106 erg cm-2 s-1 at the
Earth's orbit) experiences variations of not more than 0.1%.
[11] As it is known, the radiation spectrum of the Sun is similar to
that of the absolutely blackbody heated to 5770 K, though with a
substantial energy deficit in the near ultraviolet region. At the
same time, the intensity of solar radiation in the far ultraviolet and
X-ray regions is several orders of magnitude higher than the
relevant radiation of the absolutely blackbody. Such a difference
in the spectra of the Sun and the absolutely blackbody is
explained by the fact that the short-wave radiation in different
wavelength ranges is generated in different regions of the Sun's
atmosphere. In particular, the radiation with the wavelength
l <1500 Å
is generated in the chromosphere and corona of the
Sun, i.e., in the regions whose temperatures are much higher that
the temperature of the photosphere. It is also known that
parameters of the chromosphere and corona are rather changeable
and profoundly depend on the solar activity level. So, it is not
surprising that the intensity of the short-wave solar radiation
considerably varies from day to day and with the level of solar
activity. A relative value of cyclic variations in solar irradiance
reaches a factor of 10 at
l = 300-500 Å
and sharply decreases at
l >2000 Å.
As a result, the integral intensity of solar radiation (the
so-called solar constant equal to
K = 1.37 106 erg cm-2 s-1 at the
Earth's orbit) experiences variations of not more than 0.1%.
[12] It has recently been demonstrated in many publications that the solar irradiance in the white light can also depend on the solar activity level or, to be more exact, on the number and distribution of active formations, such as sunspots, faculae, and flocculi on the surface of the Sun.
[13] For instance, Hudson et al. [1982] showed that the appearance of relatively dark sunspots at the solar disk can lead to weakening of the solar irradiance by several tenths of a percent, while relatively bright faculae result in increasing solar radiation. Though an increase in the brightness of the photosphere in the regions of faculae is only several percent, the area occupied by them is typically larger by an order of magnitude than the total area occupied by the sunspots, and so variations in the solar irradiance due to these formations are of the same order of magnitude. Nevertheless, on the whole, the effect of facular emission dominates over the blocking effect of the sunspots in the cyclic variations of the Sun's brightness.
|
| Figure 3 |
[14] The role of the effects discussed above can be illustrated by Figure 3, which is after Wilson et al. [1981]. Figure 3 (top) presents images of the Sun with a large sunspot group crossing the central meridian in April 1980; Figure 3 (bottom) shows variations in solar irradiance S. Thin lines indicate to which point on the S(t) curve the solar image corresponds. It is evident from Figure 3 that when the sunspots approach the central meridian of the solar disk, their visible size is at a maximum and the solar irradiance reaches a minimum. The faculae surrounding the spots have a minimum brightness when near the center of the disk, while near the limb their brightness is maximum. After the sunspots pass off the solar disk, some faculae still remain at the disk, and the solar irradiance reaches a maximum. This explains the local maxima observed on the S(t) curve.
|
| Figure 4 |
[15] In a somewhat different form, this result is shown in Figure 4 (left) [after Lean, 1991]. Figure 4 presents variations in the total solar irradiance ( S ) during 1982 (Figure 4a); weakening of the irradiance due to crossing of the solar disk by sunspots: d S = SqPs, where Sq is the irradiance of the "undisturbed" Sun and Ps is the "blocking function" of this irradiance by the sunspots (Figure 4b); the "residual" solar irradiance Sc = S - Sq(1 + Ps) characterizing the surplus irradiance (Figure 4c); and the intensity of the La emission generated in the regions of bright formations outside the active region (Figure 4d). It can be seen that some irradiance minima (for instance, those that occurred in February, March, June, and July) are indeed due to the effect of sunspots, while the residual solar irradiance Sc shows a pronounced relation with the La emission intensity. Thus crossing of the solar disk by active structures indeed distinctly modulates the solar total irradiance. Along with this, the amplitude of these variations, as seen from Figure 4a, is not higher than 0.25%.
[16] The smoothed curves in Figure 4 (right) present 81-day running means of S, SqPs, Sc, and La in the course of the 11-year solar cycle. It is apparent from Figure 4 that the solar irradiance experiences a pronounced cyclic variation as well: S is maximum in 1981-1982 and 1989, i.e., in the period of solar activity maximum, and S is minimum in 1985-1987, i.e., in the period of minimum solar activity. This can be explained by the fact that during the solar cycle an increase in the brightness of the Sun associated with relatively bright and extended facular fields dominates over the blocking effect of sunspots. The relation between the annual values of the solar constant and Wolf numbers can be written as [Hoyt and Schatten, 1997]
 | (1) |
where W is the sunspot number.
[17] Along with this, it can be seen that the amplitude of the cyclic variation in the solar irradiance is larger than the amplitude of short-term variation and amounts to about 0.1% of its average value. The obtained empirical relation between solar activity and irradiance allows one to calculate the latter for preceding epochs.
|
| Figure 5 |
[18] More accurately, S was calculated for the past, to 1874, by Foukal and Lean [1990]. They used the data on the area of sunspots obtained at the Greenwich Observatory; the area of flocculi was assumed to be proportional to the Wolf numbers. The results of calculations are shown in Figure 5 taken from Foukal and Lean [1990]. It demonstrates that the amplitude d S of cyclic variations in S has been monotonically growing from the middle of the 19th century to the present. However, even maximum d S does not exceed 0.1% of the average S, which is likely to be an order of magnitude lower than the d S value that can explain the variations in the lower atmosphere temperature observed during the 11-year solar cycle [Burroughs, 1992]. Indeed, if we assume, as always, that a 1% increase in the solar constant leads to a temperature rise of 1.7o [Hoyt and Schatten, 1997], then it follows that the observed cyclic variations in S can lead to changes in air temperature of not more than 0.16% [see also Wigley and Raper, 1990]. Moreover, the phase of the observed cyclic temperature variations does not coincide in many cases with the solar cycle phase (see Figure 1). These facts indicate that the solar irradiance variability is not the only cause of cyclic variations in the atmospheric temperature and, apparently, not the main one.
[19] A fundamentally different mechanism of connection between solar activity and weather pattern was suggested by Markson [1978]. The advantage of this model is, in the opinion of the author, that it does not require significant changes in solar irradiance. Instead, the mechanism of the efficient use of the energy accumulated in the atmosphere is involved. In addition, the model explains the unexpectedly small lag of variations in the lower atmosphere parameters behind variations in relevant cosmophysical factors (solar flares, cosmic ray fluxes, passage of the sector boundaries of the interplanetary magnetic field past the Earth).
|
| Figure 6 |
[20] Let us consider this mechanism in more detail. Figure 6 [after Markson, 1978] schematically shows basic elements of the global electric circuit. According to the model under consideration, the global thunderstorm activity can be regarded as a powerful generator of electric current maintaining the electric field in the Earth's atmosphere. If the generator were switched off, the atmospheric electricity would decay with a characteristic decay time t = 15 min [Israel, 1971].
|
| Figure 7 |
[21] The electrical conductivity of air in the lower atmosphere
results mainly from ionization of air molecules by fluxes of
energetic particles of galactic cosmic rays (GCR); at higher
altitudes it is caused by fluxes of ultraviolet or X-ray emission of
the Sun. As a result, the atmospheric conductivity exponentially
grows with increasing altitude (Figure 7, from
Israel [1973]).
Let
us now consider Figure 6 from which we see that the resistance of
an air column with a cross section of 1 cm2 is about
107 W m2, in
which
2 1016 W m
2 fall at the lower layers of the atmosphere
(from the Earth's surface to the lower boundary of the cloud
layer),
5 1016 W m2
is the internal resistance of the generator ( R0 ),
and about
1016 W m2 is the resistance of the air column from the
upper boundary of the thunderstorm generator ( 
13 km) to the
ionosphere ( h = 60 km). Therefore the total resistance of the
atmosphere above the generator (the generator area is taken into
account) is
106 W,
while the integral resistance of the atmosphere
in the region of closing currents does not exceed 200
W. Thus the
total resistance of the circuit and hence the current strength in it
are mainly determined by the atmospheric resistance in the region
of the thunderstorm generator. At the same time, local variations
in the electric currents in the circuit can be due to variations in the
atmospheric conductivity immediately in the region of closing
currents.
[22] The cosmic ray fluxes modulated by variations in solar activity give rise to changes in the atmospheric conductivity in the region of the thunderstorm generator and above it, which, in turn, leads to changes in the electric field intensity and current in the entire circuit.
[23] Increasing electric field intensity leads to increasing growth rate of water droplets in the cloud (see below) and hence higher precipitation intensity [Mason, 1971]. The droplets of sufficiently large size reach the Earth's surface before they evaporate and, as a result, a large amount of latent evaporation heat is released.
|
| Figure 8 |
[24] Further development of an atmospheric disturbance in Markson's [1978] model is presented in Figure 8. The mechanism of influence of solar activity on the lower atmosphere dynamics suggested by the author is as follows. As a result of intense precipitation and release of latent evaporation heat, the temperature of the low-latitude atmosphere increases and causes the ascent of the heated air to the tropopause heights and intensification of the Hadley's convection cell, which has been, on the whole, confirmed experimentally [van Loon and Labitzke, 1994]. This, in turn, activates the middle-latitude cell and increases the convergence of airflows in the region of the cell polar front. According to Markson's calculations, the lag of the processes in the region of middle-latitude convergence behind the processes of intensification of the equatorial cell is 2-3 days.
[25] Intensification of the meridional circulation is accompanied by blocking of the zonal streams. Observations show that the meridional circulation indeed becomes stronger during the period of solar activity maximum [Schuurmans, 1969], and the height of the equatorial tropopause at this time is maximal [Rasool, 1975].
[26] Thus the model suggested by Markson [1978] qualitatively explains the mechanism of impact of solar activity on the lower atmosphere state. At the same time, the model does not contain any quantitative estimates of the changes in the atmospheric parameters it involves (temperature, pressure, wind velocity). Markson's [1978] model was significantly developed by B. A. Tinsley and his colleagues [Tinsley, 2000; Tinsley and Deen, 1991; Tinsley and Heelis, 1993; Tinsley et al., 1989, 2000].
[27] One of the fundamental parameters characterizing the optical properties of clouds and determining the precipitation intensity is the concentration of raindrops. In turn, the concentration of raindrops is determined by the concentration of nuclei of aerosol particles with a diameter of not more than 0.1 m m.
[28] Another important characteristic of clouds is the rate of formation of ice crystals in them. The presence of ice crystals sharply increases the formation rate of raindrops and hence the precipitation intensity. In addition, formation of ice affects the thermodynamic structure of the clouds, which in turn affects the cloud cover area.
[29] In the cloud, ions are rather efficiently scavenged by raindrops, and as a result, the atmospheric conductivity in the cloud significantly decreases. Since the electric current is continuous, this leads to enhancement of the electric field and accumulation of electric charges at the upper and lower boundaries of the cloud. As a consequence, the aerosol particles in this region acquire rather a large charge (to 1000 e ) [Reiter, 1992]. Tinsley and his colleagues put forward the hypothesis that charged aerosol particles are fairly effective ice forming nuclei. The experimental proof of this statement is that scavenging of aerosols appreciably increases if the aerosol particles are charged. Theoretical calculations also indicate that the rate of aerosol scavenging rapidly grows as their charge increases [Tinsley et al., 2000].
[30] Thus the model suggested by Markson [1978] can be summarized as follows.
[31] Variations in the parameters of the global electric circuit caused by variations in cosmic ray fluxes give rise to changes in the vertical current density and electric field intensity in the lower atmosphere and thus in ion concentration and electric charge of aerosol particles near the cloud boundary. As noted above, this leads to acceleration of the growth of ice crystals and raindrops, to intensification of precipitation, and an increase in the rate of release of latent evaporation heat.
[32] The works of Tinsley and his colleagues give significant insight into the physical processes relating the changes in the parameters of the global electric circuit to the lower atmosphere state and climate. However, even in such a modified form the Markson-Tinsley model does not provide any reliable quantitative estimates of the expected effects.
[33] The dynamic mechanism of the impact of solar activity on the lower atmosphere state suggested by Avdyushin and Danilov [2000] is as follows.
[34] It is known that there is a spectrum of internal waves with relatively large amplitudes in the atmosphere. Propagation of these waves into the upper layers of the atmosphere depends on the zonal air circulation in the stratosphere. For instance, the interplanetary waves can propagate into the stratosphere and then to the mesosphere only in the case of the eastward zonal circulation and if the velocity V is less than a critical velocity V cr [Avdyushin and Danilov, 2000; Geller, 1983]. Changes in solar activity are accompanied by variations in the solar ultraviolet radiation, a part of which is absorbed by the ozone layer, thereby causing changes in the air temperature and circulation in the stratosphere. Thus, if, during some period, the circulation in the stratosphere is formed by westerly winds whose velocity is close to the critical one, even insignificant changes in the circulation rate in the stratosphere induced by solar activity variations can control the exit of the internal atmospheric waves into the stratosphere or their confinement in the troposphere. In the latter case, the energy of these waves dissipates in the lower atmosphere, thereby causing its heating and changes in the topography of isobaric surfaces.
[35] This mechanism seems interesting. However, in its present form, the model does not allow assessment of either the intensity or spatial distribution of the expected disturbances in the lower atmosphere.
|
| Figure 9 |
[36] The mechanisms considered in this section are based on the observations according to which the solar irradiance reaching the Earth's surface exhibits a pronounced dependence on the solar activity level. For instance, Figure 9 [after Kondratyev and Nikolsky, 1970] shows solar irradiance as a function of the sunspot number. It can be seen that, as the sunspot number increases from zero to 100, the solar irradiance measured at the Earth's surface grows by ~3%, which can cause a temperature variation in the troposphere of 2o-3o. As the sunspot number further increases, the solar irradiance decreases. The reason for this decrease is not clear. Probably, it can be attributed to the competing effects of variations in the fluxes of galactic and solar cosmic rays [Pudovkin and Veretenenko, 1995] and X-ray radiation of the Sun.
[37] As noted above, the solar irradiance at the upper boundary of the atmosphere changes by not more than 0.1% during the solar cycle (see Figure 4). Therefore the variations in solar irradiance shown in Figure 9 can be due to changes in the atmospheric transparency alone (in a more general sense, the transmittance). What processes in the atmosphere can be responsible for the observed changes in atmospheric transparency? Kondratyev and Nikolsky [1995] suggest at least two such mechanisms, i.e., formation of cirri and ozone mechanism. Let us consider these models in more detail.
The condensation mechanism implies that the presence of ions in the atmosphere increases the rate of formation and initial growth of aerosol particles. An important source of these particles is nucleation of aerosols from water vapors and minor constituents of the atmosphere, such as, for instance, sulfur dioxide. Theoretical estimates of the nucleation rate in the electrically neutral atmosphere give the values much lower that those observed in reality. Therefore, of high importance is the fact that the presence of charged particles reduces the nucleation threshold and stabilizes embryonic particles [Turco et al., 0]. As a result, nucleation develops at a lower concentration of condensing vapors than in the unionized atmosphere. The ability of embryonic particles of aerosols with a size of 1-2 nm to grow into nuclei with sizes of ~100 nm is determined by the summary effect of the processes of condensation, coagulation, and scavenging of the particles. Calculations have shown that the growth rate of charged particles with sizes from 1 to 5 nm is twice as high as that of uncharged particles [Yu and Turco, 2001]. Since the loss rate due to coagulation for the particles with a radius of 5 nm is lower by a factor of 20 than that for the particles 1 nm in radius [Carslaw et al., 2002], the accelerated growth of charged particles results in their higher "survival" at the early stage of formation.
[38] Model calculations show that a 20% change in the ionization rate in the lower atmosphere leads to a 5-10% change in the concentration of the particles with diameters of 3-10 nm [Turco et al., 2000; Yu, 2002; Yu and Turco, 2001].
[39] In turn, the increase in the number of relatively small water droplets in the cloud must lead to an increase in its reflectivity, decrease in the intensity of precipitation from it and hence to an increase in the cloud lifetime, which is indeed observed [Brenguier et al., 2000].
[40] Laboratory investigations show that ions can, indeed, be the sources of aerosol particles [Vohra et al., 1984]. Of course, many details of this mechanism are still unclear. Nevertheless, the data described above indicate that it has rather a sound experimental and theoretical basis.
It is known that atmospheric ozone, which actively absorbs infrared radiation of the Earth, plays a significant role in formation of the global climate via the influence on the radiative, thermal, and circulation regimes in the stratosphere [Kondratyev, 1989]. The strongest influence on the ozone layer state is exerted by the solar UV radiation ( l < 242 nm) and cosmic rays. Since these sources undergo significant changes during both short-term heliospheric disturbances and solar cycle, corresponding variations in the atmospheric ozone content can be expected. Along with this, experimental data show that the link between the solar activity level and ozone content in the atmosphere is extremely complicated. For instance, Kondratyev and Nikolsky [1995] point out that the relation between the sunspot number and ozone content was positive from 1921 to 1928, it was negative from 1933 to 1958, and then this relation again became positive. This complicated and, at first glance, unstable connection between the ozone content and solar activity can be explained, first of all, by the fact that changes in the ionizing radiation intensity give rise to a chain of photochemical processes in the atmosphere that are responsible for both formation and decay of ozone molecules.
[41] Detailed investigations of disturbances in the lower atmosphere caused by penetration of energetic particles of the solar or galactic origin were performed by Hauglistaine and Gerard [1990]. According to the model suggested by these authors, penetration of energetic particles into the atmosphere gives rise to ionization and dissociation of N2 and O2 molecules. The N2+, O2+, NO+, N+, O+, and other ions then participate in the reactions of ion-atomic exchange and recombination, one of the products of which is nitric oxide NO. The latter catalytically destroys ozone molecules in the reactions
 |
 |
Thus invasion of energetic particles into the atmosphere
causes depletion of ozone and formation of a large amount of
nitrogen dioxide. This, in turn, results in significant changes in the
radiation budget of the atmosphere. In particular, the flux of solar
ultraviolet radiation with a wavelength of
l < 3250 Å
increases in
the lower atmosphere and at the Earth's surface due to its weaker
absorption by ozone. At the same time, the flux of radiation in the
blue-green region decreases due to increasing absorption of the
latter by nitrogen dioxide whose absorption cross section has a
maximum
s =6 10-19 cm
2 at a wavelength of about 4000 Å.
The calculations performed by
Hauglistaine and Gerard [1990]
have shown that a sufficiently intense solar proton event
(of the type of SPE that occurred on 4 August 1972) causes more
than a tenfold decrease in the ozone content and increases the
NO
2 concentration by about 2 orders of magnitude at an altitude
of 30-35 km. Because of the enhanced absorption of solar
radiation by nitrogen dioxide, the air temperature in the
stratosphere grows and reaches at an latitude of 30 km a
maximum of 300 K, which is 80 K higher than the norm. On the
contrary, in the troposphere the deficiency of solar radiation
causes a
10o reduction in temperature.
[42] Haigh [1996] has shown that the mechanism involving variations in UV radiations may cause significant shift in the stratospheric circulation and climate through the stratospheric ozone production. In this respect, Hoyt and Shatten [1997] note that UV variations are an excellent candidate for solar variability influences on climate, not only because solar spectral irradiance fluctuations are proportionally larger at short wavelengths, but also because they carry a significant fraction of the total solar energy variability (about 20% below 300 nm according to Lean [1991]). In a climate model experiment, Haigh [1996] analyzed the response of the atmosphere to the 11-year solar activity cycle. In this simulation, a small increase in UV radiation caused substantial stratospheric heating, as the excess in ozone absorbed more sunlight in the lower stratosphere. In addition, the stratospheric winds were also strengthened and the tropospheric subtropical jet streams were displaced poleward. The location of these westerly jets determines the latitudinal extent of the Hadley cells, and therefore the poleward shift resulted in similar displacement of the descending limbs of the Hadley cells. This ultimately led to a poleward relocation of the midlatitude storm tracks. The temperature changes in the model result were very similar, although smaller in magnitude, to the observations by van Loon and Labitzke [1994]. Moreover, the results of Haigh [1996] were supported by an analysis of Christoforou and Hammed [1997], showing a close correlation between solar activity, as expressed by mean annual sunspot numbers, and the intensity and locations of low- and high-pressure centers in the North Pacific area. Similar processes also develop in the stratosphere when the solar X-ray radiation intensity grows.
[43] Thus these are the most widely discussed physical mechanisms linking variations in solar activity to changes in the lower atmosphere state. In order to understand to what extent these mechanisms explain the phenomena observed in the lower atmosphere, let us consider in more detail the morphology of these phenomena. To exclude from the analysis, to a greatest possible degree, the variations in the atmosphere parameters caused by intra-atmospheric processes, we begin with discussing short-term (with duration of the order of several days) atmospheric disturbances associated with different manifestations of solar activity.
Citation: (2004), Influence of solar activity on the lower atmosphere state, Int. J. Geomagn. Aeron., 5, GI2007, doi:10.1029/2003GI000060.
Copyright 2004 by the American Geophysical Union