*3.4.2 Evidence for particles' influence on the lower stratospheric O3 density*

At middle and high latitudes, the Regener-Pfotzer maximum is placed well above the tropopause [60], which provides the necessary conditions for activation of the autocatalytic cycle of ozone production – i.e. a dry atmosphere and plenty of low energy electrons. As discussed in Sections 3.2 and 3.3, and shown in [52], the ionization in the Regener-Pfotzer maximum is unevenly distributed over the globe. Remind that an increased particles' flux is expected in regions of geomagnetic field strengthening. Consequently, if the autocatalytic production of ozone is significant, the longitudinal variations of the Regener-Pfotzer maximum ionization should be projected on the ozone profile.

**Figure 5** presents a comparison between ozone profiles in regions with increasing and decreasing geomagnetic field, during solar minimum in 2009. Note that the O3

#### **Figure 5.**

*Difference between ozone profiles in regions with positive (red curves) and negative (black curves) crosslongitudinal magnetic gradients; (a) for the Eastern hemisphere, and (b) for the Western one.*

values beneath the peak ozone density are higher in regions with increasing geomagnetic field (i.e. the longitudinal sector 90–50°W in the Western Hemisphere and 120– 140°E – in the Eastern one), relative to corresponding O3 values in regions with a geomagnetic field weakening in the sectors: 140–110°W and 30–50°E.

The longitudinal variations in atmospheric ozone have been noticed long ago [61, 62]. The authors have suggested that this variability could be related to the planetary wave structure. However, the maximal amplitude of the stationary planetary waves is found at 300 hPa [61], while the highest amplitude of O3 longitudinal variations in ERA Interim reanalysis is placed near 150–70 hPa [51]. These and some other problems, e.g. [63, 64] suggest that other factor(s) (e.g. energetic particles) may have an important influence on the spatial and interannual variability of the extratropical near tropopause O3.

In order to assess quantitatively the coupling between energetic particles precipitating in Earth's atmosphere and lower stratospheric ozone, as well as its spatial distribution, we have performed a cross-correlation analysis in a grid with 10° increments in latitude and longitude. Ground-based measurement of galactic cosmic rays (GCR) by neutron monitors, has been used as an indication of energetic particles flux. The Moscow record of GCR has been expanded backward in time by the paleore constructed GCR intensity [65]. The 11-year periodicity of GCR has been removed by moving averaging procedure with 22-year running window. The winter values of ozone at 70 hPa have been taken from ERA twentieth century reanalysis, covering the period 1900–2010. Data have been preliminarily smoothed by 11-year running window.

The map of ozone-GCR correlation is presented in **Figure 6** (colored shading). It is important to note that the map has been created from correlation coefficients, being preliminary weighted by the autocorrelation function of GCR with time lag corresponding to the delay of O3 response to the GCR forcing. This procedure, which reduces correlation coefficients with longer time lags, allows a comparison of correlations with different time lags. The introduction of weighs for the lagged correlation coefficients is justified by the assumption that the effect of the applied forcing in a given moment of time decreases with moving away from this moment [66].

#### **Figure 6.**

*Lag-corrected correlation map of GCR and O3 at 70 hPa (shading), compared with modeled effective vertical cutof rigidity of geomagnetic field (courtesy to Boschini MJ, Della Torre S, Gervasi M., Grandi D, Rancoita PG: Http://www.mib.infn.it, and Bobik P, Kudela K: http://space.saske.sk).*

**Figure 6** shows that the ozone responds differently to particles' impact at different regions over the world – not only by amplitude but even by sign. Thus, at high latitudes and in the Indo-Pacific region, ozone varies synchronously with GCR. On the other hand, at the Northern Hemisphere extratropics and near the southernmost edge of Latin America, both variables covariate in antiphase – meaning that in these regions ozone increases with time.

Such heterogeneity in ozone response to particles' forcing could be attributed to the different origins of impacting particles. For example, the polar regions are vulnerable to the particles from interplanetary space, propagating along the open geomagnetic field lines. The long-term variations of these particles are modulated mainly by the interplanetary magnetic field in the heliosphere. The latitudes shielded by the closed geomagnetic field lines (i.e. the tropics and mid-latitudes) are accessible to very highly energetic particles (which are very few), and to the radiation trapped in the Van Allen radiation belts. The latter are subject to geomagnetic lensing (in the lowest part of their helical trajectories along the magnetic field lines) and asymmetrical precipitation in both hemispheres, due to the asymmetry of geomagnetic field (refer to Sections 3.2 and 3.3).

**Figure 6** shows in addition the effective vertical cut-off rigidity of geomagnetic field (contours), with the values greater than 12 GV being colored in red. Note that the strongest GCR-O3 correlation over the equatorial Indo-Pacific region fairly well coincides with the higher geomagnetic cut-off rigidity. Having in mind the centennial negative trend in GCR, the positive correlation coefficients indicate ozone depletion during the examined period (1900–2010). Consequently, the reduced ozone density could be attributed to the weaker particles' fluxes accessing the said region.

On the other side, the negative GCR-ozone correlation in extratropics suggests enhancement of ozone density near 70 hPa. This result indicates that particles confined in the outer radiation belt are involved in ozone production in the lower stratosphere. Powered by the solar wind, the population of this radiation belt is highly variable [45], reflecting the changes in solar activity. The examined period is characterized by enhanced solar activity, which appears to be projected on the extratropical latitudes as enhanced ozone density at 70 hPa – due to the enhanced particles' population in the outer radiation belt.

The positive GCR-O3 correlation at polar latitudes suggests a centennial ozone depletion, which corresponds to the decreased flux of GCR, modulated itself by the stronger interplanetary magnetic field in the heliosphere during the twentieth century [51].

The centennial changes in ozone mixing ratio at 70 hPa, between the first decades of twenty-first and twentieth centuries, is presented in **Figure 7**. Note that ozone changes deduced from the correlation map in **Figure 6** fairly well corresponds to the observed changes of ozone at 70 hPa.
