E. A. Jadin
Central Aerological Observatory, Dolgoprudny, Moscow region, Russia
One of the important issues of ozone and climate changes is the understanding of interannual and decadal variations of stratospheric parameters. It is now clear that long-term changes are manifested more strongly in the stratosphere than in the troposphere, because there is less forcing of weather systems on the stratosphere. The discovery of the ozone hole in the Antarctic is one of the most prominent examples of a decadal variability of the stratosphere. In spite of a large amount of observational evidence that the ozone layer depletion is caused by man-made activities [World Meteorological Organization, 1994], the causes of interannual and decadal variations of the ozone layer and their relations with low-frequency changes of stratospheric circulation and wave activity are not well understood. Usually regression analyses are conducted to quantify the relations of total ozone anomalies with a trend, the causes of which are a priori believed to be anthropogenic impacts, the quasi-biennial oscillation (QBO), El Ni no-Southern Oscillation (ENSO) variations, and the 11-year solar cycle [Stolarski et al., 1991; Randel and Cobb, 1994]. However, stratospheric dynamics can undergo natural long-term changes of transport processes that may in turn result in the observed ozone trends. For example, Kawahira and Hirooka [1992] and Hirooka et al., [1994] pointed out that there are apparent differences between the stratospheric zonal wind of the southern hemisphere (SH) in the 1970's and 1980's, which may be a demonstration of decadal variations of planetary wave activity in the stratosphere. Labitzke and van Loon [1988, 1993] have shown decadal variations of atmospheric parameters in the northern hemisphere (NH), the cause of which they associated with the influence of the 11-year solar cycle. No explanation of this link exists hitherto. However, Jadin [1993] showed that a decadal variability of tropospheric and stratospheric parameters in the NH, which was discovered by Labitzke and van Loon [1988, 1993], can be caused by similar variability in the sea surface temperature (SST) anomalies of the North Atlantic and Pacific, resulting in decadal variations of planetary wave activity especially in the winter stratosphere. Deser and Blackmon [1993] confirmed the existence of the decadal cycles of SST anomalies in the North Atlantic.
The aim of this work is an attempt to answer the following questions:
1. Why did the ozone layer depletion start mainly in the late 1970's?
2. Are there long-term changes of stratospheric dynamics, and how are they connected with observed ozone trends?
3. Why did the relationship between ozone hole development in the Antarctic and the equatorial QBO [Garcia and Solomon, 1987] disappear during the late 1980's?
A short description of the data used and the method of analysis is presented in the next section of the paper. The subsequent sections are dedicated to the analysis of results and discussion.
The calculations of the stratospheric angular momentum (SAM), which is determined as the atmospheric angular momentum from 100 hPa to 0.4 hPa [Jadin, 1995], are a useful indicator of the interannual variability of stratospheric dynamics. On the one hand, changes of atmospheric wave activity affect zonal wind variations in the middle and high latitudes of the stratosphere during winter/spring times; on the other hand, ozone anomalies in the stratosphere depend strongly on interannual changes of eddy transport processes. The variations of zonal wind vary with altitude; therefore the stratospheric angular momentum indicating the stratospheric circulation changes as a whole can be a better indicator for the search of relations with total ozone anomalies than consideration of zonal wind changes. Besides, the use of the atmospheric angular momentum (AAM) makes is possible to conduct comparisons with data from space observations of the variations of the Earth's rotation or the length of day (LOD), which are measured with the highest accuracy providing an independent verification of atmospheric data sets [Rosen and Salstein, 1985].
The anomalies of the atmospheric angular momentum for each month were calculated
| (1) |
where a is the radius of the Earth, g is gravitational acceleration, u is zonal wind, p is pressure, j is latitude, and pb and pt are pressures at lower and upper boundaries of an atmospheric layer, respectively.
It is known that the AAM variations on seasonal and interannual timescales (up to 2-3 years) are connected with the length of day changes:
 | (2) |
where LOD is in millisecond units and AAM is in 1026 Ēg m2 s-1. This equation follows from the conservation of angular momentum between the atmospheric excitation and the Earth's variable rotation.
The daily data of zonal wind variations of the U.S. National Meteorological Center (NMC) were used for calculations of the atmospheric and stratospheric angular momentum in 1979-1992. Comparison of the calculated AAM changes with the measured LOD variations [Rosen and Salstein, 1985; Jadin and Yamazaki, 1995] has shown their excellent agreement on time scales up to 2-3 years, which testifies to the good quality of the NMC data including the stratosphere. The daily data were averaged and the monthly mean anomalies were calculated for each month during 1979-1992. The results are expressed in millisecond units using relation (2). Satellite measurements of the total ozone (TOMS, version 6) were used for calculations of ozone anomalies and trends in 1979-1991.
In order to indicate the principal components that are most responsible for the interannual variability of stratospheric dynamics and ozone, several first empirical orthogonal functions (EOF) [Peixoto and Oort, 1992] of the total ozone and stratospheric momentum anomalies were calculated. Analysis of the relations between the interannual anomalies of total ozone and stratospheric circulation was conducted by means of singular value decomposition (SVD) calculations, which maximize correlations between two fields of data [Bretherton et al., 1992].
Interannual changes of the tropospheric angular momentum AAM(100)
determined as AAM from 1000 hPa to 100 hPa are mainly connected with the
El
Ni no events, which lead to strong westerly anomalies of zonal wind in
subtropics of the middle and upper troposphere
[Rosen et al., 1984].
In
spite of the fact that only 10-20% of the global atmospheric angular
momentum is from the stratosphere, interannual
anomalies of the stratospheric angular momentum are comparable with those
of the tropospheric momentum, as is seen in Figure 1. The signals of El
Ni no 1982/1983 and 1986/1987 events are also highlighted in the westerly
SAM anomalies; however, there are large differences in the behavior of
interannual variations of the global tropospheric and stratospheric
momentum, as is seen clearly in the 1-year running means of the global
AAM(100) and SAM anomalies. These differences take place not only in the
amplitudes, but in the sign of anomalies also. For example, large
easterly SAM anomalies occurred during 1979 and the first half of 1980, in
contrast with small westerly anomalies of the global AAM(100). Similar
features were observed during 1991-1992. Conversely, westerly anomalies
of the global stratospheric momentum occurred in the second half of 1985, in
contrast with large easterly tropospheric momentum anomalies caused by the La
Ni na 1984/1985 event. A Fourier transform analysis
was conducted to
explain in the first approximation the cause of these differences of the
tropospheric and stratospheric circulation. A similar analysis has been
applied to the Southern Oscillation Index (SOI)
[Dickey et al., 1992].
The quasi-biennial (QB) oscillation in the spectral interval 19-32 months and
the low-frequency (LF) component (longer than 39 months) were separated in
the global AAM(100) and SAM anomalies (Figure 1). Large westerly
anomalies of the tropospheric momentum during the El Ni no events in
1982/1983, 1987 and 1992 can be accounted for by constructive interference
between the QB and LF components of the AAM(100) anomalies, as well as the
large easterly anomalies during the La Ni no events in 1984 and 1988/1989.
Note that the behavior of the QB and LF components of the stratospheric
momentum anomalies is distinguished from those of the tropospheric ones. The
strong easterly anomalies take place in the LF component especially during
1979/1980 and 1991/1992. Constructive interference between the QB and
LF components of the SAM anomalies in 1979/1980 led to the strong easterly
anomalies of the stratospheric circulation at the same time. Destructive
interference of the QB and LF components of the AAM(100) is observed in
1985, while the strong westerly QB component of the
SAM anomalies is
responsible for westerly anomalies in the stratosphere in the second half of
1985. The large easterly LF component of the stratospheric momentum anomalies
resulted mainly in the strong easterly anomalies of the stratospheric
circulation during 1991-1992. In contrast with AAM(100) the downward trend
of the LF component and the 1-year running mean of the SAM anomalies is
observed during 1981-1992 with an abrupt transition from easterly to
strong westerly anomalies during 1979/1980 and a clearly expressed QBO
signal.
Analysis of the zonal wind anomalies (Figure 2) showed that rapid
changes in the circulation occurred in the lower subtropical stratosphere
in the summer of 1980. The unusual structure of zonal wind anomalies
(so-called "easterly cat eyes")
[Jadin, 1997]
centered at 50 hPa near
30o S and 30o N was observed during 1979 and first half of
1980 with
slight seasonal variations. This structure did not recur in the other
years considered. The strong easterly anomalies of stratospheric momentum
in 1979/1980 are mostly associated with this structure. The easterly cat
eyes disappeared in June 1980, and then the strong westerly anomalies of
zonal winds occurred in the subtropics of the SH and near the equator in the
upper and middle stratosphere. It is known that there are
uncertainties in
the NMC data for 1979/1980, but
Jadin and Yamazaki [1995]
indicated
very good agreement between the LOD and AAM changes for this
period, which is evidence for the good quality of the NMC zonal
wind data including the stratosphere. The influence of the El Ni no
1982/1983 event reached a maximum in February 1983 (Figure 1) and penetrated
to the stratosphere in the NH subtropics. The "westerly cat eyes" centered
near the equator at 30 hPa occurred in August 1985. At
the same time the
easterly zonal wind anomalies were observed in the southern tropical
troposphere. Strong easterly anomalies arose in the upper stratosphere
in 1991/1992, which were propagated downward to the lower stratosphere and
upper troposphere. It is interesting that strong easterly anomalies of zonal
wind occurred in the upper and middle stratosphere before the Pinatubo
eruption (June 15, 1991); therefore the strong easterly anomalies in
1991/1992 (Figures 1
and 2)
cannot be fully related to the impact of the
Pinatubo eruption on stratospheric dynamics
[Rind et al., 1992;
Graf et al., 1994].
Figure 3 shows the long-term meridional propagation of stratospheric
momentum for the distinct October and March months during 1979-1992. Excepting
1979-1981, 1986, and 1988, westerly SAM anomalies dominate in middle
(30-60o S)
latitudes of the SH for October during the entire
period considered. Strong easterly SAM anomalies are observed near the
equator in the SH in October 1979, which were followed rapidly by
large westerly anomalies in 1980. Large easterly anomalies arose in
1991/1992, which resemble those in 1979. The long-term propagation of the
SAM anomalies is similar to the V structure of the propagation of the
low-frequency filtered AAM(100) anomalies
[Dickey et al., 1992],
which originate near the equator and propagate to the polar regions.
This V structure is highlighted in winter/spring seasons, which is
evidence for planetary wave forcing in the formation of the V
structure. The westerly SAM anomalies dominate also in the NH middle and
high latitudes for March with slightly less prominent V structure.
This can be explained by larger longitudinal asymmetry of winds in the NH in
comparison with those in the SH. Figure 3 also displays the QB and LF
components of the SAM anomalies in the middle latitudes for October and
March. The long-term variations of the stratospheric momentum in the belts
can also be explained in view of constructive or destructive
interference of the QB and LF components. For example, the strong westerly
SAM anomalies (50-55o S) in October 1987 are caused by their
constructive interference, while the easterly SAM anomalies in 1988 are
caused by destructive interference. Note that the development of the
ozone hole in the Antarctic corresponds to these variations of
stratospheric dynamics. The upward long-term trend is evident in the behavior
of the LF component of the SAM anomalies in middle and high latitudes
during winter/spring seasons. This may hint at a decadal decrease of the
planetary wave activity here.
It appears to be useful to separate on the principal modes
of
the total ozone and stratospheric circulation anomalies on a global scale
by means of decomposition of the corresponding fields with the empirical
orthogonal functions (EOF),
taking into account that the advective and eddy transport processes
strongly compensate each other. It is known
[Peixoto and Oort, 1992]
that the first EOF modes and their time series are most responsible for the
temporal variability of the field analyzed. The structure of the first three
EOF's of total ozone and stratospheric momentum anomalies and their time
series in 1979-1991 are shown in Figure 4. The features of the first EOF of
total ozone anomalies and the behavior of its coefficient are very similar to
those calculated by
Randel and Cobb [1994],
but with a smaller
contribution to total variability (41.1% in comparison with their 78%). It
is seen that EOF1 describes the total ozone trends on a global scale
[World Meteorological Organization, 1991].
The second EOF (15.9%
variance) is probably associated with the influence of quasi-biennial
oscillation and resembles the QBO fit in the regression analysis
[Randel and Cobb, 1994]
except for high latitudes of the northern
hemisphere during wintertime. It should be noted that the coefficient of the
EOF2 of ozone anomalies also has a downward long-term trend similar to
that of the EOF1, which can testify to long-term changes of the QBO
amplitude of ozone variations during 1979-1991. In addition, the
compensation of the EOF1 and EOF2 contributions to total ozone variability is
observed in September-October in the Antarctic, which is most prominent in
1985, 1987, and 1990. In this regard the significance of the third EOF of
ozone anomalies can be important despite its relatively small (14.3%)
role in the total variability. The EOF3 structure has distinct
interhemispheric differences, and this mode gives the most negative
contribution in the Antarctic in 1987 and to a smaller degree in 1982. The
most prominent impact of the EOF3 is observed in the midlatitudes in the NH
in March-April 1988. This is coincident with the observed variations of
ozone over the Antarctic and the middle and high latitudes of the northern
hemisphere
[World Meteorological Organization, 1994].
The first EOF of the stratospheric momentum anomalies (25.7% variance) is responsible for changes in stratospheric dynamics in the equatorial and southern subtropical latitudes. The temporal behavior of the coefficient of the EOF1 of the SAM anomalies is similar to that of the interannual variations of the global stratospheric momentum; namely, we observe a sharp jump from easterly anomalies in 1979 to strong westerly anomalies in the first half of 1980, which was most prominent in 1983 (El Ni no influence). The long-term downward trend is also seen during 1981-1992. Note that the second half of the year gives a larger contribution to the interannual variability in the tropics and subtropics of the southern Hemisphere. The second EOF of SAM anomalies (21.6%) is probably associated with semiannual oscillations of the stratospheric dynamics. The strong negative minimum of the EOF2 in tropics of the northern hemisphere during winter/spring times and the jump of its coefficient in 1980-1981 testify to a contribution of this mode to the transition toward a new regime of stratospheric circulation after summer 1980 [Jadin, 1996]. It is observed in the constructive or destructive interference of EOF1 and EOF2 modes in the total variability as well as in the case of interannual ozone anomalies. The constructive interference in the tropics and subtropics of the northern hemisphere during winter/spring occurred most strongly in 1983; the destructive one, in 1980.
The structure of the third EOF mode (16.6%) of SAM anomalies has clearly expressed minima in winter/spring in the middle and high latitudes of the northern hemisphere, subtropics (June), and high latitudes (September-November) of the southern hemisphere. It is interesting to point out that the EOF3 coefficient of the stratospheric momentum anomalies has a long-term trend and its behavior is similar to that of the EOF1 coefficient of the total ozone anomalies. The structures of the EOF1 mode of total ozone and EOF3 mode of stratospheric momentum anomalies also resemble each other, but with opposite sign.
In order to estimate how strongly the observed ozone trends
are associated
with zonal circulation anomalies of the stratosphere, the leading SVD
modes of their correlation matrix were calculated
[Bretherton et al., 1992].
Figure 5
shows the first SVD modes of the monthly mean anomalies of
the total ozone and stratospheric momentum and their time series in
1979-1991. They are expressed as the correlations (in percent) between the
coefficient of the first SVD mode of the SAM anomalies and the field of the
total ozone anomalies (heterogeneous correlations) and the field of the
calculated stratospheric momentum anomalies (homogeneous correlations). The
strongest correlations (up to 80-90%) are observed in the high latitudes of
the southern hemisphere in September-November in the ozone hole region
over the Antarctic. The estimates showed that these correlations are
statistically significant on a 95% level of variance. The maxima of
correlations also take place in winter (January-February) in the
middle and high latitudes (March) of the northern hemisphere. The local
maximum is remarkable in high latitudes of the northern hemisphere in
October.
The features of the correlations are very similar to the structure of observed ozone trends [World Meteorological Organization, 1991], and the EOF1 mode of total ozone anomalies, as well as the behavior of their coefficients. This can imply that the first SVD mode is responsible for the global ozone trend. On the other hand, the negative correlations between the interannual variations of the coefficient of the first SVD mode and the stratospheric momentum changes are also great in the high latitudes (September-October) of the southern hemisphere, in the middle and high latitudes (January-March) and low latitudes (April-June) of the northern hemisphere. It is interesting to point out that the first SVD mode of the SAM anomalies and their time series resemble the third EOF of the stratospheric momentum anomalies, which is probably responsible for the forcing of planetary waves on the zonal mean flow. This gives evidence that observed ozone trends in particular in the ozone hole region are very tightly associated with long-term anomalies of the circulation and wave activity of the atmosphere.
The relations of total ozone variations in the extratropics with the QBO of
zonal wind of the lower stratosphere at Singapore have been investigated in
some works
[Angell, 1988, 1993].
Garcia and Solomon [1987]
pointed out that the relative decreases (increases) of total ozone in the
Antarctic in October during 1979-1986 correspond to the westerly (easterly)
phase of the QBO zonal wind at 50 hPa over Singapore. However, this link was
distorted during recent years
[Herman et al., 1995].
The cause of this
is unclear; it is believed that a strengthening of anthropogenic impacts
during the past decade may lead to a chemical destruction of
ozone layer as being the dominant response to the QBO
signal. The QB components of the stratospheric momentum and total ozone
anomalies for October 1979-1991 are shown in Figure 6. In spite of the
fact that zonal winds in the 10o S-10o
N belt were calculated using
a linear interpolation between hemispheres, the sign of the SAM
anomalies near the equator is coincident with the phase of the QBO zonal wind
at 30-50 hPa over Singapore. It is clearly seen that the westerly
(easterly) SAM anomalies of the QB component in midlatitudes are
corresponding to the decrease (increase) in the QB component of total ozone
over the Antarctic in good agreement with previous results
(Figures 5
and 7).
The striking feature of the QB SAM anomalies is the
change of the sign of those between tropics and extratropics in the SH after
1986/1987. A standing structure appears to be observed during 1979-1986,
while the westerly QB stratospheric momentum anomalies at midlatitudes are
corresponding to easterly ones near equator after 1987. The analysis showed
that similar SAM features with a more complex structure (not shown) are
peculiar to the QB component of stratospheric momentum anomalies in the
NH during winter/spring seasons in 1979-1991. Thus the deviations
from the relationship
[Garcia and Solomon, 1987]
between the
equatorial QBO and the development of the ozone hole in the Antarctic can be
explained by a change in the interlatitudinal relations of the stratospheric
circulation after 1986/1987. The cause of such changes is unknown.
The presented results of the analysis testify to a strong link between the interannual anomalies of total ozone and zonal stratospheric circulation in the middle and high latitudes during the winter/spring, i.e., in the time periods when ozone trends are most prominent. Namely during these seasons, stationary planetary waves can penetrate from the troposphere to the stratosphere, strongly affecting both the zonal flow in the stratosphere and the ozone eddy transport and the polar vortex isolation in the Antarctic and Arctic. The weakening of wave activity in the stratosphere can result in an acceleration of the zonal mean flow in high latitudes (westerly anomalies of zonal wind), the strengthening of polar vortex isolation, and a decrease of ozone and heat eddy exchange in the middle and high latitudes on interannual and decadal timescales. Therefore the strong negative correlations of total ozone with stratospheric momentum anomalies (Figure 5) can be understood in view of a forcing of interannual and decadal changes of atmospheric wave activity both on the zonal dynamics and ozone layer anomalies. It is possible that an abrupt decrease of atmospheric wave activity occurred in summer 1980 on a global scale, which led first to the disappearance of strong easterly anomalies of the zonal wind in the subtropics of the lower stratosphere in the southern and northern hemispheres. These strong easterly anomalies were observed during 1979 to the first half of 1980, but then they were replaced by strong westerly anomalies in the tropical and subtropical stratosphere. During a few subsequent years, westerly wind anomalies were propagating from the equator toward the polar regions. This can testify to a weakening of the interlatitudinal eddy exchange by ozone and heat and the strengthening of the polar vortex isolation especially over the Antarctic. The El Ni no event 1982/1983 resulted also in the occurrence of strong westerly anomalies of zonal winds in the troposphere and stratosphere. The decrease of the interlatitudinal exchange by ozone and heat and a strong polar vortex isolation caused cooling of the lower stratosphere in the Antarctic, polar stratospheric cloud formation, and created favorable conditions for the chemical destruction of the ozone layer, which includes heterogeneous chemical processes. This alternative point of view on the cause of the ozone hole appearance over the Antarctic and the observed ozone trends in the northern hemisphere has been proposed by Jadin and Lysenko [1988] and Jadin [1990] and also in another form by Mahlman and Fels [1986]. The first evidence appears to represented here and by Jadin [1996] and Jadin and Diansky [1997] in favor of the wave hypothesis of the ozone hole appearance over the Antarctic and global ozone trends. It should be emphasized that the sequences of this hypothesis do not contradict available observations.
However, the results of analysis of relations between the
total ozone and stratospheric momentum anomalies (Figure 5) may be
interpreted in view of the influence of ozone concentration
decreases on the radiative regime of the lower stratosphere, which
can in turn result in the cooling of the lower stratosphere and
westerly anomalies of zonal wind. Model simulations
[Kiehl et al., 1988;
Mahlman et al., 1994]
showed that the inclusion in the simulations of
ozone hole effects in the Antarctic leads to the cooling of the lower
stratosphere at high latitudes up to 8o K.
Figure 8 shows the
calculated total ozone and zonal wind trends at 100 hPa in 1981-1991. The
model simulations give evidence that the zonal wind trends can be caused by
changes in the radiative regime owing to a decrease of ozone
concentration in the lower stratosphere. It should be noted, however, that
the most prominent ozone hole occurs in September-October, while the
maximum of the positive trends (statistically significant on 95% of
variance) of zonal winds in the high latitudes of the southern hemisphere is
observed in winter in May-August, which is not in agreement with
such a point of view. This question requires further investigation both
in view of the quality of the NMC data in high latitudes of the
southern hemisphere
[Trenberth and Olson, 1988]
and in order to
reproduce successfully planetary wave propagation in general
circulation model simulations
[Yang and Gutowski, 1994].
The other argument in favor of the wave hypothesis of the explanation of global ozone trends is the link of the interannual anomalies of the sea surface temperature with the total ozone variations. Angell [1988, 1993], Kodera and Yamazaki [1989], and Komhyr et al., [1991] have revealed that the development of the ozone hole in the Antarctic is associated with interannual SST anomalies in the central equatorial Pacific. Jadin [1992] indicated the strong correlations of the SST anomalies in the Atlantic with total ozone anomalies over Europe in 1965-1988. These relations can be explained by the influence of interannual variations of the thermal excitation of planetary waves, which depends on SST anomalies and interference between the thermal and orographic generation of planetary waves [Jadin, 1990]. The long-term propagation of anomalies of the tropospheric and stratospheric angular momentum and the strong correlations of ozone anomalies with interannual variations of the stratospheric circulation imply that the long-term ozone changes in the middle and high latitudes depend on anomalies generated in the dynamics of the tropical atmosphere, which have arisen a few years earlier. Such long-term "memory" can be provided by an interaction between the ocean and atmosphere on interannual and decadal timescales.
Finally, the conclusions can be formulated as follows:
1. Analysis of the total ozone and stratospheric angular momentum variations showed that the observed ozone trends are tightly associated with the interannual changes of the stratospheric circulation in particular in the ozone hole region over the Antarctic [Jadin and Diansky, 1997],
2. These relations together with an abrupt transition of stratospheric circulation to a new regime in summer 1980 [Jadin, 1996] imply that the primary cause of the observed trends of ozone layer can be long-term changes of atmospheric wave activity, which are associated with decadal variations in the coupled ocean-atmosphere system,
3. Cause of the deviation from the relationship of the ozone hole development with the equatorial QBO can be connected with changes of interlatitudinal relations of stratospheric dynamics after 1986.
The results can be used for comparison with model simulations.
Angell, J. K., Relation of Antarctic 100 mb temperature and total ozone to equatorial QBO, equatorial SST, and sunspot number, 1958-1987, Geophys. Res. Lett., 15, 915-918, 1988.
Angell, J. K., Reexamination of the relation between depth of the Antarctic ozone hole, and equatorial QBO and SST, 1962-1992, Geophys. Res. Lett., 20, 1559-1562, 1993.
Bretherton, C. S., C. Smith, and J. M. Wallace, An intercomparison of methods for finding coupled patterns in climate data, J. Clim., 5, 541-560, 1992.
Deser, C., and M. L. Blackmon, Surface climate variations over the North Atlantic ocean during winter: 1900-1989, J. Clim., 6, 1743-1753, 1993.
Dickey, J. O., S. L. Marcus, and R. Hide, Global propagation of interannual fluctuations in atmospheric angular momentum, Nature, 357, 484-488, 1992.
Garcia, R., and S. Solomon, A possible relationship between interannual variability in Antarctic ozone and quasi-biennial oscillation, Geophys. Res. Lett., 14, 848-851, 1987.
Graf, H. F., J. Perwitz, and I. Kirchner, Northern hemisphere tropospheric mid-latitude circulation after violent volcanic eruptions, Beitr. Phys. Atmos., 1, 3-13, 1994.
Herman, J. R., P. A. Newman and D. Larko, Meteor 3/TOMS observations of the 1994 ozone hole, Geophys. Res. Lett., 20, 3227-3229, 1995.
Hirooka, T., J. Yamashita, and K. Kawahira, Long-term variations of stratospheric circulations: a comparison of mean fields in the 1980s with CIRA 1986 models, J. Atmos. Terr. Phys., 56, 1041-1049, 1994.
Jadin, E. A., Planetary waves and interannual ozone anomalies in polar regions, Izv. Acad. Sci. USSR Phys. Atmos. Ocean, 26, 1156-1160, 1990.
Jadin, E. A., Interannual variations of total ozone over Europe and SST anomalies in the Atlantic, Meteorol. Hydrol., 7, 22-26, 1992.
Jadin, E. A., A long-term cycle in the sea surface temperature variations, lower stratospheric temperature, and total ozone in midlatitudes, Meteorol. Hydrol., 5, 52-59, 1993.
Jadin, E. A., Anomalies of ozone layer and stratospheric angular momentum, Meteorol. Hydrol., 7, 48-55, 1995.
Jadin, E. A., Long-term propagation of zonal mean anomalies in the stratospheric dynamics, Meteorol. Hydrol., 7, 36-48, 1996.
Jadin, E. A., Diagnosis of long-term changes in stratospheric dynamics, Izv. Acad. Sci. USSR Phys. Atmos. Ocean, 33, 787-794, 1997.
Jadin, E. A., and N. A. Diansky, Analysis of interannual variations of total ozone and stratospheric circulation, Meteorol. Hydrol., 9, 25-35, 1997.
Jadin E. A., and E. V. Lysenko, Long-term variations of stratospheric dynamics and ozone decrease in the Antarctic, Meteorol. Hydrol., 10, 38-42, 1988.
Jadin, E. A., and K. Yamazaki, Changes in the Earth's rotation and the atmospheric angular momentum on intra-annual and decadal time scales, Pap. Geophys. Meteorol., 45, (3), 113-120, 1995.
Kawahira, K., and T. Hirooka, Variations of mean zonal winds relating to ozone hole, J. Geophys. Res., 97, 10,157-10,163, 1992.
Kiehl, J. T., B. A. Boville, and B. P. Briegleb, Response of a general circulation model to a prescribed Antarctic ozone hole, Nature, 332, 501-504, 1988.
Kodera, K., and K. Yamazaki, A possible influence of the sea surface temperature on recent development of ozone hole, J. Meteorol. Soc. Jpn., 67 (3), 465-472, 1989.
Komhyr, W. D., S. J. Oltmans, R. D. Grass, and R. K. Leonard, Possible influence of long-term sea surface temperature anomalies in the tropical Pacific on global ozone, Can. J. Phys., 69, 1093-1102, 1991.
Labitzke, K., and H. van Loon, Association between the 11-year solar cycle, the QBO, and the atmosphere, part I: The troposphere and the stratosphere in the northern hemisphere in winters, J. Atmos. Terr. Phys., 50, 197-206, 1988.
Labitzke, K., and H. van Loon, Association between the 11-year solar cycle and the atmosphere, part V: Summer, J. Clim., 5 (3), 240-251, 1993.
Mahlman, J. D., and S. S. Fels, Antarctic ozone decreses: a dynamical cause?, Geophys. Res. Lett., 13, 1316-1319, 1986.
Mahlman, J. D., J. D. Pinto, and L. J. Umscheid, Transport radiative and dynamical effects of the Antarctic ozone hole: A GFDL "SKYHI" model experiment, J. Atmos. Sci., 51, 489-508, 1994.
Peixoto, J. P., and A. H. Oort, Physics of Climate, American Institute of Physics, New York, 1992.
Randel, W. J., and J. B. Cobb, Coherent variations of monthly mean total ozone and lower stratospheric temperature, J. Geophys. Res., 99, 5433-5447, 1994.
Rind, D., N. K. Balachandran, and R. Suozzo, Climate change and the middle atmosphere, part II: The impact of volcanic aerosols, J. Clim., 5, 189-208, 1992.
Rosen, R. D., and D. A. Salstein, Contribution of stratospheric winds to annual and semiannual fluctuations in atmospheric angular momentum and the length of day, J. Geophys. Res., 90, 8033-8041, 1985.
Rosen, R. D., D. A. Salstein, T. M. Eubanks, J. O. Dickey, and J. A. Steppe, An El Ni no signal in atmospheric angular momentum and Earth rotation, Science, 225, 411-414, 1984.
Stolarski, R. S., P. Bloomfield, R. D. McPeters, and J. R. Herman, Total ozone trends deduced from Nimbus 7 TOMS data, Geophys. Res. Lett., 18, 1015-1018, 1991.
Trenberth, K. E., and J. G. Olson, An evaluation and intercomparison of global analyses from the National Meteorological Center and the European Centre for Medium-Range Weather Forecasts, Bull. Am. Meteorol. Soc., 69, 1515-1524, 1988.
World Meteorological Organization, Global ozone research and monitoring project, Rep. 25, Scientific assessment of ozone depletion: 1991, UNEP, Geneva, Switzerland, 1991.
World Meteorological Organization, Global ozone research and monitoring project, Rep. 37, Scientific assessment of ozone depletion: 1994, UNEP, Geneva, Switzerland, 1994.
Yang, S., and W. J. Gutowski, GCM simulations of the three-dimensional propagation of stationary waves, J. Clim., 7, 414-433, 1994.