeissenberg]

Figure 15 -- Ozone in a 19-21 km layer above Hohenpeissenberg (Germany) has decreased by about 20% in the past 25 years. This 12-months running mean smoothed plot, also clearly shows the quasi-biennial oscillations at this layer. These are related to QBO fluctuations in stratoshpheric dynamics.

At the same time that stratospheric ozone is decreasing, tropospheric ozone--in the northern hemisphere at least--is increasing by about 10% per decade. An ozone increase is also noted over the savannah- fire regions in the tropics. This tropospheric ozone increase is mostly a consequence of the effect of the sun's radiation on specific air pollutants, particularly oxides of nitrogen from surface emissions, aircraft and automobile exhausts, combined with the increasing concentrations of other precursors such as methane and carbon monoxide.

Over the last one hundred years the ozone concentrations near the ground in the northern middle latitudes have more than doubled. Several sources support a lower- tropospheric ozone increase of >1% per year since the end of the nineteenth century.

They include: analytical chemical measurements at Montsouris (Paris) and a widespread network using the qualitative Schönbein method; occasional instrumented measurements from aircraft in the early 1940s; and continued monitoring in Pic du Midi (France) and in southern Germany.

During the last few decades also the ozone above the ground - in the middle and upper troposphere - has substantially increased as shown in Figure 16. The ozone increase in the troposphere, however, cannot compensate for the more severe ozone decline in the stratosphere!

Figure 16 -- Surface ozone values from Montsouris (M) and Pic du Midi (D) for the second half of the last century are less than half of those for the past few decades taken Hohenpeissenberg (HP), Arkona (A), Zugspitze (Z) and Pic du Midi (D). Evidence of troposhperic ozone increase in the northern hemisphere at 400 hPa (~7.2km) through the 1980s is shown by data from the ozone soundings made at Hohenpeissenberg

The ozone increase near the ground poses some threats to human health (e.g. eye and bronchial irritation). Furthermore, because ozone reacts strongly with other molecules, oxydizing them, it can damage the living tissue of plants and animals. The close-to ground ozone is a key component of the smog that occurs during cloudless summer days over many major cities around the world. Governments are attempting to decrease its levels by regulatory measures limiting its specific sources. Some success in this direction has been noted in Europe and North America during the last few years.


A joint publication of the World Meteorological Organization and the United Nations
Environment Programme on the occasion of the fiftieth anniversary of the United Nations

World Meteorological Organization
and United Nations Environment Programme