The Electric State of the Surface Atmosphere in the Mountain–Steppe Landscapes of Southern Siberia According to the Measurement Data in the Khakass–Tyva Expedition in 2022

: Currently, many researchers are interested in investigating the electric ﬁ eld in the fair-weather electric environment, along with its diurnal and seasonal variations across all regions of the world. However, a similar study in the southern part of Siberia has not yet been carried out. In this regard, this study aims to estimate the mean values of the electric ﬁ eld and their variations in the mountain and steppe landscapes using the measurement data from the Khakass–Tyva expedition in 2022. The maximum values of positive ion density were noted at the site in the Iyussko-Shirinsky steppe between Belyo and Tus salt lakes in the Khakass-Minusinsk Basin. The maximum values of negative ion density were observed at the site in the Shol tract in the center part of the Tyva depression. The potential gradient tends to increase with altitude and reaches a maximum in the highlands. The maximum values of the potential gradient were noted in the highlands plateau near the Mongun-Taiga Mountain Massif and Khindiktig-Khol Lake. The diurnal cycles of potential gradient at di ﬀ erent observation sites were divided into two groups: (1) a diurnal cycle in the form of a double wave; and (2) a daily cycle with a more complex course due to the strong in ﬂ uence of local factors.

For example, studies carried out at the beginning of the 20th century on board the Carnegie geophysical vessel made it possible to discover one of the global mechanisms that affected the electric field of the atmosphere [1].This is the average daily variation in the potential gradient, widely known as the Carnegie curve or unitary variation.It represents the global daily contribution of electrical activity in areas of disturbed weather [1,41] and follows universal time, globally independent of the measurement position [1].
However, the contribution of regional and local factors can significantly affect the diurnal variations in the surface electric field in different regions of the Earth.There is a strong relationship between atmospheric aerosols and atmospheric electrical characteristics: air ions sink on aerosol particles, which leads to a decrease in conductivity, and, consequently, the electric field potential gradient should increase.Thus, in the polluted urban environment, the surface values of the potential gradient are higher than those in the countryside and are subject to additional fluctuations due to changes in aerosol concentrations in the boundary layer.This effect has been found in many locations in the world [2,6,9,[42][43][44].Rising smoke plumes from wildland fires [45], intense dust storms [46,47], and eruptive clouds from volcanoes [48] spreading in the middle and upper troposphere are all factors that act above the boundary layer and contribute to a decrease in the potential gradient near the surface.
Meteorological factors have a special place in the study of atmospheric electricity.They are characterized by greater temporal and spatial variability.Fog, mist, and haze contribute to an increase in the surface value of the potential gradient [9,10,49].Short-lived (up to several hours) convective clouds, especially cumulonimbus clouds, cause shortterm deviations of the potential gradient, both positive and negative [11,[35][36][37][38][39][40].Mesoscale convective systems and cloudiness associated with atmospheric fronts and tropical cyclones, which are accompanied by strong thunderstorm activity, cause a strong disturbance of the normal atmospheric electric field over long periods of time and space [50][51][52][53].
Diurnal variations in the potential gradient on fair-weather days recorded in different regions of the globe are generally divided into three types.Diurnal variations having a minimum of about 04 UTC and a maximum of about 19 UTC (as in the case of the Carnegie curve) are Type 1. Variations having two minima, one at ~02 UTC and another at ~10 UTC, and two maxima, one at ~06 UTC and the other at ~19 UTC, belong to Type 2. Diurnal variations with a wide depression centered at ~11 UTC belong to type 3 [6,13,54].
Another quantity characterizing the electrical state of the atmosphere is the electrical conductivity of air, which is more than 90% determined by small air ions [55].Air ions play a key role in atmospheric chemistry, taking part, for example, in ion-catalyzed and ion-molecular chemical reactions and in the formation of aerosol particles induced by ions [56].The electrical conductivity of air also depends on many factors, both global and local.The formation of ions in the atmosphere (air ionization) is mainly associated with ionizing radiation (gamma radiation) and particle fluxes (alpha, beta, neutrons, etc.).In the surface layer, natural sources of ionizing radiation are, first of all, the spontaneous decay of radionuclides, in particular, radon and its daughter products.In the free atmosphere, with increasing altitude, galactic cosmic rays become the predominant source of ionizing radiation.
At the same time, it should be taken into account that in the troposphere, air ionization significantly depends on the geographical location and meteorological conditions.According to [57], fluctuations in the ionizing capacity of the environment are due to the dynamics of the mixing layer, soil type and moisture content, meteorological conditions, atmospheric transport, the presence and change in snow cover, and precipitation.At the same time, measurements carried out in Paris [58] and Shanghai [59] show that aerosol particles reduce the concentration of small ions.Cloudiness [60] and fog also lead to a decrease in the concentration of ions.Studies carried out in [61,62] showed that seasonal fluctuations in ionization were caused mainly by the presence of various air masses with relatively different chemical compositions.
As can be seen, the electrical state of the atmosphere varies greatly depending on various regional and local factors.Therefore, to fully understand the functioning of the GEC and its connection with modern climate changes, observations and analysis of the variability of atmospheric-electric quantities in different regions of the Earth are necessary.
The variability of atmospheric-electric quantities in electrically undisturbed atmospheric conditions (fair-weather conditions) in Siberia today remains poorly understood.This is especially true of Southern Siberia, which has a complex relief and includes various natural zones and types of landscapes.In this regard, the purpose of this work is to estimate the average values and variability of atmospheric-electric quantities in fair-weather conditions in Southern Siberia.

Object and Approach
Our research is based on field measurements of atmospheric-electric quantities in July-August 2022 in Southern Siberia (a central part of Eurasia).The region of this study was located in Tyva and Khakassia Republics, Russia (see red box in Figure 1a).The region is characterized by unique physical and geographical conditions.The measurements were carried out at four observation sites (see site numbers in the Table 1, and numbers and stars in the Figure 1b) at different altitudes above sea level and with different forms of landscape.Site 1 was located near Khindiktig-Khol Lake and the Mongun-Taiga Mountain Massif, the landscape of which is mountainous tundra and alpine meadows (Figure A1).Site 2 was located in the Bayan-Tala Tract, which is a steppe in the foothills of the Tannu-Ola Ridge (Figure A2).Site 3 was located in the Shol Tract, represented by a semi-desert landscape in the center of the Tyva depression (Figure A3).Site 4 was located in the Krasnaya Sopka Tract between Belyo and Tus salt lakes in the Khakass-Minusinsk Basin, represented by a steppe landscape (Figure A4).For measurements at each of the observation sites, open homogeneous areas with a landscape characteristic of the area, not subject to anthropogenic impact, were selected.The location of the measuring equipment at each of the sites is shown in Figure A5.
At each of the sites, measurements were made under the conditions of an electrically undisturbed atmosphere of the main atmospheric-electric and meteorological quantities, aerosol content, and gamma radiation background.In addition, video recording of the sky state and measurements of solar irradiance in the UV and visible ranges were carried out.Observations at each site ranged from 4 to 7 days (Table 1).In general, the obtained estimates can be considered representative of these territories and the season of the year.The only exception is the estimates of ion density variability at site 2, which are based on measurement data lasting less than a day (due to technical problems) and need to be refined.

Instruments
The positive (n+) and negative (n−) ion densities were measured with two air ion counters, AIC2 (Alpha Lab, Salt Lake City, UT, USA) [64].This instrument is a true ion density meter based on a Gerdien tube condenser design, and it contains a fan that draws air through the meter at a calibrated rate.A counter measures the number of positive or negative ions in the air whose mobility is greater than 1 cm 2 •V -1 •s -1 .The maximum measurement of ion density is two million cm −3 ; the accuracy is about 20%.The counters were placed on tripods at a height of 1 m, and the concentration of air ions was measured with a time averaging 1 s.
The electric field potential gradient (∇φ = −Ez = dφ/dz, where Ez and φ are the strength vertical component and potential of the atmospheric electric field; the values under fairweather conditions are positive) was measured using a portable electric field mill EFS-2/50 (NTCR, Moscow, Russia).The field mill was preliminary calibrated using a calibration stand (plate capacitor) and brought to the readings of an electric field mill CS110 (Campbell Scientific, Logan, UT, USA) [65] operating in monitoring mode at the geophysical observatory of IMCES SB RAS [66].EFS-2/50 measures the potential gradient in the range of 0…±20 kV/m with an error of ±1%.The field mill was placed on a grounded tripod at a height of 2 m and measured with a time averaging 1 s.
The main meteorological quantities (air temperature (t) and relative humidity (f), wind speed (V) and direction (D), atmospheric pressure, amount of precipitation, global solar radiation (SI), and UV index) were measured with 1-minute averaging using an AW003 automatic weather station (Amtast, Lakeland, FL, USA).
A NILU-UV-6T multichannel filter radiometer (Geminali, Oslo, Norway) was used to measure solar radiation at 305, 312, 320, 340, and 380 nm and in the range of 400-700 nm, as well as to determine the average and maximum values of solar irradiance in the ranges of UV-A (315-400 nm), UV-B (280-315 nm), PAR (400-700 nm), erythemal and biologically active UV radiation, total ozone column, and transparency (CLT) of the atmosphere at 340 or 380 nm due to the cloud cover and aerosols.In our case, the CLT was calculated at 380 nm as % 100 where Ee(meas.) and Ee(clear) are the measured and model (for a clear sky) irradiance at 380 nm.
Mass concentrations of particles with a diameter of less than 2.5 and 10 µm (PM2.5 and PM10) were measured using the SDS011 dust sensor (Shandong Nova Technology, Weifang, China).SDS011, using the principle of light scattering, measures the concentration of 0.3-10-µm particles in the air according to which concentrations of PM2.5 and PM10 values are calculated.The error in determining PM2.5 and PM10 is 20%.
The equivalent dose of gamma radiation (accumulation time of 3 h) was measured by a DT9501 radiation scanner (Shenzhen Everest Machinery Industry, Shenzhen, China).DT9501 is a precision instrument that measures not only gamma but alpha, beta, and Xray radiations as well.The operation of the device is based on a Geiger-Muller counter converting ionizing radiation energy into electrical impulses and radiation dose.
Radon and its decay products were measured with a Ramon-02 radiometer (Solo LLP, Almaty, Kazakhstan).The radiometer is a portable instrument measuring in semiautomatic mode the equivalent equilibrium volumetric activity of radon in the air in the range of 4-5•105 Bq/m 3 with an error of ±30%.

Data Processing
The mathematical processing of measurement data included data sampling and statistical analysis.To interpret the variability and correlation of meteorological and geophysical quantities at different observation sites, the measurement data were synchronized.Then, multi-day average hourly values were calculated.Using these data, statistical processing, correlation, and regression analyses of the daily variability in meteorological and geophysical quantities were carried out.
Also, the average hourly values of n± and ∇φ were used to calculate the conduction current density (Jλ) in the atmosphere using the formula [23]: where e is an elementary charge (1.6•10 −19 C); k+ and k− are mobility of positive and negative air ions, which were taken equal to 1.2 The main statistical estimations of the variability of atmospheric-electric quantities at observation sites in fair-weather conditions are presented in Tables A1-A3.All confidence intervals presented in the figures in this article have a 95% confidence level.
According to Figure 3, the maximum values of n+ during the days are usually observed at night at 23-03 local time (LT), and the minimum-in the daytime at 10-19 LT.At the same time, during sunrise at 04-06 LT and after one at 05-08 LT, the secondary minimum and maximum were observed, respectively.At site 3, the secondary maximum n+ was comparable to the main one, and at site 4, it even exceeded it.In addition, at all observation sites in the afternoon at 15-17 LT and at site 4, as well as in the evening at 19-21 LT, an increase in n+ was noted.
According to Figure 5, in the daily variation in n− at observation sites, in general, increased values were noted at night, and decreased values were observed during the day.However, in contrast to n+, the daily variations in n− at observation sites differ significantly from each other.Thus, at site 1, the maximum values of nwere observed in the periods of 23-02 LT, and the minimum values were observed at 17-20 LT.At the same time, the secondary maximum n-turned out to be higher than the main one.At site 3, the maximum values of n− were noted at 22-02 LT, and the minimum values at 11-16 LT.In addition, a secondary maximum was also observed at this site at 17-18 LT.At site 4, the maximum values of n− took place at 21-23 LT, and the minimum values at 02 and 04 LT.The very large confidence interval at 21 LT at site 2 is due to technical problems in the operation of the air ion counter and, accordingly, the small sample length.

Variability and Diurnal Variation in ∇φ
As seen in Figure 6 and Table A3, the average value of ∇φ at site 1 is 120 V/m (the median is 110 V/m), and its typical range of variability limited by the interval P25-P75 is 100-130 V/m.At site 2, the average ∇φ is equal to 75 V/m (the median is 69 V/m), and the typical range of variability is 60-80 V/m.At site 3, the average value of ∇φ is 92 V/m (the median is 89 V/m), and, as a rule, one varies in the range of 80-110 V/m.At site 4, the average value of ∇φ is 77 V/m (the median is 68 V/m), and its typical range of variability corresponds to 50-90 V/m.The probability distribution of ∇φ at all observation sites can be approximately described by the normal function (Figure 6).At the same time, at site 4, an additional mode is superimposed on this distribution, the values of ∇φ, which are more than two times lower in the main mode.
According to Figure 7, at sites in the daily course ∇φ, the main maximum is noted in the afternoon at 13-17 LT, and the main minimum is after midnight at 00-01 LT.In addition, in the predawn hours at 03-04 LT and at sunrise at 05-07 LT, secondary maximum and minimum are observed, respectively.According to the classification [6,13,51], the observed diurnal variation in ∇φ can generally be attributed to the continental type with two maxima and two minima.In addition to the observed maxima and minima, during the period at 09-12 LT with intense heating of the surface and boundary layer, a rapid increase in the potential gradient is observed, followed by stabilization or a slight decrease ("shoulder"), caused by a convective generator [3] (Figure 7).
At sites 2 and 4, during dawn and/or sunset, a significant increase in ∇φ was also observed (Figure 7).Moreover, at site 4, the value of ∇φ during these periods exceeded the main (afternoon) maximum.The cause of this effect is presumably the radiative cooling and formation of haze (fog) observed in anticyclonic conditions.Our assumption is confirmed by the daily variation in relative humidity at observation sites, the maximums of which, in general, are consistent with increasing the values of ∇φ.Due to orographic conditions and the relative proximity of water objects, increased air humidity is observed at these sites, which contributed to the formation of haze (fog).Thus, the average daily values of relative humidity at these sites were significantly higher (more than 10%) than the values at other sites.
Hourly average absolute and relative (% of the mean) values of the potential gradient under fair-weather conditions at observation sites are shown in Table A8.

Variability and Diurnal Variation in Jλ
The obtained values of Jλ at site 1 are in the range of ( 12

Discussion
The results of our study of the daily cycle of small ions in the territory of Southern Siberia generally coincide with the results of similar studies conducted in other regions [67,68].As in these works, the minimum ion concentrations we observed during the day and the increased concentrations in the evening, with a maximum at night, can be explained by the diurnal variability of radon concentrations and turbulent mixing during the day.These effects were reflected in the results of our observations.So, the observation sites 1 and 4 were located near relatively large lakes (Figures A1 and A4) where breeze circulation was observed when, during the day, the wind (an advective transport) moved from the water surface to the land and at nightvice versa (Figures A10j and A13j).
Since the main source of small ions on land is radon released from the soil, due to the transfer of "clean" air from the water area, where there is no radon emission, n+ and n− during this period in the surface layer will decrease.In addition, thermal turbulence, more intense in the daytime over land, enhances vertical mixing, which also leads to a "dilution" of the radon concentration in the surface layer and, accordingly, a decrease in the concentration of small ions.When southeastern advection from the mountain ridge at site 1 and western advection from the steppe at site 4 were observed, the ion concentration increased significantly.
As can be seen in Figure 3, the highest values of n+ were observed at site 4 in the steppe, located close (about 5 km) to salt lakes, which presumably can influence n+.Also, increased values of n+ in the daytime are observed at site 1 in the highlands near the Mongun-Taiga Ridge.In general, there is a similarity in the daily variations in n+ at all observation sites.The Pearson correlation coefficient for sites 3 and 4 is 0.77, and for sites 1 and 3, it is 0.36 (Table С1).The highest values of n− are observed at site 3 in the Tyva depression, and the lowest values-at site 1 in the highlands.As can be seen in Figure 5, daily variations in n− at observation sites are weakly consistent with each other.The correlation for sites 1 and 3 is 0.35, for sites 1 and 4-0.30, and for sites 3 and 4-0.33 (Table A6).
A comparative analysis of the average values of ∇φ at observation sites showed the following.In general, the increase in the absolute altitude above sea level of observation sites coincides with an increase in the average (median) values of ∇φ.A similar relationship was previously noted in the Caucasus [69] and Tibet [70].The values of ∇φ are somewhat different only at site 2 in the Bayan-Tala tract located at the foot of an extended 1500m mountain ridge.The reduced values here, we believe, are explained by the local influence of the ridge on the bending of the electric field lines.As can be seen in Figure 7, good agreement between the daily variations in ∇φ is observed between sites 1 and 3, where the correlation is 0.71 (Table A7), despite the fact that these sites are located in areas with different topography (plateau and depression), at a considerable distance from each other (about 270 km), and at different altitudes above sea level (with a difference of more than 1500 m).The observed high correlation of daily variability is presumably explained by the influence of the Carnegie curve [1,10].The similarity of daily variations in ∇φ is also noted at observation sites 2 and 4 (the correlation is 0.48 (Table A7)).There is an increase in the values of ∇φ at dawn and sunset at these sites, which is not observed at sites 1 and 3.
Based on the type of daily cycle of ∇φ, observation sites can be divided into two groups: (1) The sites with a daily cycle for typical continental regions, having two maxima and two minima (sites 1 and 3); (2) The sites with a more complex daily cycle due to the strong influence of local factors (sites 2 and 4).
The first group includes observation sites with a relatively dry climate located in open, slightly winding terrain (the central part of the basin, a high mountain plateau).The second group, on the contrary, includes sites with a more humid climate, located in areas with complex terrain (mountain valley, shady slope of a mountain range), as well as near large bodies of water.
The obtained estimates of atmospheric-electric quantities at observation sites, in general, are consistent with estimates of the characteristics of the radiation background at these sites.Thus, high values of n+ at sites 1 and 4 correspond to high values of γ-radiation dose (Figures 2, 3 and A6).At sites 2 and 3, which are characterized by relatively low n+ values, the γ-radiation dose is also low.In this case, the distribution of average values of volumetric radon activity at observation sites is the inverse of γ-radiation dose and is in better agreement with the values of n− than with n+.The minimum values of radon volumetric activity, as well as n−, are noted at site 1 and the maximum at site 3 (Figures 4, 5  and A7).A presumable explanation for the noted difference in the average values of volumetric radon activity at observation sites is the different permeability of the soils at the sites.The sandy soil at site 3 facilitates the emission of radon, and the rocky surface at site 1, on the contrary, hinders it.Based on the foregoing, the above-mentioned increase in average values of ∇φ with altitude can presumably be explained by a decrease in radon emission, which is associated with a change in the type of soil with an increase in the absolute altitude of the area.In the daily variation in γ-radiation dose and volumetric activity of radon, as well as in the daily variation in n+ and n−, the maximum, in general, occurs at night and early morning hours, and the minimum-in the daytime.Only the daily variation in the volumetric activity of radon at site 1 stands out from this dependence.This is presumably due to the weak emission of radon directly at this site and its transfer from adjacent territories.The daily variation in ∇φ, in general, is opposite to the daily variation in the above values, which is explained by the inverse relationship between the potential gradient and the electrical conductivity of the air, which is determined by n+ and n−.
There is also some consistency between atmospheric-electric quantities, the aerosol content in the air, and the transparency of the atmosphere.The maximum values of PM2.5 and PM10 in the surface layer (at a height of 1 m) were recorded at site 4, which has the lowest altitude above sea level and is located on a steppe area near salt lakes, and the minimum values were recorded at site 1, located at a high mountain plateau (see Figure A8).Lower aerosol content was observed at site 2, located at the foot of the mountain range.Atmospheric transparency at a wavelength of 380 nm, on the contrary, was at its maximum at site 1 and minimum at site 4 (Figure A9).At the same time, a direct relationship was noted between the absolute altitude of the sites and the average values of CLT380, which is most clearly manifested in the afternoon (13-15 LT).Since the measurements were carried out mainly in clear and partly cloudy weather, the increase in CLT380 with altitude is mainly due to a decrease in the aerosol content with altitude.The latter should lead to a decrease in the electrical resistance of the atmosphere and an increase in the conduction current from the ionosphere to the Earth's surface.The noted feature, in general, is consistent with the results obtained, according to which Jλ in the daytime increases with increasing absolute altitude of the area (Figure A9).
Next, we will estimate the variability and relationship of hourly average values of atmospheric-electric quantities during the day with the main meteorological quantities (t, f, V, D), SI, CLT380, PM2.5, and equivalent dose of gamma radiation at each observation site.The average daily variations in these quantities are shown in Figures A10-A13, and the results of correlation analysis are presented in Tables A9-A12.
Basically, at observation sites, the daily variability of n+, n− and ∇φ correlates with the variability of other measured quantities as moderate (R = 0.4-0.6)and strong (R > 0.6).However, there is a weak relationship (R < 0.4) at some sites.
Finally, we compared the data from our expeditionary measurements of ∇φ with data from other studies conducted, in particular, in mountainous regions.First of all, we compared the average daily variations in ∇φ at our observation sites with the unitary variation in ∇φ (the Carnegie curve).So, the main maxima in daily variations in ∇φ at sites 1 and 3 are observed 9-10 h before the main maximum of the Carnegie curve and at sites 2 and 4-12-13 h before.The secondary maxima observed in daily cycles around 3-4 LT (20-21 UTC) at all four sites coincide with the main maximum of the Carnegie curve.This coincidence is especially clear at site 1 on the high mountain plateau.Since secondary maxima of ∇φ occur in the early morning hours before dawn, we cannot explain their appearance by the influence of local factors at observation sites.We assume that their occurrence is associated with the influence of unitary variation.However, additional research is required to confirm or refute this.
A comparison of daily variations in ∇φ at expedition observation sites with daily variations in ∇φ at stationary observation sites in regions with similar physical-geographical conditions and time zones (from UTC+5 to UTC+8) showed the following.Daily variations in ∇φ at our observation sites, especially in the afternoon hours, are qualitatively similar to the daily variations observed at stations located on the West Siberian Plain and the spurs of the Kuznetsk Alatau [21], the Tibetan Plateau [22], and the foothills and slopes of the Himalayas in Northern India [69] and Pakistan [70,71].At all continental observation sites under good weather conditions, the increase in ∇φ in the daytime and the appearance of the main maximum is caused by radiative heating of the surface, the amplification of con-vective movements, increased turbulent mixing, and, accordingly, the spatial redistribution of aerosols.Therefore, the formation of the main maximum ∇φ corresponds to the afternoon, when convective-turbulent mixing reaches its maximum.Only according to expeditionary observations, maxima ∇φ occur at 06-09 UTC, and according to stationary observations, at 13-15 UTC.We believe that this difference is due to the fact that stationary observations are carried out in industrial areas, where an increased content of anthropogenic aerosols is recorded, and expeditionary observations were carried out far from sources of aerosol pollution.

Conclusions
Using the field measuring data in the southern part of Siberia in the mountain-steppe landscapes of Khakassia and Tyva in July-August 2022, estimates of the general and daily variability of atmospheric electrical quantities under electrically undisturbed atmospheric conditions were obtained.
The maximum values of n+ were noted at the site in the Iyussko-Shirinsky steppe between Belyo and Tus salt lakes in the Khakass-Minusinsk Basin.The maximum values of nwere observed at the site in the Shol tract in the center part of the Tyva depression.
The values of ∇φ tend to increase with altitude and reach a maximum in the highlands.The maximum values of ∇φ were noted in the highlands plateau near the Mongun-Taiga Mountain Massif and Khindiktig-Khol Lake.
The maximum values of n+ during the day at observation sites are mainly observed at night (23-03 LT), and the minimum values are observed during the day at 10-19 LT.At 04-06 and 05-08 LT, a secondary minimum and maximum of n+ are observed, respectively.The daily cycles of nare similar to n+ but differ from each other at different observation sites.
The main maximum of ∇φ is observed in the afternoon at 13-17 LT, and the main minimum-after midnight at 00-01 LT.In addition to the main extremes of ∇φ, at 03-04 and 05-07 LT, there are secondary maximums and minimums, respectively.An increase in humidity at dawn (sunset) led to an increase in ∇φ.
The diurnal cycles of ∇φ at different observation sites can be conditionally divided into two groups: (1) a diurnal cycle in the form of a double wave; and (2) a daily cycle with a more complex course due to the strong influence of local factors.
The Jλ values at the observation sites varied in the range of 10 −12 -10 −9 A/m 2 ; the maximum of Jλ in the daily variation, as a rule, was observed at night at 23-03 LT, and the minimum during the day at 10-19 LT.Table A9.Сorrelation сoefficients (R) for the hourly average quantities measured at site 1.

Figure 1 .
Figure 1.Locations of the study region (red box on panel (a)) and observation sites (numbers and stars on panel (b)).The maps are shown based on the global digital elevation model ETOPO2 [63].

Figure 2 .
Figure 2. Box plot (left panels) and histogram (right panels) of the distribution of positive air ion density in fair-weather conditions at observation sites.

Figure 3 .
Figure 3.Diurnal variation in hourly means of positive air ion density in fair-weather conditions at observation sites.

Figure 4 .
Figure 4. Box plot (left panels) and histogram (right panels) of the distribution of negative air ion density in fair-weather conditions at observation sites.

Figure 5 .
Figure 5.Diurnal variation in hourly means of negative air ion density in fair-weather conditions at observation sites.

Figure 6 .
Figure 6.Box plot (left panels) and histogram (right panels) of the distribution of the potential gradient in fair-weather conditions at observation sites.

Figure 7 .
Figure 7.Diurnal variation in hourly means of potential gradient in fair-weather conditions at observation sites.
-130)•10 −12 A/m 2 ; the average is 70•10 −12 A/m 2 , and the median is 40•10 −12 A/m 2 .Respectively, the values of Jλ at site 3 are (8-43)•10 −12 A/m 2 , 43•10 −12 A/m 2 , and 11•10 −12 A/m 2 , and at site 4 they are (12-130)•10 −12 A/m 2 , 70•10 −12 A/m 2 , and 40•10 −12 A/m 2 .According to Figure8, the daily variation in Jλ at observation sites, in general, resembles the daily variation in n+.The maximum values of Jλ are observed at night at 23-03 LT, and the minimum values are observed during the day at 10-19 LT.During sunrise at 04-06 LT and after sunrise at 05-08 LT, secondary minimum and maximum are observed, respectively.Moreover, at site 4, the secondary maximum Jλ exceeded the main maximum both in amplitude and duration.The increase in Jλ at site 4 in the afternoon at 15-17 LT and in the evening at 20-21 LT was also observed.

Figure 8 .
Figure 8.Diurnal variation in hourly means of conduction current density in fair-weather conditions at observation sites.

Figure A3 .
Figure A3.Location of site 3 (red star) in the Shol tract in the center of the Tyva depression.

Figure A4 .
Figure A4.Location of site 4 (red star) in the Krasnaya Sopka tract between Belyo and Tus salt lakes in the Khakass-Minusinsk Basin.

Figure A10 .
Figure A10.Hourly average variations in negative (a) and positive (b) ion densities; potential gradient (c); conduction current density (d); atmospheric transparency at 380 nm (e); mass concentration of PM2.5 (f); air temperature (g); air relative humidity (h); wind speed (i) and direction (j); global solar radiation (k) and equivalent dose of gamma radiation (l) under fair-weather conditions at site 1.

Figure A11 .
Figure A11.Hourly average variations in negative (a) and positive (b) ion densities; potential gradient (c); conduction current density (d); atmospheric transparency at 380 nm (e); mass concentration of PM2.5 (f); air temperature (g); air relative humidity (h); wind speed (i) and direction (j); global solar radiation (k) and equivalent dose of gamma radiation (l) under fair-weather conditions at site 2.

Figure A12 .
Figure A12.Hourly average variations in negative (a) and positive (b) ion densities; potential gradient (c); conduction current density (d); atmospheric transparency at 380 nm (e); mass concentration of PM2.5 (f); air temperature (g); air relative humidity (h); wind speed (i) and direction (j); global solar radiation (k) and equivalent dose of gamma radiation (l) under fair-weather conditions at site 3.

Figure A13 .
Figure A13.Hourly average variations in negative (a) and positive (b) ion densities; potential gradient (c); conduction current density (d); atmospheric transparency at 380 nm (e); mass concentration of PM2.5 (f); air temperature (g); air relative humidity (h); wind speed (i) and direction (j); global solar radiation (k) and equivalent dose of gamma radiation (l) under fair-weather conditions at site 4.

Table 1 .
Description of the observation sites.

Table A1 .
Statistical parameters of the positive ion density variability under fair-weather conditions at observation sites.

Table A2 .
Statistical parameters of the negative ion density variability under fair-weather conditions at observation sites.
* Positive and negative ion densities and conduction current density were measured in less than a day.

Table A3 .
Statistical parameters of the electric field potential gradient variability under fair-weather conditions at observation sites.

Table A4 .
Statistical parameters of the conduction current density variability under fair-weather conditions at observation sites.
* Positive and negative ion densities and conduction current density were measured in less than a day.

Table A5 .
Сorrelation сoefficients (R) for the hourly average positive ion density (cm −3 ) at observation sites.Positive and negative ion densities were measured in less than a day. *
* Positive and negative ion densities were measured in less than a day.

Table A7 .
Сorrelation сoefficients (R) for the hourly average potential gradient (V/m) at observation sites.

Table A8 .
Hourly average absolute (V/m) and relative (% of the mean) values of potential gradient under fair-weather conditions at observation sites.

Table A10 .
Сorrelation сoefficients (R) for the hourly average quantities measured at site 2.

Table A11 .
Сorrelation сoefficients (R) for the hourly average quantities measured at site 3.

Table A12 .
Сorrelation сoefficients (R) for the hourly average quantities measured at site 4.