1. Introduction
The Northern Hemisphere winter conditions are significantly influenced by the North Atlantic Oscillation (NAO), defined as the sea level pressure gradient between the Icelandic minimum and the Azores maximum (Hurrell et al. 2003). The synchronous variations of the subtropical and subarctic atmospheric pressure were noticed long ago (Walker 1925), manifesting itself as a simultaneous strengthening of the subpolar low pressure and the subtropical high pressure (the positive NAO phase), or synchronous weakening of the pressure in both centers of action (the negative phase of NAO). This seesaw in the North Atlantic pressure variability is accompanied by corresponding changes in the winter storm-track over the North Atlantic and Europe. For example, the global thermally driven subtropical jet (i.e., a local maximum of the wind speed)—typical for the negative NAO phase—is bifurcated during the positive phase of NAO (Woollings et al. 2010, their Fig. 1). The appearance of the additional midlatitude jet (propagating from the North America toward the Scandinavian peninsula) is supposedly due to the interaction of the synoptic-scale departures from the mean flow’s strength and position (known as “transient baroclinic eddies”; Thompson et al. 2003) and the mean flow itself.
Supposing that the bifurcation of the subtropical jet (during the positive NAO phase) is forced by the eddy–mean flow interactions, we have to elucidate the factor(s) responsible for the variability of the amplitude of circulation eddies (anomalies). At synoptic time scale, the intensification or weakening of eddies is usually related to the strength of the thermally driven subtropical jet. For example, the stronger jet—with its vertical wind shear and meridional gradient of potential vorticity (limiting itself the eddy movement in meridional direction)—organizes the baroclinic eddies, ensuring their vertical propagation. The latter, in turn, reinforces the jet. Oppositely, the weaker jet allows self-organization of eddies in baroclinically unstable flow (poleward of the subtropical jet)—even in the absence of thermal forcing (Thompson et al. 2003). However, the drivers(s) affecting the subtropical jet’s strength are still unclear.
Despite the short life cycle of NAO atmospheric mode, it possesses distinct decadal variations (e.g., Pozo-Vázquez et al. 2001; Raible et al. 2005; Athanasiadis et al. 2020; Christiansen et al. 2022). Moreover, the proxy reconstructions of NAO reveal the existence of well pronounced trends, during the last few centuries (Appenzeller et al. 1998). On the other hand, the absence of atmospheric memory from one winter season to the next one, suggests that the long-term variability of NAO reflects either the greater heat capacity of the ocean (and, respectively, its long memory), or it is a projection of some external influences on the climatic system.
Some authors are tempted to explain the long-term NAO variability as an aggregated effect of the short-term stochastic variability of tropospheric circulation (e.g., Stephenson et al. 2000; Feldstein 2000). The other types of explanations relate the long-term NAO variability to (i) the other components of climatic system like the ocean, the cryosphere and the other atmospheric modes (e.g., Grötzner et al. 1998; Cassou 2008; Ineson and Scaife 2009; Cagnazzo and Manzini 2009; Petoukhov and Semenov 2010; Warner 2018; Årthun et al. 2021), or (ii) factors external to the climate system—for example, anthropogenic and volcanic forcing (Wallace and Thompson 2002; Shindell et al. 2003; Fischer et al. 2007), and solar variability (Kodera 2002; Ruzmaikin and Feynman 2002; Kirov and Georgieva 2002; Georgieva et al. 2007; Gray et al. 2016).
Among the external forcing, the increased concentration of greenhouse gases is the most frequently analyzed. However, the anthropogenic explanation of the NAO positive trend—observed during the second half of twentieth century—becomes very unlikely after the phase transitions of NAO latter on (see Fig. 1). Moreover, according to Christiansen et al. (2022) the modeling skill of historical ensembles including well mixed greenhouse gases or anthropogenic aerosols is not significantly different from zero. The effect of the other candidate—the volcanic forcing—is quite short lasting. For example, the effect of gasses, ash, and water vapor—tossed by the volcanic eruption in the atmosphere—becomes negligible after 2–3 years (Shindell et al. 2003).
Centennial evolution of the NAO index, derived by the University of East Anglia (https://crudata.uea.ac.uk/cru/data/nao/index.htm; Jones et al. 1997), compared with CO2 density.
Citation: Earth Interactions 27, 1; 10.1175/EI-D-23-0007.1
Centennial evolution of the NAO index, derived by the University of East Anglia (https://crudata.uea.ac.uk/cru/data/nao/index.htm; Jones et al. 1997), compared with CO2 density.
Citation: Earth Interactions 27, 1; 10.1175/EI-D-23-0007.1
Centennial evolution of the NAO index, derived by the University of East Anglia (https://crudata.uea.ac.uk/cru/data/nao/index.htm; Jones et al. 1997), compared with CO2 density.
Citation: Earth Interactions 27, 1; 10.1175/EI-D-23-0007.1
With regard to the other external forcing—that is, the solar activity—there is large disagreement in the scientific community. This controversy is rooted in the nonstationary temporal synchronization of NAO and solar variability—being stronger in periods of increased solar activity (e.g., Maruyama et al. 2018; Drews et al. 2022). Moreover, the correlation between NAO and sunspots changes its sign alternatively, with a periodicity of approximately 50 years (Gruzdev and Bezverkhnii 2020). This situation motivates Thiéblemont et al. (2015) to conclude that 11-yr solar cycle simply synchronizes the internally inherent quasi-decadal NAO variability.
Some authors interpret the NAO variability as a remnant of breaking synoptic waves (during their interaction with the three-dimensional winter flow)—evolving in NAO like anomalies (Franzke et al. 2004). The wave breaking critically depends on the strength of the mean flow (Thompson et al. 2003), and accordingly on the sea surface temperature gradients. Consequently, one might expect that the spatial–temporal evolution of the near-surface temperature and pressure should be synchronized—at least to some extent. Analysis of instantaneous correlation map of the air temperature at 2 m above the surface and the sea level pressure reveals, however, that this is not the case (see Fig. 2).
Instantaneous cross-correlation map of air surface temperature and sea level pressure (unfiltered winter values), calculated for the period 1900–2019. Correlation coefficients higher than 0.195 are statistically significant, according to the two-sided Student’s t test.
Citation: Earth Interactions 27, 1; 10.1175/EI-D-23-0007.1
Instantaneous cross-correlation map of air surface temperature and sea level pressure (unfiltered winter values), calculated for the period 1900–2019. Correlation coefficients higher than 0.195 are statistically significant, according to the two-sided Student’s t test.
Citation: Earth Interactions 27, 1; 10.1175/EI-D-23-0007.1
Instantaneous cross-correlation map of air surface temperature and sea level pressure (unfiltered winter values), calculated for the period 1900–2019. Correlation coefficients higher than 0.195 are statistically significant, according to the two-sided Student’s t test.
Citation: Earth Interactions 27, 1; 10.1175/EI-D-23-0007.1
Figure 2 shows clearly that the pressure–temperature temporal synchronization varies over the globe not only by strength, but also by sign. Note that the in-phase covariance of temperature and pressure is found over the North Atlantic Ocean, as well as in a latitudinal belt—roughly corresponding to the subtropical atmospheric jet. Over the rest of the world, the temperature and pressure covariate in antiphase. This means that at longer time scales the troposphere could not be treated as an ideal gas and the diabatic heating has an important impact on the atmospheric thermodynamics [refer to Eq. (1)]. This could be an explanation for the problems of the contemporary climatic models to reproduce adequately the NAO long-term variability (Simpson et al. 2018; Blackport and Fyfe 2022; Christiansen et al. 2022).
On the other hand, the climate sensitivity to the near-tropopause ozone variations has been established long ago, and by many authors (e.g., Manabe and Wetherald 1967; Ramanathan et al. 1976; Forster and Shine 1997; Stuber et al. 2001; Kilifarska 2012; Kilifarska et al. 2020). Numerical experiments (Stuber et al. 2001; Kilifarska et al. 2018) confirm that models’ climate is indeed sensitive to the near-tropopause ozone variation. However, the climate models fail to reproduce the real spatial–temporal variation of the lower-stratospheric ozone, due to the nonrecognition of the second source of ozone at these levels (Kilifarska 2013). The recent models are based on the assumption that the lower-stratospheric ozone density is controlled solely by stratospheric circulation.
The ozone production in the lower stratosphere is activated by the low energy electrons in the Regener–Pfotzer ionization maximum (Kilifarska 2013). The heterogeneous spatial distribution of the latter is projected on the lower-stratospheric ozone density (Kilifarska et al. 2020). Combined with the strong ozone influence on the near-surface climate, energetic particles in the Regener–Pfotzer maximum could be a reasonable explanation of the regional specificity of climate variations.
The present study is focused on the analyses of coupling between the Northern Hemisphere air surface temperature and sea level pressure, particularly over the North Atlantic Ocean—the region of formation of NAO atmospheric mode. The role of the lower-stratospheric ozone variability and its relation to the spatial–temporal synchronicity of atmospheric pressure and temperature is also studied. We have found that the lower-stratospheric ozone influences the sea level pressure mainly in a latitudinal belt characterized by a centennially lower ozone mixing ratio at 70 hPa. The strongest winter ozone impact coincides with the two active circulation centers in the Northern Hemisphere—that is, the Aleutian low and region of NAO formation.
2. Data and methods
a. Data
Monthly mean data of air temperature at 2 m above the surface, ozone mixing ratio at 70 hPa and sea level pressure are taken from the merged first ECMWF reanalysis of the 20th century (ERA-20C) and ERA-Interim, covering the period 1900–2019. Both reanalyses were merged at the year 2000. The merging procedure includes an equalization of the decadal means of both records, taken over the period 2001–10—smoothing in such a way the transition between the two reanalyses, and avoiding possible steplike changes between their means. Monthly records of all atmospheric variables have been derived in a grid with 5° step in latitude and longitude. The wider winter seasonal means (covering the period November–April) have been used in our analysis. Our attention was focused on the multiannual variability of atmospheric variables, so the fluctuations shorter than 5 years have been suppressed by applying 5-point moving average procedure.
Long record of galactic cosmic rays’ data (from 1700 to 1951) is provided by the World Data Centre for Paleoclimatology located in Boulder, Colorado, and the NOAA Paleoclimatology Program. After 1951, the record was extended by calibrated data from the Moscow neutron monitor. The 11-yr solar cycle, existing in the data, has been suppressed by 11-yr running average procedure.
b. Statistical methods
We have also used the lagged cross-correlation analysis to study the coherence in temporal variability of galactic cosmic rays’ intensity and ozone mixing ratio, as well as the covariance between near-surface pressure and temperature. This classical approach has been chosen, because it provides an assessment of the time delay of responding variable to the applied forcing. The cross-correlation coefficient is calculated by shifting the independent variable m steps back in time, which determines the negative sign of the responding variable’s time lag.
Although the applied software provides an estimation of statistical significance of correlation coefficients, we cannot rely on the standard Student’s t test, because our smoothed time series are serially correlated. This is due to the fact that a standard test of significance could inflate the calculated Z values, providing false positives. For this reason, we have used the method suggested by Afyouni et al. (2019), based on the recalculation of the correlation coefficients’ variance, taking into account not only the autocorrelation coefficients of both time series but also their cross correlation. Other available approaches, providing a solution of the problem of serially correlated records, are based on recalculation of effective degree of freedom, and the assumption that analyzed time series do not correlate. This requirement is not met by our time series, because they covariate in time. Moreover, our main purpose is to estimate the strength and reliability of their cross correlation.
3. Results
a. Temperature–pressure covariance in the Northern Hemisphere
The map of simultaneous correlation between near-surface temperature and pressure illustrates that a strong temperature–pressure coupling is found only in limited regions of the Northern Hemisphere (see Fig. 2). Having in mind that causality is of greatest importance for detection and attribution analyses, we have performed a two-way lagged correlation—changing consequently the leading factor (temperature or pressure). The derived correlation maps are shown in Fig. 3. The top-left panel of Fig. 3 illustrates the temperature impact on the sea level pressure, while the right one—the pressure effect on the air surface temperature. The lower panels present the time lags of the response to the applied forcing—that is, that of the sea level pressure (left) and the near-surface air temperature (right).
Lagged correlation between air temperature at 2 m above the surface (T2m) and the sea level pressure (SP), calculated for the period 1900–2019: (top left) the T2m impact on the pressure; (top right) the pressure influence on the temperature; (bottom left) the time delay of SP response to the temperature forcing; (bottom right) the time lag of T2m response to the pressure forcing.
Citation: Earth Interactions 27, 1; 10.1175/EI-D-23-0007.1
Lagged correlation between air temperature at 2 m above the surface (T2m) and the sea level pressure (SP), calculated for the period 1900–2019: (top left) the T2m impact on the pressure; (top right) the pressure influence on the temperature; (bottom left) the time delay of SP response to the temperature forcing; (bottom right) the time lag of T2m response to the pressure forcing.
Citation: Earth Interactions 27, 1; 10.1175/EI-D-23-0007.1
Lagged correlation between air temperature at 2 m above the surface (T2m) and the sea level pressure (SP), calculated for the period 1900–2019: (top left) the T2m impact on the pressure; (top right) the pressure influence on the temperature; (bottom left) the time delay of SP response to the temperature forcing; (bottom right) the time lag of T2m response to the pressure forcing.
Citation: Earth Interactions 27, 1; 10.1175/EI-D-23-0007.1
There are several findings, following from Fig. 3, that should be noted: (i) the long-term pressure–temperature coupling is dominantly in antiphase (oppositely to the expected from the ideal gas law); (ii) at higher latitudes, the pressure impact on the air temperature is stronger and covers wider regions in the Northern Hemisphere (although temperature response is delayed by 7–13 years); (iii) at tropical latitudes the temperature influence on the sea level pressure is slightly better pronounced, and pressure response is delayed by 4–7 years; (iv) over the North Atlantic region the lagged pressure–temperature correlation is very poor.
Comparison of Figs. 2 and 3 shows that pressure–temperature variations over the North Atlantic region evolve more or less independently in time. This is well illustrated in the Fig. 4, presenting the evolution of both variables at two main centers of NAO mode—that is, Ponta Delgada (Azores) and Reykjavik (Iceland). The short-time fluctuations (i.e., shorter than 5 years) are dumped by smoothing the raw data with 5-point moving window.
Time series of sea level pressure and air temperature measured in (a) Ponta Delgada and (b) Reykjavik, smoothed by a 5-point moving window.
Citation: Earth Interactions 27, 1; 10.1175/EI-D-23-0007.1
Time series of sea level pressure and air temperature measured in (a) Ponta Delgada and (b) Reykjavik, smoothed by a 5-point moving window.
Citation: Earth Interactions 27, 1; 10.1175/EI-D-23-0007.1
Time series of sea level pressure and air temperature measured in (a) Ponta Delgada and (b) Reykjavik, smoothed by a 5-point moving window.
Citation: Earth Interactions 27, 1; 10.1175/EI-D-23-0007.1
Conclusively, the long-term synchronization between subtropical and subarctic pressure variability (determining NAO climatic mode) could not be attributed to corresponding changes of the air surface temperature. This implies that the long-term evolution of both variables is determined either by different factors, or at least by different influential mechanisms of the same forcing.
b. Statistical evidence for lower-stratospheric ozone influence on the near-surface pressure
Prediction of the North Atlantic pressure evolution is not trivial, because recently it becomes clear that positive trend in global temperature is not simply projected on the sea level pressure. So, the expectations for positive and increasing NAO pattern during twenty-first century (Hurrell at al. 2003) remained elusive. On the other hand, the recent discovery of NAO sensitivity to the lower-stratospheric ozone variations (Velichkova and Kilifarska 2019) gives us a motivation for further analysis of the statistical relation between sea level pressure and ozone near the tropopause.
Prior studies reveal that ozone density at 70 hPa is significantly influenced by the energetic particles in the lower-atmospheric ionization layer, known as the Regener–Pfotzer maximum (Kilifarska et al. 2022). Beyond the auroral oval, the Regener–Pfotzer maximum is determined by the secondary ionization and the products of nuclear decay, resulting from the interaction of primary cosmic rays with the atmospheric atoms and molecules. At lower latitudes, however, the Regener–Pfotzer maximum is filled mainly by energetic particles trapped in Earth’s radiation belts. In regions with weaker geomagnetic field or with higher azimuthal magnetic gradient, these particles are lost in the atmosphere, where they contribute to the ionization in the Regener–Pfotzer maximum. At middle latitudes, this low-atmospheric ionization layer is placed above the tropopause, where the atmospheric humidity is severely reduced—due to the freeze drying process near the tropopause (the tropospheric cold point). Thus, the availability of low energy electrons in a dry lower stratosphere activates the process of ozone production at these levels (Kilifarska 2013). Consequently, the heterogeneously distributed humidity, and low atmospheric ionization, over the globe is the main reason for irregular production of ozone in the lower stratosphere. This is the physically based explanation of the hemispherical asymmetry and longitudinal variations of the lower-stratospheric ozone density (Kilifarska 2017), reported from many authors (Hood and Zaff 1995; Pan et al. 1997; Steblova 2001; Peters et al. 2008).
Examination of the ozone and pressure time series (with suppressed subdecadal periodicities) reveals that the synchronized antiphase pressure variations in Azores and Iceland are accompanied by corresponding opposite changes of ozone in the lower stratosphere (see Fig. 5). Interestingly, the strongest decrease of pressure in Ponta Delgada corresponds to the highest values of ozone at 70 hPa, during the examined period. Oppositely, the highest pressure values in Reykjavik correspond to the lowest ozone density during the period 1925–75. Calculated cross-correlation coefficients, presented in Table 1, confirm the good synchronicity of these variables in both cases—with suppressed variations shorter than 5 years, and shorter than 11 years.
Time series of ozone at 70 hPa (black or blue curve) and sea level pressure (red curves), smoothed by an 11-point moving window, shown for (a) Ponta Delgada and (b) Reykjavik. Note that at both sites the ozone abundance is accompanied by reduction of the sea level pressure.
Citation: Earth Interactions 27, 1; 10.1175/EI-D-23-0007.1
Time series of ozone at 70 hPa (black or blue curve) and sea level pressure (red curves), smoothed by an 11-point moving window, shown for (a) Ponta Delgada and (b) Reykjavik. Note that at both sites the ozone abundance is accompanied by reduction of the sea level pressure.
Citation: Earth Interactions 27, 1; 10.1175/EI-D-23-0007.1
Time series of ozone at 70 hPa (black or blue curve) and sea level pressure (red curves), smoothed by an 11-point moving window, shown for (a) Ponta Delgada and (b) Reykjavik. Note that at both sites the ozone abundance is accompanied by reduction of the sea level pressure.
Citation: Earth Interactions 27, 1; 10.1175/EI-D-23-0007.1
Simultaneous cross-correlation coefficients of ozone at 70 hPa and sea level pressure, calculated for the period 1900–2019, at Reykjavik (Iceland) and Ponta Delgada (Azores). The center column shows the coherence in temporal variability of time series smoothed by a 5-point moving window. The correlation coefficients in the right column are derived for the time series with suppressed subdecadal variations of analyzed variables. All correlations are statistically significant at confidence level α = 0.05.
Having in mind the nonlinear character of climatic time records, and deficiencies of linear statistics, furthermore we have applied the artificial neural network technique (regression problem). The spatial distribution of the ozone–pressure temporal synchronization has been estimated by mapping ozone temporal variations on the atmospheric pressure, applying the neural network in each node of our grid with resolution 5° in latitude and longitude. The map presented in Fig. 6 shows the correlation coefficients between observed and modeled values of the sea level pressure, as a function of the lower-stratospheric ozone. The figure shows that the strongest ozone impact on the sea level pressure is found over the main “centers of action” of atmospheric circulation—that is, the Aleutian low pressure system, and the North Atlantic region between Azores and Iceland islands (see Fig. 6). This finding reveals that the near-surface pressure could be affected distantly by the variations in the lower-stratospheric ozone density. The mechanisms for ozone influence on the sea level pressure are discussed in the following subsection.
Spatial distribution of the lower-stratospheric ozone impact on the sea level pressure, estimated by the neural network technique, during the period 1900–2019.
Citation: Earth Interactions 27, 1; 10.1175/EI-D-23-0007.1
Spatial distribution of the lower-stratospheric ozone impact on the sea level pressure, estimated by the neural network technique, during the period 1900–2019.
Citation: Earth Interactions 27, 1; 10.1175/EI-D-23-0007.1
Spatial distribution of the lower-stratospheric ozone impact on the sea level pressure, estimated by the neural network technique, during the period 1900–2019.
Citation: Earth Interactions 27, 1; 10.1175/EI-D-23-0007.1
Meanwhile we have examined the spatial–temporal covariance between ozone at 70 hPa and the air surface temperature (T2m). The neural network technique has been applied again in each node of our grid, comparing the similarities in temporal evolution of ozone and temperature. The result presented in Fig. 7 illustrates the spatial distribution of correlation between observed and modeled temperature values, with ozone used as independent variable. The glance on the correlation map reveals that the strongest ozone–temperature coupling is detected at tropical regions of the Indo-Pacific Ocean and Indonesia. Comparison of Figs. 6 and 7 shows that the strongest ozone impact, on the sea level pressure, does not match the ozone influence on the air temperature.
Spatial distribution of the air temperature response to variations in the lower-stratospheric ozone, during the period 1900–2019, as determined by the artificial neural network technique.
Citation: Earth Interactions 27, 1; 10.1175/EI-D-23-0007.1
Spatial distribution of the air temperature response to variations in the lower-stratospheric ozone, during the period 1900–2019, as determined by the artificial neural network technique.
Citation: Earth Interactions 27, 1; 10.1175/EI-D-23-0007.1
Spatial distribution of the air temperature response to variations in the lower-stratospheric ozone, during the period 1900–2019, as determined by the artificial neural network technique.
Citation: Earth Interactions 27, 1; 10.1175/EI-D-23-0007.1
This result indicates that there are different mechanisms for the lower-stratospheric ozone influence on the near-surface pressure and temperature. More details about these mechanisms can be found in the next subsection.
c. Mechanisms of lower-stratospheric ozone influence on the near-surface temperature and pressure
The atmospheric ozone is a strong radiatively active gas, adsorbing in several spectral bands of electromagnetic radiation (Forster and Shine 1997). Consequently, changes in the lower-stratospheric ozone density impact the local stratospheric temperature, and as a result—the pressure at the certain levels (due to the changes in the diffusion velocity of atmospheric molecules). These pressure changes are detected also by barometers on the sea level because the atmospheric pressure is an integral characteristic of the weight of the whole atmospheric column above the sea surface. Consequently, the changes in the lower-stratospheric ozone cold have a direct impact on the sea level pressure through changing temperature and diffusivity at the levels of ozone changes.
On the other hand, the Eq. (2) illustrates that the near-surface atmospheric pressure depends also on the near-surface temperature. The latter, in turn, depends on many factors, one of which is greenhouse warming. It is well known that the strongest greenhouse gas is atmospheric water vapor. However, despite the greatest amount of water vapor in the lower troposphere its impact in the greenhouse effect is only 10% (Inamdar et al. 2004). The remaining 90% of the greenhouse effect belongs to the upper-tropospheric water vapor, regardless of its tiny concentration (Sinha and Harries 1995; Spencer and Braswell 1997). Consequently, the understanding of the near-surface temperature variability goes through understanding of factor(s) and mechanism(s) controlling the spatial–temporal variability of the upper-tropospheric water vapor.
As mentioned above, the changes in the lower-stratospheric ozone density affect the near-tropopause temperature, and consequently—the upper-tropospheric static stability (North and Eruhimova 2009; Young 2003). Thus, the abundance of ozone warms the tropopause, making the upper troposphere more stable, followed by its gradual drying—due to the suppressed upward propagation of moisture from the lower-atmospheric layers. The greenhouse power of the depleted water vapor is significantly weakened, and the near-surface temperature cooled. Oppositely, the reduced amount of the lower-stratospheric ozone cools the tropopause, making the upper troposphere more unstable. Consequently, more water vapor is lifted up in the upper troposphere, and the greenhouse warming of Earth’s surface becomes stronger. In adiabatic systems the pressure follows temperature changes and consequently this is an indirect mechanism for ozone influence on the near-surface pressure—that is, through modulation of the temperature field. A schematic illustration of lower-stratospheric ozone influence on the near-surface temperature and pressure is shown in Fig. 8.
Block scheme of the various mechanisms of lower-stratospheric ozone influence on the near-surface temperature and pressure.
Citation: Earth Interactions 27, 1; 10.1175/EI-D-23-0007.1
Block scheme of the various mechanisms of lower-stratospheric ozone influence on the near-surface temperature and pressure.
Citation: Earth Interactions 27, 1; 10.1175/EI-D-23-0007.1
Block scheme of the various mechanisms of lower-stratospheric ozone influence on the near-surface temperature and pressure.
Citation: Earth Interactions 27, 1; 10.1175/EI-D-23-0007.1
d. Factors influencing the spatial–temporal variability of the lower-stratospheric ozone
In the lower stratosphere, a direct photochemical production of ozone (i.e., O2 + O + M → O3 + M) is impossible, because the solar UV radiation reaching to the lower stratosphere, cannot dissociate any more the molecular oxygen. This extends the lifetime of the lower-stratospheric ozone up to 3–4 years (Brasseur and Solomon 2005), suggesting that the spatial distribution of the lower-stratospheric ozone is controlled mainly by the stratospheric Brewer–Dobson circulation. However, recent multimodeling experiments show that the amount of extratropical ozone is determined mainly by the local production, and that the impact of the tropical ozone, transported by the Brewer–Dobson circulation, is no more than 30% (Grewe 2006). This result immediately raises the following question: What is the source of ozone production at these latitudes and altitudes?
An answer could be found in Kilifarska (2013), presenting evidence for the existence of a new ozone source in the lower stratosphere. This conclusion was completely unexpected, because to that moment the energetic particles have been considered only as destroyers of atmospheric ozone (e.g., Jackman et al. 1980; Rohen et al. 2005). The author reveals that the energy of the secondary electrons in the Regener–Pfotzer maximum is enough to produce the short lived tetraoxygen ion
The validity of this concept could be checked by analysis of spatial–temporal coherence of energetic particles flux entering Earth’s atmosphere, and ozone density. Attempting to resolve the problem of causality, the lagged correlation analysis has been applied, which determines the time delay of the linear ozone response to particles’ forcing. The lagged correlation coefficients have been calculated in each node of our grid (with a resolution of 5° in latitude and longitude). The coherence in ozone-particles’ evolution has been examined comparing the long-term variability of galactic cosmic rays (GCR) reaching the planetary surface (with 11-yr solar cycle suppressed) and ozone mixing ratio at 70 hPa (with periodicities shorter than 5 years eliminated). Results are shown in Fig. 9, illustrating the spatial heterogeneity of GCR impact on the ozone. Note the particle–ozone correlation varies not only in strength, but also in sign. These peculiarities should be attributed to the geomagnetic lensing of trapped energetic particles (in the lowest part of their trajectories) by heterogeneous geomagnetic field (Kilifarska et al. 2022).
(top) Lagged correlation map of galactic cosmic rays and ozone mixing ratio at 70 hPa, calculated for the period 1900–2019. Values greater than 0.37 are statistically significant at α = 0.05 level (colored shading). (bottom) Time delay of ozone response to GCR forcing in years.
Citation: Earth Interactions 27, 1; 10.1175/EI-D-23-0007.1
(top) Lagged correlation map of galactic cosmic rays and ozone mixing ratio at 70 hPa, calculated for the period 1900–2019. Values greater than 0.37 are statistically significant at α = 0.05 level (colored shading). (bottom) Time delay of ozone response to GCR forcing in years.
Citation: Earth Interactions 27, 1; 10.1175/EI-D-23-0007.1
(top) Lagged correlation map of galactic cosmic rays and ozone mixing ratio at 70 hPa, calculated for the period 1900–2019. Values greater than 0.37 are statistically significant at α = 0.05 level (colored shading). (bottom) Time delay of ozone response to GCR forcing in years.
Citation: Earth Interactions 27, 1; 10.1175/EI-D-23-0007.1
Comparison of Figs. 6 and 9 reveals that the strongest direct ozone impact on the sea level pressure corresponds to the regions with a negative GCR–ozone correlation. Moreover, the evolution of winter ozone at 70 hPa indicates that ozone mixing ratio over latitudinal belt 45°–75°N has been significantly lower during the twentieth century relative to the first decade of the twenty-first century (see Fig. 10). This could be a reasonable explanation for the positive NAO trend detected in the second half of the twentieth century.
Ozone temporal evolution at 70 hPa, during the twentieth century, presented as a difference between the corresponding decadal mean and the first decade of the twenty-first-century mean.
Citation: Earth Interactions 27, 1; 10.1175/EI-D-23-0007.1
Ozone temporal evolution at 70 hPa, during the twentieth century, presented as a difference between the corresponding decadal mean and the first decade of the twenty-first-century mean.
Citation: Earth Interactions 27, 1; 10.1175/EI-D-23-0007.1
Ozone temporal evolution at 70 hPa, during the twentieth century, presented as a difference between the corresponding decadal mean and the first decade of the twenty-first-century mean.
Citation: Earth Interactions 27, 1; 10.1175/EI-D-23-0007.1
4. Discussion
The results presented in this study give a new perspective on the question: Why the Aleutian low and the region of the NAO formation act as centers determining the circulation variability in the Northern Hemisphere?
We found that besides the equator to pole temperature gradient, the circulation pattern could be affected also by pressure variations, imposed from the lower-stratospheric ozone—especially in latitudinal belt 45°–75°N (refer to Figs. 6 and 9). Usually, the appearance of circulation anomalies is thought to be a stochastic process, related to the atmospheric/flow instability. This instability, however, could have its real physical forcing, that is, the pressure fluctuations, induced by the variations of the lower-stratospheric ozone density. Being a radiatively active gas, ozone abundance or depletion directly affects the sea level pressure, through local changes of the lower-stratospheric temperature and, respectively, pressure. Figure 6 shows that the winter ozone’s impact on the sea level pressure is strongest in the regions of Bering Sea and Aleutian Islands, as well as in the North Atlantic—among Azores and Iceland islands—that is, just over the main “centers of action.”
One reasonable question (related to the establishment of causality chain) is about the factor(s) determining the spatial–temporal variability of the lower-stratospheric ozone. Our previous investigations uncover that it could be attributed to the variations of Earth’s radiation belts, feeding the lower-atmospheric ionization in the Regener–Pfotzer maximum (Kilifarska et al. 2020, 2022). The number of particles trapped in the Van Allen radiation belts is modulated by the solar variability and galactic cosmic ray fluxes, reaching Earth’s magnetosphere (refer to a schematic diagram in Fig. 11).
Schematic diagram of solar and geomagnetic modulation of cosmic rays (reaching Earth’s atmosphere), and their role for ozone production in the lower stratosphere, which in turn results in regional specificity of climate variations.
Citation: Earth Interactions 27, 1; 10.1175/EI-D-23-0007.1
Schematic diagram of solar and geomagnetic modulation of cosmic rays (reaching Earth’s atmosphere), and their role for ozone production in the lower stratosphere, which in turn results in regional specificity of climate variations.
Citation: Earth Interactions 27, 1; 10.1175/EI-D-23-0007.1
Schematic diagram of solar and geomagnetic modulation of cosmic rays (reaching Earth’s atmosphere), and their role for ozone production in the lower stratosphere, which in turn results in regional specificity of climate variations.
Citation: Earth Interactions 27, 1; 10.1175/EI-D-23-0007.1
The latitudinal belt with reduced lower-stratospheric ozone (within most of the twentieth century) is under the influence of energetic particles trapped in the Van Allen radiation belts (see Figs. 9 and 10). The latter are fed primarily by galactic cosmic rays, the intensity of which gradually decreases—up to the end of the twentieth century (a modulation effect of the more active sun; Rouillard and Lockwood 2007). In agreement with the concept of the secondary source of ozone in the lower stratosphere (Kilifarska 2013), the ozone mixing ratio at 70 hPa is significantly lower than its values during the first decade of twenty-first century. The enhanced ozone density in the latter period is reasonably attributed to the severely reduced solar activity during the 24th solar cycle, and corresponding rise of galactic cosmic ray flux.
The coupling between lower-stratospheric ozone and atmospheric pressure and temperature could be traced at different time scales. This article shows evidence for their covariance at multidecadal time scales. A recent study illustrates, however, that such a coupling is well visible at daily and weekly time scales (Kilifarska and Peqini 2023). These authors show that short lasting reduction of cosmic rays’ intensity is followed by well detectable depletion of near-tropopause ozone density. The ozone response is regionally specified because of geomagnetic spatial irregularities, modulating the access of energetic particles into the lower atmosphere. This finding suggests that the stochastic nature of atmospheric circulation could be attributed—at least to some extent—to the spatial–temporal variability of the near-tropopause ozone, due to its influence on the near-surface pressure and temperature.
5. Conclusions
This paper shows statistical evidence for synchronization of the long-term variations of winter ozone at 70 hPa, sea level pressure, and near-surface air temperature in the Northern Hemisphere. The different patterns of ozone influence on the pressure and temperature are an indication for different mechanisms of ozone effect. For example, the ozone impact on the sea level pressure could be direct (through local changes of the lower-stratospheric temperature and pressure) or indirect—through changes of the near-surface temperature, which in turn affects the pressure—in accordance with the energy conservation law.
The mechanism of distant lower-stratospheric ozone effect on the near-surface temperature consists of (i) influence on the near-tropopause temperature, (ii) alteration of the upper-tropospheric static stability, followed by moistening or drying of the upper troposphere (Kilifarska et al. 2020); (iii) strengthening or weakening of the water vapor impact in the atmospheric greenhouse effect (remind that the upper-tropospheric water vapor ensures 90% of the impact of the whole atmospheric water vapor; Inamdar et al. 2004).
The regions of the strongest ozone influence on the winter sea level pressure correspond to the atmospheric “centers of action”—that is, the Aleutian low and the region of formation of North Atlantic Oscillation (NAO). These regions coincide also with the strongest impact of energetic particles trapped in Earth’s radiation belts on the lower-stratospheric ozone density. This conjunction we interpret as a confirmation of our mechanism for energetic particles’ influence on the regional specificity of climate variability.
Acknowledgments.
The authors are thankful to the project teams of ECMWF’s first atmospheric reanalysis of the 20th Century and the ERA-Interim reanalysis, providing gridded data for ozone, temperature, and pressure. We are also grateful to the World Data Center for Paleoclimatology, Data Contribution Series 2008-013, NOAA/NCDC Paleoclimatology Program, Boulder, Colorado, as well as to the IZMIRAN neutron monitors data base (http://cr0.izmiran.ru/common/links.htm) for providing data for galactic cosmic ray intensity. This research was funded by National Science Fund of Bulgaria Contracts KP-06-N34/1/30-09-2020, and KP-06-M54/1/15-11-2021.
Data availability statement.
Access to ERA-Interim data was deactivated on 1 June 2023, but ECMWF provides an updated ERA5 (https://rda.ucar.edu/datasets/ds633.0/), from which atmospheric ozone, pressure and temperature can be retrieved (registration is required). Data for cosmic ray flux, measured at the ground surface, are freely available online (https://www.nmdb.eu/nest/ and http://cr0.izmiran.ru/common/links.htm). Graphs of NAO with CO2, and ozone with sea level pressure and temperature, were created with the STATISTICA commercial software (StatSoft’s license). All maps were created with the SURFER program, the license of which is held by Golden Software. Schematic diagrams were created with the MS Word software.
REFERENCES
Afyouni, S., S. M. Smith, and T. E. Nichols, 2019: Effective degrees of freedom of the Pearson’s correlation coefficient under autocorrelation. Neuroimage, 199, 609–625, https://doi.org/10.1016/j.neuroimage.2019.05.011.
Appenzeller, C., T. F. Stocker, and M. Anklin, 1998: North Atlantic Oscillation dynamics recorded in Greenland Ice cores. Science, 282, 446–449, https://doi.org/10.1126/science.282.5388.446.
Årthun, M., R. C. J. Wills, H. L. Johnson, L. Chafik, and H. R. Langehaug, 2021: Mechanisms of decadal North Atlantic climate variability and implications for the recent cold anomaly. J. Climate, 34, 3421–3439, https://doi.org/10.1175/JCLI-D-20-0464.1.
Athanasiadis, P. J., S. Yeager, Y.-O. Kwon, A. Bellucci, D. W. Smith, and S. Tibaldi, 2020: Decadal predictability of North Atlantic blocking and the NAO. npj Climate Atmos. Sci., 3, 20, https://doi.org/10.1038/s41612-020-0120-6.
Bishop, C. M., 1994: Neural networks and their applications. Rev. Sci. Instrum., 65, 1803–1832, https://doi.org/10.1063/1.1144830.
Blackport, R., and J. C. Fyfe, 2022: Climate models fail to capture strengthening wintertime North Atlantic jet and impacts on Europe. Sci. Adv., 8, eabn3112, https://doi.org/10.1126/sciadv.abn3112.
Brasseur, G. P., and S. Solomon, 2005: Aeronomy of the Middle Atmosphere: Chemistry and Physics of the Stratosphere and Mesosphere. 3rd ed. Atmospheric and Oceanographic Sciences Library, Vol. 32, Springer, 644 pp.
Cagnazzo, C., and E. Manzini, 2009: Impact of the stratosphere on the winter tropospheric teleconnections between ENSO and the North Atlantic and European region. J. Climate, 22, 1223–1238, https://doi.org/10.1175/2008JCLI2549.1.
Cassou, C., 2008: Intraseasonal interaction between the Madden–Julian Oscillation and the North Atlantic Oscillation. Nature, 455, 523–527, https://doi.org/10.1038/nature07286.
Christiansen, B., S. Yang, and D. Matte, 2022: The forced response and decadal predictability of the North Atlantic Oscillation: Nonstationary and fragile skills. J. Climate, 35, 5869–5882, https://doi.org/10.1175/JCLI-D-21-0807.1.
Drews, A., W. Huo, K. Matthes, K. Kodera, and T. Kruschke, 2022: The Sun’s role in decadal climate predictability in the North Atlantic. Atmos. Chem. Phys., 22, 7893–7904, https://doi.org/10.5194/acp-22-7893-2022.
Feldstein, S. B., 2000: The timescale, power spectra, and climate noise properties of teleconnection patterns. J. Climate, 13, 4430–4440, https://doi.org/10.1175/1520-0442(2000)013<4430:TTPSAC>2.0.CO;2.
Fischer, E. M., J. Luterbacher, E. Zorita, S. F. B. Tett, C. Casty, and H. Wanner, 2007: European climate response to tropical volcanic eruptions over the last half millennium. Geophys. Res. Lett., 34, L05707, https://doi.org/10.1029/2006GL027992.
Forster, P. M. F., and K. P. Shine, 1997: Radiative forcing and temperature trends from stratospheric ozone changes. J. Geophys. Res., 102, 10 841–10 855, https://doi.org/10.1029/96JD03510.
Franzke, C., S. Lee, and S. B. Feldstein, 2004: Is the North Atlantic Oscillation a breaking wave? J. Atmos. Sci., 61, 145–160, https://doi.org/10.1175/1520-0469(2004)061<0145:ITNAOA>2.0.CO;2.
Georgieva, K., B. Kirov, P. Tonev, V. Guineva, and D. Atanasov, 2007: Long-term variations in the correlation between NAO and solar activity: The importance of north–south solar activity asymmetry for atmospheric circulation. Adv. Space Res., 40, 1152–1166, https://doi.org/10.1016/j.asr.2007.02.091.
Gray, L. J., T. J. Woollings, M. Andrews, and J. Knightc, 2016: Eleven-year solar cycle signal in the NAO and Atlantic/European blocking. Quart. J. Roy. Meteor. Soc., 142, 1890–1903, https://doi.org/10.1002/qj.2782.
Grewe, V., 2006: The origin of ozone. Atmos. Chem. Phys., 6, 1495–1511, https://doi.org/10.5194/acp-6-1495-2006.
Grötzner, A., M. Latif, and T. P. Barnett, 1998: A decadal climate cycle in the North Atlantic Ocean as simulated by the ECHO coupled GCM. J. Climate, 11, 831–847, https://doi.org/10.1175/1520-0442(1998)011<0831:ADCCIT>2.0.CO;2.
Gruzdev, A. N., and V. A. Bezverkhnii, 2020: Manifestation of the 11-year solar cycle in the North Atlantic climate. IOP Conf. Ser.: Earth Environ. Sci., 606, 012018, https://doi.org/10.1088/1755-1315/606/1/012018.
Hood, L. L., and D. A. Zaff, 1995: Lower stratospheric stationary waves and the longitude dependence of ozone trends in winter. J. Geophys. Res., 100, 25 791–25 800, https://doi.org/10.1029/95JD01943.
Hurrell, J. W., Y. Kushnir, G. Ottersen, and M. Visbeck, 2003: An overview of the North Atlantic Oscillation. The North Atlantic Oscillation: Climate Significance and Environmental Impact, Geophys. Monogr., Vol. 134, Amer. Geophys. Union, 1–35, https://doi.org/10.1029/134GM01.
Inamdar, A. K., V. Ramanathan, and N. G. Loeb, 2004: Satellite observations of the water vapor greenhouse effect and column longwave cooling rates: Relative roles of the continuum and vibration-rotation to pure rotation bands. J. Geophys. Res., 109, D06104, https://doi.org/10.1029/2003JD003980.
Ineson, S., and A. A. Scaife, 2009: The role of the stratosphere in the European climate response to El Nino. Nat. Geosci., 2, 32–36, https://doi.org/10.1038/ngeo381.
Jackman, C. H., J. E. Frederick, and R. S. Stolarski, 1980: Production of odd nitrogen in the stratosphere and mesosphere: An intercomparison of source strengths. J. Geophys. Res., 85, 7495–7505, https://doi.org/10.1029/JC085iC12p07495.
Jones, P. D., T. Jónsson, and D. Wheeler, 1997: Extension to the North Atlantic Oscillation using early instrumental pressure observations from Gibraltar and south-west Iceland. Int. J. Climatol., 17, 1433–1450, https://doi.org/10.1002/(SICI)1097-0088(19971115)17:13<1433::AID-JOC203>3.0.CO;2-P.
Kilifarska, N. A., 2012: Mechanism of lower stratospheric ozone influence on climate. Int. Rev. Physiol., 6, 279–290.
Kilifarska, N. A., 2013: An autocatalytic cycle for ozone production in the lower stratosphere initiated by galactic cosmic rays. C. R. Acad. Bulg. Sci., 66, 243–252.
Kilifarska, N. A., 2017: Hemispherical asymmetry of the lower stratospheric O3 response to galactic cosmic rays forcing. ACS Earth Space Chem., 1, 80–88, https://doi.org/10.1021/acsearthspacechem.6b00009.
Kilifarska, N. A., and K. Peqini, 2023: Impact of Forbush decreases and geomagnetic storms on the atmospheric ozone profiles. Earth Space Sci., 10, e2023EA002954, https://doi.org/10.1029/2023EA002954.
Kilifarska, N. A., T. Wang, K. Ganev, M. Xie, B. Zhuang, and S. Li, 2018: Decadal cooling of East Asia—The role of aerosols and near tropopause ozone forcing. C. R. Acad. Bulg. Sci., 71, 937–944, https://doi.org/10.7546/CRABS.2018.07.10.
Kilifarska, N. A., V. G. Bakhmutov, and G. V. Melnyk, 2020: The Hidden Link Between Earth’s Magnetic Field and Climate. 1st ed. Elsevier, 215 pp., https://doi.org/10.1016/C2018-0-01667-9.
Kilifarska, N. A., V. G. Bakhmutov, and G. V. Melnyk, 2022: Coupling between geomagnetic field and Earth’s climate system. Magnetosphere and Solar Winds, Humans and Communication, K. S. Essa, K. H. Mahmoud, and Y.-H. Chemin, Eds., IntechOpen, https://doi.org/10.5772/intechopen.103695.
Kirov, B., and K. Georgieva, 2002: Long-term variations and interrelations of ENSO, NAO and solar activity. Phys. Chem. Earth, 27, 441–448, https://doi.org/10.1016/S1474-7065(02)00024-4.
Kodera, K., 2002: Solar cycle modulation of the North Atlantic Oscillation: Implication in the spatial structure of the NAO. Geophys. Res. Lett., 29, 1218, https://doi.org/10.1029/2001GL014557.
Manabe, S., and R. T. Wetherald, 1967: Thermal equilibrium of the atmosphere with a given distribution of relative humidity. J. Atmos. Sci., 24, 241–259, https://doi.org/10.1175/1520-0469(1967)024<0241:TEOTAW>2.0.CO;2.
Maruyama, F., K. Kai, and H. Morimoto, 2018: Solar influences on the North Atlantic Oscillation by wavelet-based multifractal analysis. J. Geosci. Environ. Prot., 6, 133–150, https://doi.org/10.4236/gep.2018.68010.
North, G. R., and T. L. Eruhimova, 2009: Atmospheric Thermodynamics: Elementary Physics and Chemistry. Cambridge University Press, 267 pp.
Pan, L., S. Solomon, W. Randel, J.-F. Lamarque, P. Hess, J. Gille, E.-W. Chiou, and M. P. McCormick, 1997: Hemispheric asymmetries and seasonal variations of the lowermost stratospheric water vapor and ozone derived from SAGE II data. J. Geophys. Res., 102, 28 177–28 184, https://doi.org/10.1029/97JD02778.
Peters, D. H. W., and A. Gabriel, and G. Entzian, 2008: Longitude-dependent decadal ozone changes and ozone trends in boreal winter months during 1960–2000. Ann. Geophys., 26, 1275–1286, https://doi.org/10.5194/angeo-26-1275-2008.
Petoukhov, V., and V. A. Semenov, 2010: A link between reduced Barents-Kara sea ice and cold winter extremes over northern continents. J. Geophys Res., 115, D21111, https://doi.org/10.1029/2009JD013568.
Pozo-Vázquez, D., M. J. Esteban-Parra, F. S. Rodrigo, and Y. Castro-Díez, 2001: A study of NAO variability and its possible non-linear influences on European surface temperature. Climate Dyn., 17, 701–715, https://doi.org/10.1007/s003820000137.
Raible, C. C., T. F. Stocker, M. Yoshimori, M. Renold, U. Beyerle, C. Casty, and J. Luterbacher, 2005: Northern Hemispheric trends of pressure indices and atmospheric circulation patterns in observations, reconstructions, and coupled GCM simulations. J. Climate, 18, 3968–3982, https://doi.org/10.1175/JCLI3511.1.
Ramanathan, V., L. B. Callis, and R. E. Boughner, 1976: Sensitivity of surface temperature and atmospheric temperature to perturbations in the stratospheric concentration of ozone and nitrogen dioxide. J. Atmos. Sci., 33, 1092–1112, https://doi.org/10.1175/1520-0469(1976)033<1092:SOSTAA>2.0.CO;2.
Rohen, G., and Coauthors, 2005: Ozone depletion during the solar proton events of October/November 2003 as seen by SCIAMACHY. J. Geophys. Res., 110, A09S39, https://doi.org/10.1029/2004JA010984.
Rouillard, A. P., and M. Lockwood, 2007: Centennial changes in solar activity and the response of galactic cosmic rays. Adv. Space Res., 40, 1078–1086, https://doi.org/10.1016/j.asr.2007.02.096.
Ruzmaikin, A., and J. Feynman, 2002: Solar influence on a major mode of atmospheric variability. J. Geophys. Res., 107, 4209, https://doi.org/10.1029/2001JD001239.
Shindell, D. T., G. A. Schmidt, R. L. Miller, and M. E. Mann, 2003: Volcanic and solar forcing of climate change during the preindustrial era. J. Climate, 16, 4094–4107, https://doi.org/10.1175/1520-0442(2003)016<4094:VASFOC>2.0.CO;2.
Simpson, I. R., C. Deser, K. A. McKinnon, and E. A. Barnes, 2018: Modeled and observed multidecadal variability in the North Atlantic jet stream and its connection to sea surface temperatures. J. Climate, 31, 8313–8338, https://doi.org/10.1175/JCLI-D-18-0168.1.
Sinha, A., and J. E. Harries, 1995: Water vapour and greenhouse trapping: The role of far infrared absorption. Geophys. Res. Lett., 22, 2147–2150, https://doi.org/10.1029/95GL01891.
Spencer, R. W., and W. D. Braswell, 1997: How dry is the tropical free troposphere? Implications for global warming theory. Bull. Amer. Meteor. Soc., 78, 1097–1106, https://doi.org/10.1175/1520-0477(1997)078<1097:HDITTF>2.0.CO;2.
Steblova, R. C., 2001: North-south asymmetry of the ozone layer. Rep. Russ. Acad. Sci., 379, 675–679.
Stephenson, D. B., V. Pavan, and R. Bojariu, 2000: Is the North Atlantic Oscillation a random walk? Int. J. Climatol., 20, 1–18, https://doi.org/10.1002/(SICI)1097-0088(200001)20:1<1::AID-JOC456>3.0.CO;2-P.
Stuber, N., M. Ponater, and R. Sausen, 2001: Is the climate sensitivity to ozone perturbations enhanced by stratospheric water vapor feedback? Geophys. Res. Lett., 28, 2887–2890, https://doi.org/10.1029/2001GL013000.
Thiéblemont, R., K. Matthes, N.-E. Omrani, K. Kodera, and F. Hansen, 2015: Solar forcing synchronizes decadal North Atlantic climate variability. Nat. Commun., 6, 8268, https://doi.org/10.1038/ncomms9268.
Thompson, D. W. J., S. Lee, and M. P. Baldwin, 2003: Atmospheric processes governing the Northern Hemisphere Annular Mode/North Atlantic Oscillation. The North Atlantic Oscillation: Climate Significance and Environmental Impact, Geophys. Monogr., Vol. 134, Amer. Geophys. Union, 81–112, https://doi.org/10.1029/134GM05.
Velichkova, T., and N. Kilifarska, 2019: Lower stratospheric ozone’s influence on the NAO climatic mode. C. R. Acad. Bulg. Sci., 72, 219–225, https://doi.org/10.7546/CRABS.2019.02.11.
Walker, G. T., 1925: Correlation in seasonal variations of weather—A further study of world weather. Mon. Wea. Rev., 53, 252–254, https://doi.org/10.1175/1520-0493(1925)53<252:CISVOW>2.0.CO;2.
Wallace, J. M., and D. W. J. Thompson, 2002: Annular modes and climate prediction. Phys. Today, 55, 28–33, https://doi.org/10.1063/1.1461325.
Warner, J. L., 2018: Arctic Sea ice—A driver of the winter NAO? Weather, 73, 307–310, https://doi.org/10.1002/wea.3399.
Woollings, T., A. Hannachi, B. Hoskins, and A. Turner, 2010: A regime view of the North Atlantic Oscillation and its response to anthropogenic forcing. J. Climate, 23, 1291–1307, https://doi.org/10.1175/2009JCLI3087.1.
Young, J. A., 2003: Static stability. Encyclopedia of Atmospheric Sciences, 2nd ed. J. R. Holton, J. A. Curry, and J. A. Pyle, Eds., Vol. 2, Academic Press, 423–430.