**4. Ozone as a mediator of geomagnetic field influence on climatic variables**

The sensitivity of atmospheric temperature profiles and climate to the ozone density (particularly near the tropopause) has been noticed long ago [67–71], etc. The detected synchronization between the spatial and temporal variability of particles' flux reaching the ground, and the lower stratospheric ozone, is a hint that ozone could serve as a mediator of the geomagnetic field-energetic particles' influence on climatic

**Figure 7.** *Spatial distribution of centennial ozone changes between the first decades of twenty-first and twentieth centuries.*

variables (i.e. temperature, pressure, etc.) [72]. The following section throws some more light on this problem.

#### **4.1 Ozone imprints on climatic variables**

#### *4.1.1 Hemispherical and longitudinal asymmetries of ozone-temperature covariance*

The potential synchronization between ozone at 70 hPa and near-surface temperature variability, within the period 1900–2010, is examined by the use of lagged crosscorrelation analysis. The leading role of winter ozone in the ozone-temperature correlation, have been analyzed in a spatial grid with 10° steps in latitude and longitude. The time series of both variables are taken from the monthly values provided by the ERA twentieth century reanalysis. The correlation map presented in **Figure 8** is created from the preliminary weighted correlation coefficients by the autocorrelation function of ozone, with lag corresponding to the time delay of temperature response – to account for the reduced weigh of covariances being away from the moment of applied forcing.

The most impressive of the results shown in **Figure 8** is the asymmetry of the temperature response to ozone variations. The positive O3 –T2m correlation coefficients – over Eurasia and the extratropical Pacific Ocean, unlike the overall negative correlation, require their explanation. In addition, the analysis of the long-term variations of ozone and temperature at 60°N latitude, and at longitudinal zones, 140 and 70°W (corresponding to the regions with positive and negative GCR-ozone correlation) are presented in **Figure 9**. It is important to note that the short-term variations are preliminarily filtered by data smoothing through 11-year running average procedure.

**Figure 9** clarifies that the lower temperature trend of Eastern Asia corresponds to the higher ozone density at 70 hPa. Oppositely, the stronger warming in south-eastern Canada corresponds to a lower ozone density at 70 hPa, with a negative centennial trend. Examination of the global picture of twentieth century warming (presented in **Figure 10**) reveals that the "hot spots" of contemporary global warming (i.e. northeastern Canada and Greenland, and the Southern Ocean – southward of Africa) correspond to the regions of negatively correlated ozone and temperature (refer to **Figure 8**). In opposite, the regions with in-phase co-varying ozone and temperature are characterized by weaker warming.

#### **Figure 8.**

*(top) correlation map of winter ozone at 70 hPa and air surface temperature, calculated over the period 1900– 2010; (bottom) time lag in years of temperature response following ozone changes.*

#### **Figure 9.**

*(left) Time series of winter ozone at 70 hPa and 60°N latitude, obtained at Eastern (140°E longitude) and Western (70°W) longitude; (right) air surface temperature at the same latitude and longitudes.*

#### **Figure 10.**

*Centennial changes of the air surface temperature between the first decades of twenty-first and twentieth centuries, derived from the ERA twentieth century reanalysis.*

Conclusively, the above results indicate that the strongest warming during the twentieth century is observed in regions with reduced density of the lower stratospheric ozone.

#### *4.1.2 Climatic modes and lower stratospheric ozone density*

Climate variability is not homogeneous in space and is usually described as a combination of some "preferred" spatial regimes, called *modes*. In meteorology and climatology, the term '*mode'* is used to describe a spatial structure with at least two strongly connected centers of action [73]. The most famous of these spatial structures – known as *climatic modes* – affect weather and climate on different spatial and temporal scales. Most climatic *modes* are defined by statistical classifications of the observed variability of surface temperature, sea-level pressure, precipitations, etc. They could be a result of the action of fundamental physical processes such as the instability of the climatic mean flow, mesoscale interactions between the atmosphere and the ocean, etc. [74]. However, these statistical patterns may also be artifacts of nature, whereby they are not stable over long periods of time, or they may be statistical artifacts.

Although the spatial-temporal variations of climatic modes are extensively studied, the reasons for their occurrence and variability over time are not fully understood. Internal variations of the climate system are usually associated with the processes of energy exchange and redistribution between the planetary atmosphere and ocean. The huge heat capacity of the ocean is the reason for its inertia in response to short-time fluctuations of atmospheric variables, which transforms them into long-period variations of the ocean surface temperature. This understanding does explain the phase alteration, but it is not able to explain neither the various manifestations of climatic modes [75] nor their long-term changes.

Analysis of the spatial-temporal variability of GCR and ozone at 70 hPa reveals the important role of the latter in the formation of regional specificity of air surface temperature variability (refer to Subsection 4.1.1, or to [76]). Examination of the temporal synchronization between two of the most important *climatic modes* – North

Atlantic Oscillation (NAO) and El Niño Southern Oscillation (ENSO) –confirms the existence of statistical relation in the regions of *modes*' manifestation [77].

**Figure 11** illustrates the projection of the long-term variations of ozone at 70 hPa on the NAO index (which describes the variability of the surface pressure between Azores and Iceland). The coupling between both variables has been estimated by the use of the lagged cross-correlation analysis between annual values of NAO index (smoothed by 5 points averaging) and winter ozone values at 70 hPa (smoothed by 11 points moving window). The stronger smoothing of ozone is due to its higher temporal variability. The leading factor (i.e. the "forcing") in calculated ozone-NAO variability is ozone. As in the previous case, the correlation coefficients have been preliminarily weighted (according to different delay of NAO response) with the ozone's autocorrelation function. The physical reasoning behind this weighting is that the memory of the climate system for the applied impact weakens with time. This suggests that the high correlation coefficients with a large delay are more or less random.

**Figure 11** shows that the ozone's impact on the NAO climatic pattern fairly well coincides with both centers of action (Azores and Iceland) determining the phase of NAO mode. Unlike the previous results (stressing the leading role of the northern [78] or the southern part of NAO spatial structure [79]), **Figure 11** indicates that the variations of lower stratospheric ozone density can impact each center of action

**Figure 11.**

*(top) Cross-correlation maps of the winter lower stratospheric ozone and NAO index, calculated for the period 1900–2010; (bottom) time lag of NAO response in years.*

(Azores or Icelandic), or simultaneously both of them – altering in such a way the phase of NAO mode [76].

Analysis of the time delay of NAO response to ozone changes shows that surface temperature near the Icelandic Low respond with a delay of 1–2 years. In the subtropical center of action, however, the atmospheric response is delayed approximately by a decade (see the bottom panel in **Figure 11**).

**Figure 12.**

*Comparison of correlation maps of ozone at 70 hPa with GCRs (dark shading) and water vapor at 150 hPa (contours), for winter (a) and (c), and summer (b) and (d) panels.*

#### **4.2 Mechanism of ozone influence on climatic variables**

Direct ozone influence on the surface temperature is quite small due to the mutually exclusive effect of stratospheric and tropospheric ozone in the planetary radiation balance [70]. Ozone's ability to absorb the incoming solar radiation (and to a lesser extent the longwave radiation emitted from the Earth), makes it a radiatively active gaz. The covariance between the near tropopause ozone and temperature has been noticed long ago [80, 81]. However, the tropopause temperature determines the moist adiabatic lapse rate and accordingly the static stability of the upper troposphere [82, 83], which in turn alters the humidity near the tropopause [51]. For example, ozone depletion cools the near tropopause region making the upper troposphere more unstable [82, 83]. The upward propagation of the more humid air masses from the lower atmospheric levels moistens the upper troposphere, and strengthens the greenhouse warming of the planet. The satellite measurements show that water vapor at these levels ensures 90% of the greenhouse warming of the total atmospheric humidity [84]. Consequently, ozone variability in the lower stratosphere is projected on the planetary surface through the modulation of the strength of greenhouse warming.

**Figure 12**, which compares the lag-corrected correlation maps of ozone mixing ratio at 70 hPa with: (i) GCR, and (ii) humidity at 150 hPa, is a good illustration of our hypothesis validity. Note that the latitudinal band of antiphase correlation between GCRs and ozone (dark shading), and in phase correlation between ozone and water vapor (red contours), coincide impressively well. In the Northern Hemisphere, this coincidence persists round the year, although being slightly reduced in summer season (compare panels (a) and (b) in **Figure 12**). In the winter Southern Hemisphere, the area of synchronous variations of GCR, ozone, and humidity is narrower and practically disappears in summer (**Figure 12d**). The results presented in **Figure 12** are a good indication that ozone–humidity variations, which are projected down to Earth's surface by the strengthening or weakening of the greenhouse effect, are actually related to GCR variability.

### **5. Conclusions**

Historical and contemporary changes in climate system put a lot of questions, the answers to which are difficult. This motivates scientists from different branches to look for various factors with a potential influence on the climate system. Geomagnetic field is one of the proposed factors, due to the rendered multiple evidence for spatially or temporary co-varying geomagnetic field and climate, at different time scales. In this chapter, we clarify that hypothesized geomagnetic influence on climate could be reasonably explained through the mediation of energetic particles, propagating in Earth's atmosphere, and their influence on the ozone density in the lower stratosphere.

More specifically, the non-dipolar part of geomagnetic field creates irregularities in the spatial distribution of lower atmospheric ionization in the Regener-Pfotzer maximum [51]. The bulk of low-energy electrons and dry lower stratosphere favors activation of autocatalytic ozone production at these altitudes. Thus geomagnetic irregularities are projected on the ozone density near the tropopause. Being a radiatively active gas, the ozone itself affects the temperature and humidity in the tropopause region, altering in such a way the greenhouse effect and consequently – the near-surface temperature.

This chapter provides evidence for the validity of this chain of sequences, which gives an adequate explanation of hemispherical and longitudinal asymmetry of the lower stratospheric ozone distribution, regionality of climate change, formation of regional climate patterns, known as climatic modes, etc.
