3. Short-Term Disturbances in the Lower Atmosphere

Figure 10

[44]  First of all, let us see what kind of disturbance develops in the lower atmosphere under the action of solar flares. For this purpose, Figure 10 [after Pudovkin and Babushkina, 1992] shows variations in the Blinova index [Blinova, 1978; State Committee of USSR on Hydrometeorology and Environmental Control, 1977-1989] calculated through the superposed epoch method for the latitude range j = 45^o-65^o at isobaric levels 700, 500, and 300 mbar for intense geomagnetic storms. The key day is taken to be the day of the geomagnetic disturbance onset. Let us remind that the Blinova index A_{j 1 j 2} characterizes the average angular velocity of zonal atmospheric circulation a in the latitude range [ j 1, j 2 ]; A_{j 1 j 2} = 10³ a_{j 1 j 2}/W, where W is the angular velocity of the Earth's rotation. At the latitude of St. Petersburg, the linear velocity of the zonal circulation corresponding to A=1 is about 25 cm s^-1.

[45]  As can be seen from Figure 10, the evolution of the geomagnetic disturbance is accompanied by considerable changes in the zonal atmospheric circulation rate. The changes in the rate occur synchronously at all the levels considered and are characterized by increasing A at the first stage of disturbance and by a significant decrease in A during the subsequent period. According to calculations of Pudovkin and Babushkina [1992] presented in Figure 10, variations in A are due to 20 m changes in the heights of isobaric surfaces, which requires an additional energy input (or output) of (5-8) 10²⁶ erg to (from) the lower atmosphere.

Figure 11

[46]  What is the source of this energy? Depending on the suggested mechanism, this may be either the latent evaporation heat or deficit (or surplus) of solar irradiance modulated by variations in the atmospheric transparency or cloudiness. In this connection, Figure 11 from Pudovkin and Babushkina [1992] shows variations in the daily Kp index and intensity of the GCR flux (Figure 11b) obtained by the method of superposed epochs as well. Figure 11a presents variations in the direct solar radiation at the Earth's surface (at the plane perpendicular to the line of sight) in the auroral ( j = 65^o-75^o ) and subauroral ( j = 55^o-62^o ) zones. It is evident that the development of the Forbush decreases is accompanied by a considerable (5-10%) increase in the solar irradiance and hence atmospheric transparency [Veretenenko and Pudovkin, 1997]. The increase in the solar radiation flux given above corresponds to an additional input of energy to the lower atmosphere, depending on the season, of the order of (0.2-2.6) 10²⁶ erg day^-1, which is rather close to the energy of the disturbances considered here [Veretenenko and Pudovkin, 1999].

Figure 12

[47]  Because of a relatively small amount of data used for plotting Figure 11, it can be supposed that the obtained relation between variations in the cosmic ray flux and atmospheric transparency is occasional. However, if the increase in the atmospheric transparency observed during a Forbush decrease were indeed due to the reduction in the cosmic ray flux, the increase in the flux would be accompanied by a lower atmospheric transparency. It is possible to check this hypothesis by analyzing variations in atmospheric transparency in the course of solar proton events (SPE) associated with precipitation to the Earth's atmosphere of energetic (with energies to several hundreds of MeV) solar protons generated on the Sun during intense chromospheric flares. Figure 12 [from Roldugin and Vashenyuk, 1994] shows variations in the atmospheric transparency at a wavelength of l=3440 Å at stations Murmansk and Arkhangelsk in the course of several proton events. It is evident from Figure 12 that the SPE development is accompanied by a considerable decrease in the atmospheric transparency, which confirms the hypothesis that it is the intensity of cosmic ray fluxes that determines the optical properties of the atmosphere.

[48]  Naturally, then the question arises as to what is the reason for changes in atmospheric transparency: variations in the ozone content in the atmosphere or condensation mechanism? To a certain extent, the question was answered by Roldugin and Vashenyuk [1994], who stated that a decrease in the atmospheric transparency is accompanied (or is caused?) by a twofold to fourfold increase in the concentration of aerosol particles with a radius of 0.1-1  m m, which apparently points to a significant role of the condensation mechanism.

Figure 13

[49]  Of interest are also the data shown in Figure 13 [after Veretenenko and Pudovkin, 1994] that demonstrate variations in the cloud cover (in percent) during Forbush decreases of GCR at different latitudes. As can be seen from Figure 13, at latitudes 60^o- 64^o (most probably, at higher latitudes as well), a pronounced decrease in the total cloud cover of the order of 10% is observed on the 1st and 2nd day after the Forbush decrease onset.

[50]  In addition, Figure 13 also indicates that the effect of the Forbush decreases in the cloudiness vanishes at a latitude of the order of 50^o. If we assume that such a latitudinal behavior of this effect is explained by the magnetic cutoff of particles, the energy of the particles should be about 1 GeV. Protons with such energies penetrate into the atmosphere to the heights of the order of 10 km, where the phenomena considered here are likely to occur. The latter supposition is confirmed by the fact that the effect of Forbush decreases is observed most distinctly at the stations where the high-level clouds prevail, and the main type of clouds are cirri [Pudovkin and Veretenenko, 1996; Veretenenko and Pudovkin, 1994].

Figure 14

[51]  In view of the above said, it can be supposed that, in contrast to Forbush decreases, solar proton events associated with an increase in the flux of energetic particles must give rise to increasing cloudiness. Figure 14 [after Veretenenko and Pudovkin, 1996] presents variations in the cloud cover during the SPE at a number of meteorological stations at which the low-level cloudiness is typically weak. It can be seen from Figure 14b that as expected, the SPE is accompanied by a considerable increase in cloudiness. Like in the case of the Forbush decreases, the effect of SPE depends on the latitude of the observation point and decreases from 40% at a latitude of 70^o to zero at 55^o. The decrease in the cloudiness on the minus third day is most likely to be due to the X-ray radiation of solar flares [Veretenenko and Pudovkin, 1996].

[52]  Taking the estimates of Schneider [1972], according to which a 8% change in the cloud cover is equivalent to a 2% change in the "solar constant," we infer that the Forbush decreases cause the effect at the Earth's surface equivalent to the 2.5% change in the "solar constant."

Figure 15

[53]  On the other hand, solar proton events lead to an increase in the cloud cover of up to 40%, which is equivalent to the reduction in the "solar constant" by 5%. Such changes in the solar irradiance in the lower atmosphere must cause rather large changes in the vertical temperature profile in the troposphere and lower stratosphere. Variations in the vertical temperature profile D t^o (deviations from the "quiet" level) during the Forbush decrease were obtained (also by the superposed epoch method) from the data of the Sodankyla observatory ( j =67^o ) in Finland by Pudovkin et al. [1995, 1995] obtained results are shown in Figure 15. It can be seen that as Forbush decreases of GCR develop, the temperature in the troposphere increases by several degrees and reaches a maximum on the second day after the Forbush decrease onset. At the same time, in the lower stratosphere, the temperature decreases by several degrees.

Figure 16

[54]  Figure 16 [after Pudovkin et al., 1995] shows variations (also deviations from the "quiet" level) in the vertical temperature profile at Sodankyla during 19 solar proton events taking place in 1981-1988. As could be expected, the temperature in the troposphere decreases by several degrees on the second day after the disturbance onset. Along with this, attention should be paid to the fact that on the first day of the disturbance the temperature in the troposphere rises by 1^o-2 ^o rather than decreases. This is likely to be the evidence of changes in the optical properties of the clouds in the course of disturbance or, more probably, of formation of two different types of clouds, each with its own optical properties and lifetime.

[55]  It is obvious from Figures 15 and 16 that in the case of Forbush decreases of GCR and also during solar proton events, D t^o reverses its sign at an altitude of 10 km, which points, evidently, to the fact that the cloud or aerosol layer responsible for the observed changes in temperature is located precisely at this altitude.

Figure 17

[56]  Thus the data presented above indicate that variations in cosmic ray fluxes are, indeed, accompanied by significant changes in air temperatures in the lower atmosphere. Therefore it can be anticipated that these changes will lead to variations in the atmospheric circulation as well. Figure 17 [from Veretenenko and Pudovkin, 1993] shows variations in the Blinova indices at an isobaric level of 500 mbar during Forbush decreases of GCR (Figure 17a) and solar proton events (Figure 17b). It is obvious from Figure 17 that a decrease in the cosmic ray flux is, indeed, accompanied by decreasing zonal circulation rate, while an increase in the SCR flux leads to increasing circulation rate.

[57]  On the basis of the results described above, we can interpret the data on variations in the atmospheric circulation rates during geomagnetic disturbances shown in Figure 10 as resulting from the superposition of the effects of penetration of solar protons (and solar X-ray radiation) and Forbush decreases of GCR developing with a time shift of 1-2 days.

[58]  Analysis of the results obtained in the investigations mentioned above leads to the conclusion that at the 55^o-70 ^o latitudes the disturbances in the lower atmosphere associated with different manifestations of solar activity are mainly due to variations in the solar irradiance reaching the lower atmosphere [Veretenenko and Pudovkin, 1998]. In turn, these variations are related to changes in atmospheric transparency and cloudiness at the altitudes of the upper troposphere and lower stratosphere that are modulated by variations in cosmic ray fluxes. This suggests that the key role in changes of atmospheric transparency, at least in the latitudinal range considered here, is played by the condensation mechanism. This does not mean, of course, that the ozone mechanism does not provide any contribution into changes of the optical properties of the atmosphere.

Figure 18

[59]  Unfortunately, there are few papers devoted to specific features of solar weather connection at low latitudes ( <55^o ). In all probability, the most comprehensive treatment has been given in a series of papers of E. Friis-Christensen, H. Svensmark, and N. Marsh [Friis-Christensen and Svensmark, 1997; Marsh and Svensmark, 2000; Svensmark, 1998; Svensmark and Friis-Christensen, 1997]. One of the most important results of their studies is presented in Figure 18, which shows variations in cosmic ray intensity (12-month running mean deviations from the norm), from the data of Climax, and cloudiness (also deviations from the norm, in percent) of, mainly, the low-level cloud cover [Marsh and Svensmark, 2000] over oceans derived from the data of a geostationary satellite. It is obvious from Figure 18 that there is a pronounced relation between variations in cloudiness and cosmic ray flux in the region involved. The observed cyclic variations in cloudiness exert a considerable influence on the input of solar energy into the lower atmosphere and, consequently, the air temperatures in it. On the basis of these facts, they concluded that variations in cosmic ray fluxes play a fundamental role in formation of weather and climate at low latitudes. The physical mechanisms that are developed in a cloud under the action of cosmic rays are not discussed in detail in the works mentioned above. According to Carslaw et al. [2002] the processes associated with the direct influence of cosmic rays on the atmosphere and also with changes in the parameters of the global electric circuit can develop at low latitudes. However, relative contributions of these processes are not clear.

[60]  So far we have studied morphology of elementary disturbances (i.e., those associated with single isolated heliospheric perturbations) in the lower atmosphere. This allowed us to exclude, to a large extent, many phenomena, secondary in nature though no less important. These were, for instance, changes in air temperature, cloudiness, and atmospheric transparency arising due to transfer of air masses caused by excitation of the additional atmospheric circulation system or variations in the thermal balance of the ocean-atmosphere system. In analysis of long-term variations in the lower atmosphere parameters and their connection with variations in solar activity, the impact of all these phenomena cannot be neglected. Their net effect should be considered, which appreciably complicates the picture [Veretenenko and Pudovkin, 1999].

[61]  As an example, let us return to Figure 1. As it has been mentioned above, the data shown in Figure 1 illustrate a rather specific situation. On the one hand, variations in air temperatures in St. Petersburg exhibit the periodicity coinciding in its characteristics with the solar activity periodicity. On the other hand, on the whole, there is no long-term statistical relation between changes in these variables, and Bucha and Bucha [1998] point to the fact that this situation is typical. It would be relative to the response of ocean-atmosphere system, which can have regional peculiarities that have own temporal variations.