1. Introduction

[2]  The global climate change and anomalous weather conditions observed during the last decade in different regions of the Earth require a deep understanding of the natural and anthropogenic factors responsible for these phenomena. For this reason, an ever increasing attention of researchers has been devoted to the problem of influence of variations in solar activity on the lower atmosphere state, weather, and climate.

[3]  The problem of existence of solar-terrestrial connection and the physical mechanism underlying it has a long history. However, systematic investigations of the link between variations in solar activity and weather phenomena started, in all probability, only from the end of the 19th century. For instance, Koeppen [1873] studied the relation between the solar activity level (Wolf numbers) and mean air temperatures in the Northern Hemisphere. He showed that air temperatures exhibit pronounced 11-year variations correlating with similar oscillations in solar activity. However, the sign of the correlation was found to be different: from 1777 to 1790 the correlation was positive, while from 1815 to 1854 it was negative. Further investigations of Koeppen [1914] convincingly confirmed the results obtained by him earlier. From that time, hundreds of publications concerned with different manifestations of solar activity in changes of the lower atmosphere parameters, weather, and climate of the Earth have appeared. For instance, Ohl [1969], Shiyatov [1972], King [1975], and Vitinskiy et al. [1976] reported the data pointing to the existence of 11-, 22-, 45-, and 95-year cycles in quite different weather conditions, such as, for example, precipitation intensity, variations in surface temperatures and pressures in different regions of the Earth, annual tree ring widths, and drought rhythm in the western United States. There are a number of publications that demonstrate that the solar activity affects circulation of the lower atmosphere. Wilcox et al. [1974] have revealed that the area of the low-pressure regions in the troposphere decreases and goes to a minimum in approximately a day after the Earth crosses the sector boundary of the interplanetary magnetic field. Bucha [1984, 1988] and Bucha and Bucha [1998] have concluded that an enhanced corpuscular radiation leads to a reduced pressure in the polar regions and hence to enhanced zonal circulation, which results in a considerable rise in air temperatures in Europe. They also point to the existence of a pronounced cyclicity in weather changes.

[4]  Thus a vast body of experimental evidence indicating that there are statistically significant relations between different weather phenomena and solar activity has been accumulated. Nevertheless, many specialists, and especially meteorologists, show a good deal of skepticism about the idea that solar activity exerts influence on the lower atmosphere state [Burroughs, 1992; Monin, 1969; Pittock, 1978; Salby and Shea, 1991; Siscoe, 1978] (see also reviews of Avdyushin and Danilov [2000], Pudovkin and Raspopov [1992], Hoyt and Schatten [1997], and Carslaw et al. [2002]).

[5]  The reasons for the doubts are rather strong. The main arguments are [Bucha and Bucha, 1998] as follows: (1) the absence of steady correlation relations between the solar activity level and different weather conditions, (2) an appreciable unbalance (of several orders of magnitude) between the power of atmospheric processes and intensity of variations in the fluxes of solar wave and corpuscular radiation (a possible solution of the problem will be discussed in sections 2.2-2.4), and (3) the absence of probable physical mechanisms responsible for the possible solar-terrestrial connection.

Figure 1

[6]  The assertions given above are illustrated in Figure 1 which shows variations in annual mean Wolf numbers ( W ) and air temperatures in St. Petersburg ( d T ) (the 5-year running mean) for the last more than 100 years [Zaitseva et al., 2003]. As can be seen from Figure 1, variations in surface temperatures exhibit pronounced quasi-11-year cycles, which is in full agreement with the results obtained by Koeppen [1873, 1914]. However, the relation between these variations and the relevant oscillations in solar activity is rather complicated. For instance, in the periods from 1870 to 1910 and from 1945 to 1995, variations in temperature and solar activity are in phase, while from 1910 to 1930 they are in opposite phase. From 1930 to 1945, there is no relation between variations in temperature and solar activity at all. On the whole, the correlation coefficient during the period considered here is low, and the relation between W and d T is statistically insignificant. Therefore Figure 1 can be interpreted as a visual demonstration of the occasional nature of the connection between solar activity and the lower atmosphere state observed during some periods [Burroughs, 1992; Carslaw et al., 2002; Pittock, 1978]. This connection can be explained by the fact that some intra-atmospheric processes have periodicities similar to those of the sunspot activity of the Sun [Donarummo et al., 2002; Hurrel, 1995; James and James, 1989; Ram and Stolz, 1999]. The reversal of the sign of correlation between variations in solar activity and air temperature is attributable to gradual accumulation of the phase shift between them.

[7]  Note, however, that the data shown in Figure 1 can be interpreted in an entirely different way. If we take into account the fact that variations in the surface air temperatures are brought about by a set of different factors (wind system, cloudiness, atmospheric transparency) rather than one factor, it can be supposed that the observed reversal of the sign of correlation between d T and W variations can be explained by a fairly complicated dependence of these factors on the solar activity level, the connection between temperature and each of these factors being stable. The task of a researcher is to identify these factors and to find their relation with variations in solar activity.

Figure 2

[8]  To illustrate this idea, let us consider the following example. Figure 2a shows variations in winter air temperatures t^o in the polar stratosphere at the 30-mbar height and solar activity level (intensity F of radio emission with a wavelength of 10.7 cm) from 1956 to 1988. As can be seen from Figure 2a, the relation between variations in t^o and F during the entire period is rather weak (r = 0.14 ). However, if we separate the analyzed data in accordance with the quasi-biennial oscillation (QBO) phase, the picture becomes appreciably different. Namely, for the years with QBO in the westward phase, there is a distinct positive correlation (r = 0.76 ) between t^o and F (Figure 2b), while the years with QBO in the eastward phase are characterized by a negative correlation between t^o and F, though it is less pronounced (Figure 2c). Thus the impression that there is no relation between variations in the polar stratosphere temperature and solar activity is gained because the different states of the atmosphere have different responses to variations in solar activity, but the relation does exist. This example shows how careful one should be when discussing whether the solar-terrestrial connection exists or not.

[9]  As far as the physical mechanisms underlying the observed Sun-weather relation are concerned, it should be noted that the abundance of the hypotheses rather than the lack of them seems rather suspicious. However, in our opinion, this is the consequence of the diversity of the solar-terrestrial connections and of a wide spectrum of atmospheric processes and their relations with variations in different cosmophysical factors rather than the fact that the subject has been poorly studied.

[10]  The basic mechanisms responsible for the specific features of solar-terrestrial connection are as follows: (1) changes in the solar irradiance associated with short- and long-term variations in solar activity, (2) changes in cosmic ray fluxes and the resulting variations in the parameters of the global electric circuit and variations in the cloud formation rate caused by these changes, (3) changes in the lower atmosphere dynamics and variations in the propagation of planetary waves and the energy carried by them, and (4) changes in atmospheric transparency and cloudiness (also under the action of varying cosmic ray fluxes). Let us consider in more detail some of the hypotheses listed above and their experimental and theoretical bases.