**3. Empirical data and analysis procedure**

4 Atmospheric Aerosols – Regional Characteristics – Chemistry and Physics

due to tundra and forest fires) on AOD.

**2. Russian actinometric network data** 

climatic forcing and ecology. The study of current spatiotemporal variations in the atmospheric aerosol component is of scientific interest and presents a problem. Current ground-based networks of monitoring (in particular, AERONET) are the results of such interest (Holben et al., 1998). There are eight AERONET stations in Russia; seven of them are located in Siberia [8]. The maps, which show a global distribution of the sources of different anthropogenic, natural, organic, mineral, marine, and volcanic) aerosols arriving in the atmosphere, the total aerosol optical depth in the atmospheric thickness according to model data (IPCC, Climate Change 2001) and the aerosol optical depth according to satellite (MODIS) monitoring (IPCC, Climate Change 2007), show Russia as a territory of decreasing aerosol optical depth (AOD) going from south to north. At the same time, Russia occupies the entire northeastern part of Eurasia ( 30°E –180°E; 50°N – 80°N) and includes different climatic zones which differ in water content, air temperature, cloudiness, solar radiation flux incident upon the land surface, underlying surface, and air-mass circulation. In addition, the density of population and the degree of industrialization of different Russian regions are very inhomogeneous in space. In the studies (Plakhina et al., 2007, 2009) we have shown that an analysis of the AOD of a vertical atmospheric column can be made on the basis of observational data obtained at the Russian actinometric network, in particular, on the basis of data on the integral atmosphere transparency ( *P* ), because *P* variations are, to a great extent, determined by the aerosol component of the attenuation of direct solar radiation; other components of the attenuation (water vapor and other gases) have little effect on its time variations. Thus, on the basis of data on the homogeneous (calibrated against a single standard and obtained with a unified method) observational series of direct solar-radiation fluxes at the land surface and estimates of the integral (total and aerosol) transparency, it is possible to analyze variations in the AOD of a vertical atmosphere. Now we continue this analysis on the basis of an extended database (the number of stations -- 53, and the period of observations – 1976 -2010 years. Now we present the character of multiyear seasonal variations in AOD, the simplest statistical parameters (means, extrema, and variation coefficients) of spatial variations in AOD annual means, the "purification" of the atmosphere from aerosol over the past 15 years (1995-2010 y.y.). Also we compare the effects of the two natural factors (the global factor—the powerful volcanic eruptions in the latter half of the 20th century which resulted in the formation of a stratospheric aerosol layer and the regional tropospheric factor—for example, the arrival of aerosol in the atmosphere

Fig. 1 gives a map showing the location of 53 actinometric stations of the Russian network (Makhotkina et al., 2005, 2007; Luts'ko et al., 2001)for which the AODs of vertical atmospheric columns were estimated for a wavelength of 0.55 μ from the measured fluxes of direct solar radiation at land surface. These stations cover a large part of Russia and are located outside the zones of direct local anthropogenic sources of industrial and municipal aerosol emissions (suburbs, rural areas, uplands, etc.). In other words, the considered spatiotemporal variations in AOD are formed under the influence of natural factors: the The special-purpose Atmosphere Transparency database formed at the Main Geophysical Observatory makes it possible to analyze both the integral and aerosol transparencies of the atmosphere. The stations given in Fig. 1 were selected with consideration for the quality and completeness of the instrumental series. The integral air transparency :

$$\mathbf{P} = (\mathbf{S}/\mathbf{S}\mathbf{o})^{1/2} \tag{1}$$

Where Sis the direct solar radiation to the normal-to-flux surface, reduced to the average distance between the Earth and the Sun and a solar altitude of 30°; *S*0is the solar constant equal to 1.367 kW/m2. The Linke turbidity factor is unambiguously correlated with Р:

$$\text{IT} = \text{lgP} / \text{lgPi} = (\text{lgS} - \text{lgS} \,) / (\text{lgS} - \text{lgSi} \,) = \text{-lgP} / 0.0433 \,\tag{2}$$

The AOD of the vertical atmosphere was calculated with a method specially developed and used at the MoscowStateUniversity meteorological observatory (Abakumova et al., 2006) with consideration for its limitations and errors:

Variations in the Aerosol Optical Depth Above the Russia

from the Data Obtained at the Russian Actinometric Network in 1976–2010 Years 7

**Figure 3.** Spatial distributions of the multiyear means of AOD over the observation periods 1976–1994

The annual values of AOD for each of the stations multiyear means over 1976–1994 years, over 1995–2010 years, their standard deviations are given in Fig.2. Each column of the diagram corresponds to the longitude positionof the station in accordance with Table 1. The means of the AOD characteristics are corresponds to the multiyear (1976–1994 years) annual means of AOD (black), grey corresponds to the multiyear (1995–2010 years) annual means, red corresponds to standard deviations of the annual values of AOD from its mean for each station. A spatial distribution of AOD is shown in more detail in the maps (Fig. 3) drawn by interpolating the data obtained at 53 stations to Russia's territory. For this interpolation, the technologies of the MATLAB 7.5.0. program package were used: there are options to create a uniform grid for the entire region, onto which the given functions Z= F(x,y*)*were projected, where *x* and *y*are the latitude and longitude, respectively, for each of 53 observation points, and *Z* is the AOD mean. In addition, a bilinear interpolation of data was performed. Under bi-cubic and bi-square interpolations, the results, in principle, do not differ from those given in Fig. 3. The spatial distribution of the AOD means over the 35 - year period is in a good agreement with the results of modeling a spatial atmospheric-aerosol distribution, which are given in the IPCC third report (IPCC, Climate Change 2007). The model described in this report takes into account aerosols of different origins anthropogenic and natural sulfates, organic particles, soot, mineral aerosol of natural origin, and marine saline particles) which have certain specific properties of distribution over the globe, and it yields a decrease in AOD over Eurasia from the southern to the northern latitudes in the presence of areas with increased atmospheric turbidity over southern Europe, the Middle East, southeastern Asia,

(upper part) and 1995–2010 (lower part).

AOD={lnS-[0.1886w(-0.1830) + (0.8799w(-0.0094) -1)/ sinh]}/{0.8129w(-0.0021) – 1 + (0.4347w(-0.0321) -1)/sinh} (3)

An index of the Angström spectral attenuation which depends on the size distribution of particles and the coefficient of particle reflection—is assumed to be equal to 1; Sis the direct solar radiation reduced to the average distance between the Earth and the Sun, W/m2 ; and w is the water content of the atmosphere, g/cm2 . The conditions of observations at the stations, as a rule, correspond to the weather of an anticyclonic type (clear or slightly cloudy) when the Sun is not blocked by clouds.
