Влияние факторов космической погоды на работу радиосредств

The problems of space weather impact on the developed technological society, and in particular on operation of radio electronic devices, have recently become especially acute. In connection with the inclusion of computer and robotic technologies in most of our daily life, a natural question arises: how steadily and correctly this electronic equipment (not always controlled by ordinary users) and its software can operate under varying external conditions [Goodman, Aarons, 1990].

The problem arose long ago owing to interference effects in wire systems [Barlow, 1849] and failures in electric power transmission networks [Love, Coїsson, 2016], especially strong at high latitudes.

Nowadays, a sharp increase is being observed in the number of high precision equipment, which sometimes has inconspicuous peculiar properties insignificant under normal conditions. However, under conditions that are different from those expected, such peculiar properties may be critical for functioning of radio electronic devices, including commonly used household appliances [Whiteson et al., 2014].

The problem of space weather impact on radio devices for regular users has become most evident from the analysis of data acquired with global positioning systems, which are currently the de facto basic element of positioning and timing systems. The main function of this system – the precise positioning – turns out to depend on characteristics of a medium. In particular, during geomagnetic disturbances, the systems can go wrong more often and more seriously and sometimes even fail [Afraimovich et al., 2004; Afraimovich et al., 2007; Kim et al., 2014]. This effect manifests itself in positioning of both terrestrial and space objects [Xiong et al., 2016].

Sudden space weather disturbances leading to powerful scattered signals in radars, radio communications, and radiosondes [Bagaryatsky, 1961; Sverdlov, 1982], require developing systems for predicting such interference and reducing the degree of its influence on radio devices.

Thus, the assessment of space weather impact on operation of radio devices, the forecast of its consequences, the readiness for the problems caused by this impact, and the elimination of its possible effects are the urgent tasks facing any technologically advanced society [The Sun to the Earth - and Beyond ..., 2003; Solar and Space Physics ..., 2013]. The interval between putting equipment into operation, emergence of operational problems, construction and putting into operation of new, more reliable equipment in many cases comprises several years. These intervals are especially long for space-based equipment. A natural solution to this problem will be to take into account the possibility of failure and to estimate space weather effects on the final result of operation of this equipment before it is replaced with a new one, as well as to predict possible failure periods.

The problem of assessing the space weather impact on different areas of human activity and reducing its consequences is usually solved in a variety of ways – from implementation of national strategies [Solar and Space Physics ..., 2013; National Space Weather Strategy, 2015], plans [National Space Weather Action Plan, 2015], legislative acts [Obama, 2016], and available information systems [Space Weather – Effects on Technology, 2012] to the participation of enthusiasts and the use of capabilities of household devices and computers (the so-called citizen science [Barnard et al., 2014, Au-rorasourus, 2016; Wikipedia, 2016]). Various monitoring and forecasting systems, both global [ ] and local ones designed for specific aspects of space weather [Love et al., 2016] are stimulated and supported. In-depth reviews of space weather effects on equipment of different types can be found in the monographs [The Sun to the Earth ..., 2003;

Solar and Space Physics ..., 2013; Effects of Space Weather ..., 2004; Goodman, 2005; Space Weather ..., 2007].

The main geo-effective space weather effects extensively studied today include [National Space Weather Strategy, 2015]: solar radio bursts affecting the operation of radio receivers; induced geoelectric fields influencing wired energy supply and communication systems; ionizing radiation affecting the efficiency of electronic equipment and the vital activity of organisms; expansion of upper atmospheric layers leading to an increase in temperature and density of these layers and affecting the dynamics and lifetime of artificial Earth satellites; as well as ionospheric disturbances affecting radio wave propagation and scattering.

The launch of the system of impulse decameter coherent radars at ISTP SB RAS, in particular under the project "National Heliogeophysical Complex of the Russian Academy of Sciences", raises questions of continuous space weather monitoring for solving not only fundamental but also applied problems important for a technologically advanced society.

ISSUES OF SPACE WEATHER FORMATION

The term “severe space weather” has arisen relatively recently to describe the influence of solar and geomagnetic activity on equipment operation and infrastructure performance [Severe Space Weather Events, 2008; Solar and Space Physics, 2013], although space weather effects have long been known [Barlow, 1849]. A decisive influence on the major part of large-scale terrestrial phenomena is exercised by the source of radiation and particles closest to Earth – the Sun. Although there are examples of the feedback effect of human activity on large-scale natural processes and the artificial generation of some natural phenomena [Baker et al., 2014; Gombosi et al., 2017], but the Sun can be considered nowadays as the main and permanent source of space weather formation. The particle and radiation flux from the Sun, highly dynamic in space and time, is associated with internal solar processes, and, since it is impossible to monitor deep processes on the Sun in real time, the flux could only be averagely predicted. A human acts mainly as an observer of solar activity variations and as a researcher of regular processes occurring in the upper layers of the Sun. The 11–12-year solar activity cycles, which are related to reversals of the solar magnetic field and reveal themselves in all its parameters from radio emission ( F 10.7) to the number of sunspots (Wolf number) optically observed for several centuries, are well-known. The axial rotation of the Sun with a period 25–30 days also cause periodic variations in particle-wave radiation fluxes.

The propagation velocity of such fluxes from the Sun to Earth varies: wave radiation propagates at the speed of light and reaches Earth in about 8 minutes, the corpuscular component moves approximately thousand times slower. Thus, if in the first case the motion of radiation in most problems may be considered to be straight and the motion of objects in the solar system may be ignored, for the particle motion we should take into account the Sun rotation, the orbital motion of Earth and its daily rotation, and calculate the particle motion in the resulting complex trajectory in terms of geometry (cone) of particle release from the Sun and distribution of their velocities. Therefore, it is difficult to assess solar particle-wave flux effects, in view of possible delays in appearence of these effects, their extension in time and space (due to the difference in particle velocities and angles of their release from the Sun) [Odstrcil, 2003], as well as the possibility of particle accumulation in Earth's magnetosphere (which leads to additional time delays of the effects). Solving this problem requires creating complex physical models, including numerical, large computational resources, long-term and accurate experiments, as well as a large set of diverse diagnostic instruments at various points of Earth and outer space, which operate in the mode of continuous monitoring and data transfer to centers of their storage, automatic processing, and real-time cosimulation. An essential contribution to the solution of this problem is measurements of these fluxes by the ACE and DSCVR satellites at the Lagrange point L1, at a distance of about 1 million km from Earth. They allow us to increase the accuracy and to detail the short-term predictions of composition and dynamics of particle-wave radiation [Machol et al., 2012] as compared to more forehand, albeit less accurate and detailed predictions based on remote observations of solar activity with various (ground- and space-based) telescopes.

Particle radiation, reaching the boundary of Earth's magnetosphere, interacts with it. Trajectories of charged particles bend significantly, and the particles begin to move under strong influence of the geomagnetic field, generating electric fields and currents in the magnetosphere.

An important role here is played by high-latitude regions around magnetic poles (cusps), where the direction of the magnetic field is close to vertical. This causes charged particles to precipitate from the magnetosphere into this region toward Earth's surface. The observed optical effects – aurora borealis – have been well known for a long time and represent one of the consequences of the arrival of solar disturbance in Earth's magnetosphere. Regular observations of similar effects have also been conducted since the last century and carried out by scientists with various special-purpose optical instruments such as all-sky cameras, photometers, as well as by amateur photographers.

Precipitating particles cause a change in ionizationrecombination processes in the lower layers of the ionosphere (D and E) and, in turn, increase the electron density there. This increase leads to an increase in radio wave absorption, which manifests itself as an amplitude decrease or the loss of radio signals on the paths passing through these regions.

Such effects are monitored by observing the amplitude of radio signals over long radio paths (e.g., with networks of inclined sounding ionosondes or by receiving a signal from broadcasting stations) or the amplitude of radio signals from space sources (e.g., with riometers). A decrease in signal amplitude is also one of the consequences of the arrival of particles and radiation in Earth's magnetosphere.

The voltage difference arising from separation of charges moving in the solar wind in Earth's magnetosphere under the influence of the magnetic field brings about the formation of field-aligned (along magnetic field lines) currents. These currents close through the E layer (90–120 km above Earth's surface), which has a maximum electrical conductivity due to peculiarities of distributions of collision frequencies of charged and neutral particles. The strong horizontal currents forming in the E layer of the polar ionosphere generate magnetic field disturbances recorded on Earth's surface with instruments for measuring the full magnetic field vector – magnetometers. These geomagnetic disturbances, observed since the nineteenth century, are also one of the consequences of the arrival of charged particles in Earth's magnetosphere.

Besides the processes caused by the motion of charged particles in the ionosphere, the geomagnetic field structure changes due to the appearance of additional charged sources. The main manifestation of this effect, associated with the regular particle-wave radiation of the Sun (solar wind), is, of course, the difference between the external geomagnetic field and the dipole field, including the existence of a sunward sharp flat transition region and an antisunward strongly elongated region.

The solar wind can change the size and shape of the magnetosphere. As a result, the geographic area of the phenomena emerging from particle precipitation (auroras borealis, strong ionospheric currents, radio wave absorption) shifts from high to middle latitudes. In this case, we can observe, say, auroras borealis at latitudes of central regions and southern borders of the Russian Federation (up to the 40 degree magnetic latitude), where it is usually not observed [Feldshtein et al., 2010]. The remaining effects (radio wave absorption, strong currents in the ionosphere, and geomagnetic disturbances) demonstrate similar dynamics during intensification of solar wind fluxes.

It is obvious that solar radiation comes much earlier than corpuscular radiation and also influences processes occurring in Earth's upper atmosphere. The main effect is the very existence of the ionosphere – a plasma layer ionized by solar radiation in Earth's neutral atmosphere. Accordingly, any variations in solar radiation cause variations in the ionospheric electron density at heights corresponding to the lines of radiation absorption by gases constituting Earth's atmosphere and ionosphere. Thus, electron density variations at different heights above Earth's surface [Mikhailov, Perrone, 2016] may be one of the consequences of solar wind disturbances.

An important fact is the existence of well separated zones in Earth's ionosphere, magnetosphere, and atmosphere, on the boundary of which characteristics of the medium change drastically. This leads to the existence of eigenoscillations in these zones, which propagate in the medium relatively independently.

Such oscillations may appear as Schumann resonances in the layer between Earth's surface and ionosphere [Schumann, 1952], internal gravity waves in the atmosphere and their effects in the ionosphere [Lognonné et al., 1998], fast magnetosonic waves in the magnetosphere [Leonovich, Mazur, 2008], and so on. Therefore, many processes occurring in the magnetosphere–ionosphere–atmosphere system can be considered as a superposition of eigenoscillations of this system. Those oscillations that least fade out with time exist the longest in this system and produce aftereffects when the system continues to change despite that the cause of these changes has already disappeared.

This system is also characterized by dynamics under the driving force. For example, gravity variations due to the periodic motion of the Moon around Earth lead to the formation of tidal waves, which have an effect not only on the ocean, but also on the atmosphere and ionosphere [Alpert, 1949]; and the motion of the day-night boundary (solar terminator) in the atmosphere, to the formation of internal gravity waves.

Thus, Earth's magnetosphere, ionosphere, and atmosphere have both eigen and forced oscillations, which can lead to the formation of additional disturbances during the periods when the solar wind effect is absent or has already disappeared. This sometimes causes an additional solar wind effect in time and space, including the appearance of the "memory" effect in the magnetosphere–ionosphere–atmosphere system.

SPACE WEATHER IMPACT ON RADIO DEVICES

The operation of radio devices depends on a combination of electromagnetic processes inside and outside the devices. We call effects direct if a malfunction is caused by processes inside a radio device, and indirect if it occurs outside the device.

Direct effects involve the induction of electromagnetic fields in conductors inside a radio device, a change in potentials due to additional ionization by background radiation, the emergence of auxiliary currents due to penetration of additional charges from outside, as well as an increase in the background electromagnetic radiation of various types and concentration of different particles during disturbances. This can cause radio equipment malfunctions under the influence of induced current, which leads to hardware and software malfunctions, a decrease in the signal-to-noise ratio, additional ionization of the equipment by electromagnetic radiation, and particle-induced changes in equipment characteristics.

Indirect effects include changes in the medium of radio signal propagation, such as a change in the refractive index of the ionosphere. In this case, the malfunction of the equipment is associated with a change in the medium it uses to transfer data or operate. Depending on types of medium, indirect effects can be classified as changes of conditions in the magnetosphere, ionosphere, and atmosphere, on or under Earth's surface.

Solar radio bursts

The most intense solar effect is the electromagnetic radiation observable in various parts of the solar spectrum. Solar radio noise and radio bursts (sudden enhancements of radio emission), discovered in the 1940s, have been quite actively investigated to this day [Bastian et al., 1998; Solar and S pace W eather R adiophys-ics ..., 2004; Lee, 2007; Shibasaki et al., 2011].

In addition to the general substantial increase in the level of radio emission, variations in the intensity are possible within the radio burst with periods from milliseconds to seconds [Chernov, 2011]. This leads to an additional increase in the instantaneous intensity of radio emission by tens of decibels compared to the average level of the burst [Benz, 1986], which is already higher than the level of the quiet-Sun radio emission. Due to these features, the main effect of radio bursts is reduced to the occurrence of unexpected interference in radar, radio communication, and radio reception devices [Bala et al., 2002].

Induced geoelectric fields

Geomagnetic disturbances can cause an amplification of currents in the earth's crust, mainly due to the amplification of auroral currents in the polar ionosphere [Boteler, 1994; Pirjola, 2000]. Geomagnetically induced currents affect the stable operation of electric systems [Campbell, 1978; Pulkkinen et al., 2005; Thomson et al., 2011]. As such, they are direct mechanisms of influence on radio electronic devices. These currents are probably the first observable manifestation of the space weather effect on electrical devices [Barlow, 1849].

Table 1

Modes of radio wave propagation in the ionosphere at different frequencies

Range	Frequencies	Propagation mode
ULF	<3 kHz	Waveguide, surface wave
VLF	3–30 kHz	Waveguide, surface wave
LF (LW)	30–300 kHz	Waveguide, surface wave
MF	300–3000 kHz	Surface wave, ionospheric wave
HF	3–30 MHz	Surface wave, ionospheric wave with significant refraction, meteor scatter, hop propagation
VHF	30–300 MHz	ionospheric wave (weakly refractive), Meteor scattering
UHF	300–3000 MHz	ionospheric wave (weakly refractive)
SHF	3–30 GHz	ionospheric wave (weakly refractive)
EHF	30–300 GHz	ionospheric wave (weakly refractive)

Expansion of the upper atmosphere

Monitoring of density, winds, temperature, and composition of the neutral atmosphere is an important task, which is also closely related to space weather monitoring.

Space-based systems provide solutions to a large number of practical problems today. Many of these satellites are low-orbital and affected by the neutral atmosphere, which causes their deceleration and premature orbit reduction. This, in turn, shortens the lifetime of a satellite and complicates its tracking. The neutral atmosphere is mainly controlled by solar activity through surface and atmosphere

At present, some organizations use geoelectric field forecasting systems [Erinmez et al., 2002; Thomson et al., 2011]. The increasing interest in geoelectric fields is associated with effects in electric power networks. These effects often cause long-term malfunctions in the networks in North America, Sweden, and Australia [Béland, Small, 2004; Pulkkinen et al., 2005; Marshall et al., 2011].

Ionizing radiation

The effect of radiation on various electronic devices has been known for quite a long time [Ionizing Radiation Effects ..., 2015]; it reduces to a change in characteristics (constant and temporary) of the devices due to incoming radiation or its related atmospheric processes. This effect can be observed even on consumer devices, such as smartphones [Whiteson et al., 2014].

These effects are most severe in space vehicles. On average, according to CLUSTER data, a solar-radiation-induced decrease in the efficiency of solar panels on board satellites is about 5 % per year. This limits the time of their operation [Keil, 2007]. An even more important effect is the degradation of optical and electronic equipment on board satellites, which may also lead to their loss [Lotóaniu et al., 2015].

Currently, more than 8000 flights a year pass over the North Pole [Space Weather – Effects on Technology, 2012], hence the need to take into account the effect of solar radiation on health of flight personnel, passengers, and on electronic equipment.

heating by solar radiation and through heat transfer from high-latitude regions, in which ohmic heating is caused by processes and currents in the ionosphere and magnetosphere [Buonsanto, 1999]. Besides, the relationship is being studied between sudden stratospheric warming events and effects of expansion of the upper atmosphere [Liu et al., 2013].

Ionospheric disturbances

The mechanisms discussed above are related to the direct effect of "severe space weather" on radio devices. The main mechanism of the indirect effect of space weather on radio devices are ionospheric disturbances [Buonsanto, 1999; Kutiev et al., 2013]. The ionosphere is a partially ionized gas divided into several basic layers (usually denoted by D, E, and F depending on their distance from Earth's surface). It is located at a height of about 60 to 2000 km and has a strong influence on radio wave propagation. The interaction of radio waves with the ionosphere depends on frequency, distance from a receiver to a transmitter, ionospheric conditions, and the underlying Earth surface. A fairly detailed description of the radio wave propagation processes can be found in the monographs [Ginzburg, 1960; Budden, 1988].

The main radio devices affected by space weather are HF radio communication devices, surface-to-space communication systems, global navigation systems, over-the-horizon radars, satellite altimeters, and spacebased radars [Goodman, Aarons, 1990]. The stable operation of most of these devices depends on ionospheric conditions [Cannon et al., 2004].

Table 1 lists the main mechanisms of radio wave propagation in different frequency ranges.

In the lower part of the spectrum (VLF, ULF), radio wave propagation can be described as waveguide propaga- tion in the effective waveguide formed by Earth's surface and ionosphere. In the upper part of the spectrum (SHF, UHF), radio wave propagation can be considered almost rectilinear, weakly affected by the ionosphere. Between these ranges, the ionospheric impact on radio wave propagation is most considerable, and the HF band is worst affected by solar disturbances and is best suited for designing tools to monitor such effects [Goodman, 2005].

Table 2 shows the main functions of radio equipment, indicating corresponding radio ranges.

The ULF range (<3 kHz) has been studied in sufficient detail [Bannister, 1986; Pappert, Moler, 1978]. In its analysis, the ionosphere and Earth are assumed to be ideal, homogeneous, and with sharp boundaries. The ionosphere acts at these wavelengths as an ideal conductor and generally has no effect on propagation of these waves. Nevertheless, the lower part of the ionosphere, especially the sporadic E layer, can influence radio wave characteristics (mainly phase ones) due to the interference of waves reflected from the regular and sporadic layers [Pappert, 1980].

Table 2

Some applications of different frequency ranges

Range	Functions
ULF and VLF	Navigation, time and frequency signals
LF	navigation, broadcasting
MF	amplitude modulation broadcasting
HF	radio communication, standard time signals, radiolocation, amateur radio communication, positioning systems
VHF	Television, broadcasting with signal frequency modulation, aircraft radio communication
UHF and EHF	GPS/GLONASS navigation, radiolocation, television

The basic model for describing oscillation propagation in VLF and LF ranges is waveguide propagation. This approach considers the space between the ionosphere and Earth's surface as a single waveguide. In this case, transverse electric (TE) or transverse magnetic (TM) modes (each with its own characteristics) arise depending on the type of antenna used to generate radiation. However, they are not independent due to the presence of an inhomogeneous magnetic field.

As regards the ionospheric effect on radio wave propagation, the VLF/LF range more strongly depends on ionospheric conditions than ULF, and therefore it is more affected by space weather disturbances. Without considerable space weather disturbances, VLF/LF radio wave propagation is more of less stable. Due to interference of different modes and their propagation conditions, space weather disturbances can cause, for example, sudden phase anomalies (SPA) arising from the enhancement of the D layer during solar flares. When particles penetrate into the polar cap, as during magnetic storms or polar cap absorption events (PCA), propagation conditions also change, resulting in phase and am- plitude signal distortions. Other factors such as ground conductivity have also a strong effect on signal characteristics. These factors are most significant in Polar Regions, which are affected both by the ionosphere and by seasonal dynamics of the underlying surface.

Many researchers combine MF and HF radio ranges because they are characterized by the so-called spatial (sky or ionospheric) wave – a signal path highly refractive in the ionosphere. Moreover, the so-called surface (ground) wave related to radio wave propagation along Earth’s surface can be observed in both the ranges. Weakening of the sky wave, for instance, due to absorption in the D layer, increases the probability of observing the ground wave. However, sometimes it is convenient to consider the MF and LF ranges at a time, for instance, to predict characteristics of emission intensity at these wavelengths [Wang, 1985; Ghasemi et al., 2013]. A review of experimental data on MF radio wave propagation can be found in [Knight, 1983; Vilensky et al., 1983].

In the VHF range, space weather effects are largely related to phase and amplitude variations in propagating signals due to ionospheric irregularities of different scales. If the irregularities have sufficiently small transverse spatial scales, the phenomenon is observed as scintillation [Basu, Basu, 1981; Basu et al., 1985; Aarons, 1982; Priyadarshi, 2015]; however, if they are relatively large-scale, it appears as smooth changes in characteristics of the received signal.

The main ionospheric effects (signal group delay, Faraday fading, Doppler frequency shift, etc.) depend on the integral value of electron density along a propagation path. This value is usually called total electron content (TEC) and can be measured from data obtained by dual-frequency GPS receivers [Klobuchar, 1975]. Most errors (up to 70 %) of global satellite positioning can be corrected by taking into account this ionospheric correction. Even greater success can be achieved by knowing the complete three-dimensional distribution of electron density in real time. Ionospheric effects of global large-scale space weather disturbances can be roughly estimated more easily using the global electron content – the total amount of electron plasma in the entire ionosphere [Afraimovich et al., 2008] derived by integrating TEC maps all over the world.

The HF range intermediate between MF and VHF ranges is the most difficult to describe. This is because the critical (plasma) frequencies of the main ionospheric layers (except the D layer) are within this range, and the gyrofrequencies are comparable with the lower boundary of the range. At the same time, HF radio wave propagation can be described in terms of the geometrical optics (hop propagation) [Ginzburg, 1960] and mode propagation [Kurkin et al., 1981], combining features of VHF and LF ranges. High sensitivity to absorption also makes it similar to the lower frequency ranges, especially MF. At the same time, some HF signals with frequencies above the critical frequency can propagate under weak distortion of their paths, but under the strong influence of polarization effects. This makes them close to the VHF range. The existence of irregularities of the order of wavelength, especially at polar latitudes, leads to strong backscattering by natural plasma irregularities (radio aurora) as in the VHF range. The existence of natural ionospheric irregularities smaller than the Fresnel radius results in amplitude-phase fading (scintillation). Doppler frequency shifts exceeding 1 Hz also make this range similar to VHF and UHF ranges.

Therefore, the space weather impact on the HF range is very strong and includes practically the entire spectrum of effects observable in other ranges: absorption due to ionization during solar flares, absorption in the polar cap, radio aurora, multimode propagation, group and phase delays caused by refraction, Faraday and Cotton-Mouton polarization effects [Ginzburg, 1960; Goodman, 1991], etc.

The extensive use of HF systems requires an understanding of the propagation medium, which in turn is affected by space weather. The strongest space weather disturbance, which affects radio wave propagation, is a geomagnetic storm manifesting itself at all latitudes, including the least disturbed middle latitudes [Akasofu,

1977]. Doppler frequency shifts and signal frequency distortions also depend on the temporal dynamics of the ionospheric channel [Basler et al., 1988] and are responsible for stable receiving. These effects are especially strong at polar and equatorial latitudes.

The main radio equipment that reacts to all or almost all space weather effects is a short-wave over-the-horizon radar with pulsed or continuous signal emission. Their scientific equivalent is SuperDARN pulsed radars [Chisham et al., 2007], ionosonde-direction finders with a continuous chirp signal [Uryadov et al., 2013] or oblique sounding ionosondes [Ivanov et al., 2003]. The principle of the radar operation is to transmit a complex radio signal, which is partially scattered by ionospheric irregularities, partially refracted in the ionosphere, and partially scattered back by Earth's surface. We use data from the Ekaterinburg HF radar (EKB) of ISTP SB RAS to illustrate the space weather impact on radio devices.

OVER-THE-HORIZON RADIOLOCATION AS A METHOD FOR MONITORING SPACE WEATHER EFFECTS

The main tasks of the over-the-horizon radiolocation are to detect and examine characteristics of scatterers beyond the horizon, using radio wave propagation effects and complex algorithms for rejecting noises from natural sources. The influence of a propagation medium on group and phase delays usually remains considerable. Reviews of these radio devices can be found in [Headrick, Skolnik, 1974; Alebastrov et al., 1984; Headrick, 1990; Principles of Modern Radar, 2010 ]. In scientific problems, irregularities of a propagation medium (mainly the ionosphere) serve as a scatterer.

As HF radio waves propagate, the emergence of additional ionospheric layers leads to the emergence of additional propagation paths [Tsunoda et al., 2016], and, as a result, complicates matching of the radar range (group delay of a signal) and azimuth to real positions of scattering objects or directions to them [Reinisch et al., 1997; Berngardt et al., 2016; Chen et al., 2016; Warrington et al., 2016]. The frequency dependence of propagation paths causes strong phase distortions, thus making the detection of complex signals over long paths even more difficult. Moreover, the signal amplitude can vary due to defocusing/focusing of the signal [Berngardt et al., 2016] and its absorption in the lower ionospheric layers [Berngardt et al., 2016; Gauld et al., 2002; Setti-mi et al., 2014; Settimi et al., 2015; Sonnenschein et al., 1997]. All these are supplemented with the previously described effects: changes in group and phase delays and polarization distortions. Even accurate measurements of the speed of scatterers require us to correctly take into account the background ionosphere [Gillies et al., 2011].

The operation of over-the-horizon radio devices is maintained with systems for modeling radio signal propagation in an inhomogeneous magnetized ionospheric plasma [Fridman et al., 2016; Landeau et al.,

1997; Rei n isch et al., 1997; Settimi et al., 2015; Warrington et a l., 2016] because they ca n reduce iono s pheric errors [ R einisch et al., 1997; Bern g ardt et al., 2 0 15b]. To correct l y solve the problem of s ignal propag a tion, we should know the propagation medium or at least have a go o d model of the medium. There are v a rious models fo r predicting and correcting propagation characteristics in applications to diverse radio system s (for example, IONCAP, VOACAP, ICE P AC, ASAPS ) , designed mainly to predict different ion o spheric char a cteristics [Zol e si, Cander, 2014]. Med i an monthly ionospheric m o dels, developed more tha n 50 years ag o and constantly improved [Bilitza et al., 2014, 2017], can also be employed to predict the ope r ation of HF radio devices under different conditions. E xamples of such improvem e nts are NeQuick [Radicella, Leitinger, 2 001] and PIM [ Daniell et al., 1995]. O n e of the co m mon models is the International Referenc e Ionosphere (IRI) adjusted b y different data [Bilitza e t al., 2014, 2 017; Settimi et al., 2015]. Most of these models are e ither statistical or smoothed, thus impedi n g effective m onitoring of l o cal changes, which are m ost pronoun c ed at high latitudes.

The most accurate is the real-time monitori n g of ionospheric characteristics based on d ata from net w orks of instruments and their use to correc t radio propa g ation models [Settimi et al., 2015; Bilitza et al., 2017; Her-nández-Pajares et al., 2017].

The EKB radar (56.5° N, 58.5° E ) , put into tri a l operation in D ecember 2012, is a pulse d decameter c oherent radar installed jointly with the In s titute of Geo p hysics of the U ral Branch of the Russia n Academy o f Sciences (IG P UrB RAS) at the IGP Ur B RAS obser v atory Arti. The transmitter/receiver equip m ent of the radar was devel o ped at the University o f Leicester ( G reat Britain) a n d purchased at the expens e of SB RAS. The installation of the radar array was funded by R osgi-dromet (R u ssian Hydrometeorologic a l Service). A t present, this r adar is the only scientifi c pulse deca m eter over-the-h o rizon radar in the Russian Federation.

The tra n sceiver antenna system o f the radar is a linear phased array; it provides a 3 – 6° beamwidth de p ending on fre q uency and a 50° field o f view scann e d by fixed bea m positions one by one. T h e spatial and temporal resol u tions of the radar are 15 – 45 km and 2 min respectivel y . The 8–20 MHz freque n cy range e n ables the radar to operate as an over-the- h orizon rada r , and the peak p ower of 10 kW ensures i t s operation i n the range up t o 3500–4500 km. Short so u nding signal s provide a low ( about 600 W) mean radar power, thus en a bling it to operat e in the round-the-clock mo n itoring regi m e. An approxima t e field of view of the radar is shown in F igure 1. Refracti v e effects make this sector s o mewhat large r ; and when solving specific problems it is n e cessary to ta k e into account co n ditions of the background ionosphere to c alculate the region from which a signal co m es.

In the geomagnetic storm main pha s e, the ionosp h eric

Figure 1. Approximate field of view of the EKB radar in geographical coordinates, excluding refraction in the ionosphere. Numbers indicate radar beam numbers plasma frequency can decrease to 50 % of its pre-storm value and then recover within a few days. The refraction coefficient, radio signal propagation paths, group and phase propagation delays also vary a great deal.

Electron densi t y variations i n the E and F layers usually cause a radio signal propa g ation path t o distort; the lower is the frequ e ncy, the stro n ger are these d istortions.

The weakest d istortions th at amount to a change of gro up delays in io n ospheric pro p agation or to polarization dist o rtions associ a ted with F a raday fading [Ginzburg, 196 0 ; Budden, 1988] occur in V HF and UHF ranges. In this c ase, the main effect refers t o the errors in determining the range calculat e d from group or phase del a ys. Such errors are peculiar t o various rad a r systems an d are widely kno w n in data o f Global Po s itioning Sys t ems (GPS) [Klo b uchar et al., 1 987]. Farada y fadings can l ead to addition a l variations i n signal powe r , which are n oticeable in radio astronomical observations [ Afraimovich, 2007].

The main fac t or affecting H F radio wa v e propagation is refraction caused by a large-scale ionospheric irre g ularity. To illustrate the e ffect of vari a tions in the bac k ground ionospheric chara ct eristics on a radio signal (Figure 2), we sh o w the power of a scattere d signal as a function of the ra d ar range and time as infer r ed from the EK B radar data fo r three sele c ted days: M a y 19, 2016, Aug u st 30, 2016, and Septemb e r 22, 2016. F igure 2, a – c shows that to t h e nighttime ( 22–24, 00–0 8 LST) correspond large rad a r ranges (zo n e II); and to the daytime (09–19 LST, zo n e I), small r a dar ranges. Figure 2, b indi c ates that diu r nal variatio n s of the rad a r range can exc e ed 1000 km. Figure 2, d explains t h is effect in ter m s of daytime (black color) and nighttime (gray color) refraction. It i s seen that wi t h a nighttim e decrease in the e lectron dens i ty N e (right i n Figure 2, d ) and an increase in its maximum, the tr a jectory becomes longer, and so does the r a dar (group) d elay to the b oundary of the dead zone. T h is effect is a r egular daily one and depen d s on the thr e e-dimension a l electron d e nsity distributi o n over the e n tire propagat i on path of th e signal.

A)

E

Qi

В)

E

С)

E

Qi

1ST. hours

Time

Figure 2 . Power of a scattered EKB r adar signal for May 19, 2016 ( a ), August 3 0 , 2016 ( b ), an d September 2 2 , 2016 ( c ) vs radar range of scattering and time, LST = UT+4; an app r oximate signa l propagation t r ajectory in da y time (black) a nd nighttime

(gray) ( d )

Eigen w aves propagating in the magnetized ionosphere are ordinary and extraordina r y waves wit h different polarizations, depending on the angle wit h the magnetic f ield. When waves prop a gate in the ionosphere at d ifferent velocities, their s u perposition u ndergoes Fara d ay fadings. The fading phase in the fir s t approximation is proportional to the to t al electron c o ntent [Kravtsov, Orlov, 1980].

The ionosphere being a dispersive medium, differ- ent frequencies propagate in it with different velocities. Hence, propagating in the medium, a signal is distorted. Distortions of narrowband HF signals were discussed, for example, in [Zasenko et al., 1993]. Such distortions can be related both to the refraction and signal focusing effects and to the frequency dispersion of the ionospheric absorption. Besides F-layer ionization, electron density can change in the D and E layers, thus causing higher radio wave absorption.

Radio w ave absorption occurs w hen the eff e ctive frequency of collision of electrons w ith other pa r ticles becomes s ufficiently high. Thus th e wave ener g y is transferred to heavier particles – ions and neutrals , converting int o heating. This occurs mos t often during solar flares whe n the electron density in th e D layer inc r eases [Rogov et a l., 2015]. Basically, the a b sorption is h i gh in the HF ra n ge and at lower frequen c ies, while at VLF frequencie s it becomes rather low.

Figure 3, a–f shows the behavi o r of the sca t tered signal po w er during several X-ray fl a res: May 08, 2014 (М5.2), A pril 11, 2013 (М6.5), O ctober 25, 2013 (M9.4), a n d respective variations in X-rays as d e rived from GO E S data. It is seen that often during X -ray flares the s ignal power can decrease substantially without changi n g the scattering range; he n ce, the main e ffect is associat e d with the appearance o f an ionized r e gion that has weak refraction but high abs o rption, such a s the D layer. Fi g ure 3, g illustrates a chan g e in signal in t ensity during propagation in the presenc e of a strongl y ionized lowe r region; Figure 3, h , in its absence . The dashed lin e in Figure 3, g schematic a lly shows th e signal amplit u de that decreases as the signal pass e s the absorbing l ayer. The signal trajector y changes sli g htly, but the sig n al amplitude drops consid e rably as the s ignal propagates in the absorbing layer.

Ionization of lower ionospheric layers, which is caused by energetic solar protons and leads to radio wave absorption in the polar cap (PCA), usually lasts from hours to several days and is particularly critical during HF radio wave propagation along polar radio paths [Perrone et al., 2004], resulting in a large decrease in the amplitude of the propagating wave (up to 100 dB).

T he existence of abrupt c ha nges in bac k ground param e ters of the i onosphere o r solar wind, as well as stee p spatial grad i ents of ionos p heric param e ters, brings abo u t the forma t ion of vari o us irregulari t ies. In the pres e nce of eigenoscillations a n d variations in the magnetosphere–ionos p here–atmos p here system under the acti o n of driving forces, thes e effects can oscillate in space and time a c cording to c omplex laws determined by c haracteristics of the syste m 's eigenoscillations and the d ynamics of e x ternal actio n . In practice, this leads to the o ccurrence of spatio-temp o ral variations of all these para m eters with d ifferent spa c e-time scale s : planetary wav e s [Liu et al., 2010], inte rn al gravity w aves [Hun-suc k er, 1982], et c ., and, thus, to the temporal modulation of the above e ffects.

F igure 4 sho w s the power o f a scattere d EKB radar signal as a func t ion of the r a dar range o f scattering and time in the p resence of t r aveling ionospheric disturb a nces. Figure 4, a , b illust r ates cases o f large-scale irre g ularities of t he F layer w ithout formation of an addi t ional signal p ropagation m ode (path), w hich generally reduce to a c h ange in the s hape of the e l ectron density profile with o ut changin g its monoto n y. Similar irre g ularities ar e usually co n sidered as internal or acoustic-gravity w aves [Oinats et al., 2016]. The principle of formation o f such effect s is similar t o that shown in F i gure 2 d . Fig u re 4, c , d de m onstrates ca s es of large-scal e irregularities with the f o rmation of a n additional propagation mode , which usual l y reduce to t h e formation of wavelike vertic a l disturbance s disrupting th e mono-

Figure 3 . Power of a scattered EKB r adar signal vs t he radar rang e of scattering a nd time durin g X-ray flares M 5.2 on May 08, 2014 ( a ) , M6.5 on April 11, 2013 ( c ) , M9.4 on Oc t ober 25, 2013 ( e ), and respe c tive variation s in X-rays as i nferred from GOES data ( b , d , f ); approximate trajectories and ampli t udes of signal s as they propa g ate in the pres e nce ( g ) and i n the absence ( h ) of the a bsorbing layer at a height of 100 km

tonic elect r on density profile and app e aring as fila m ents in the ran g e-time diagram, the dire c tion of thes e filaments indi c ating the direction of mot i on of these i r regularities [Stocker et al., 2000]. Si m ilar irregul a rities arose, for example, in the passage o f waves fro m the Chelyabin s k bolide on February 15, 2 013 [Berng a rdt et al., 2015c; Kutelev, Berngardt, 2013 ] or in the p a ssage of shock w aves from earthquakes [ O gawa et al., 2 012; Berngardt et al., 2017].

Wave p hase changing due to pro p agation in t h e refracting io n osphere, the signal acqui r es a time-v a rying phase shif t , described in the first ap p roximation b y the Doppler fr e quency shift. The freque n cy shift in t h e HF range can b e up to tens of hertz dep e nding on ex t ernal conditions.

As the spatial scale of the irreg u larities decr e ases, the oscilla t ions become faster and m ore pronoun c ed in the Doppl e r frequency shift. Figure 5 shows the p ower and Doppl e r frequency shift of the s c attered EKB radar signal in v elocity units (hereinafter c alled Doppl e r velocity) as a function of the radar ran g e of scatterin g and time durin g such medium-scale irreg u larities.

A close correspondence can be seen between the Doppler velocity and the power variations associated probably not only with the motion of the reflection point along the range (Figure 2, d), but also with the focusing effects [Stocker et al., 2000; Kutelev, Berngardt, 2013], i.e., they are a combination of the effects discussed in Figure 4. It is clear to what a further reduction in scales of irregularities will lead – to signal fadings with even smaller periods, i.e., to the scintillation effect.

Ionospheric ir r egularities g e nerating sig n al scintillation have been st u died in man y works, in pa r ticular their dep e ndence on v a rious manif e stations of s o lar and geoma g netic activit y . A positive correlation b e tween their app e arance and s o lar activity i n dex is typica l for equatorial a nd high latitudes [Aaron s et al., 1980; Rino, Matthews, 1980]. Th e scintillation can be cause d by irregularit i es of different scales – fr om meter t o kilometer, incl u ding mediu m -scale irre gu larities, whose dimension s are compa r able and les s than the r a dius of the Fres n el zone [Basu, Basu, 198 1 ; Basu et al., 1985, 1988; Mul l en et al., 19 8 5; Aarons, 1982; Weber e t al., 1985; Gro v es et al., 19 9 7; Wernik e t al., 2003; G herm et al., 201 1 ]. Notice th a t the scintil l ation is mor e intense at equ a torial latitud e s, although t h ey can also occur at high latit u des, and is l a rgely associa t ed with the development

10           15

Time (LIT,hours)

07/11/2015

E

ф

CD

C ro

(U

OJ o:

12/11/2016

21/11/2015

15/03/2016

2500 -

1500 - .

20 1

10 -

0 Й

Figure 4 . Power of a scattered EKB r adar signal vs the radar rang e of scattering a nd time in th e presence of traveling ionospheric dist u rbance – large-scale irregula r ities of the F l a yer – without ( a , b ) and wit h formation ( c , d ) of an additi o nal propagation mode

of various instabilities [Tsunoda, 1988]. In the eq u atorial ionosph e re, the main mechanism o f scintillation is the plume (pl a sma irregularity). Plumes h ave charact e ristic dimension s of the order of several h u ndreds of ki l ometers in ho r izontal and vertical directions. At hig h latitudes, sci n tillation is associated w i th the grow t h of small-scal e irregularities [Rino, Matthews, 1 980; Aarons, 1 9 82]. Scintillation at middl e latitudes is u sually caused by a combination of high-l a titude mecha n isms and the e ff ect of irregularities in th e subauroral ionosphere, in c luding the emergence o f sporadic l ayers [Goodman, 1967].

Small- s cale irregularities (wit h a charact e ristic scale of t h e order of half the signa l wavelength) can also lead t o substantial backscatter i ng in the H F and VHF ranges, known as radio auro r a. There ar e numerous ty p es of such irregularitie s typical for both polar [ B agaryatsky, 1961; S verdlov, 1 982; Haldoupis , 1989] and equatorial io n ospheric re g ions [Patra et a l ., 2014; Chau, Kudeki, 2 0 13]. A feat u re of most of t h ese irregularities is thei r extension a long geomagne t ic field lines.

The morphology of the high-latitude radio aurora has been extensively studied. Its main characteristic is the relationship with the polar oval position (which, for example, is responsible for the well-known diurnal dependence of the radio aurora with the nighttime enhancement) and electric field intensity, which is one of the main effects accompanying space weather disturbances. Radio aurora has been the objective of many studies [Bagaryatsky, 1961; Sverdlov, 1982; Uryadov et al., 2013], including recent research based on EKB radar data [Berngardt et al., 2015a]. Radio aurora observations by the SuperDARN radar network are reviewed in [Chisham et al., 2007].

It should be noted that the probability of the occurrence of small-s c ale irregula r ities is ofte n associated with regions of intensive cu r rents. As al r eady mention e d, a change i n solar wind characteristi c s leads to a cha n ge in this c u rrent system,, in particula r to the expan s ion of the polar oval regi o n in both eq u atorial and polar directions.

T he ionospheric currents a ccompanying the polar oval move to more equatori a l regions a n d manifest the m selves in dat a from magn e tic stations d ue to generatio n of magneti c fields [Rost o ker, 1972], t h us allowing us t o estimate the degree of g eomagnetic disturbance in the ionospher e . Moreover, indices that feature the deg r ee of growth of the polar oval can be c onstructed

Figure 5 . Power of a scattered EKB r a dar signal ( a , c , e ) and Dopp l er frequency s h ift in velocity u nits ( b , d , f ) v s of the radar range of sca t tering and time in the presen c e of medium-s c ale irregularit i es

Figure 6 . Power of a scattered EKB radar signal ( a ) , the Doppler v elocity ( b ), an d the signal sp e ctral width in v elocity units ( c ) vs the ra d ar range of scattering and ti m e; approxima t e paths and a m plitudes of sig n als propagati n g without ( d ) a nd with ( f ) a layer with i r regularities at a height of 100 km

through a direct analysis of interplanetary ma g netic field distu r bances (Troshichev et al., 2006).

Figure 6 shows the power of a scattered EKB radar signal ( a ), the Doppler velocity ( b ), and the sp e ctral width in u n its of the Doppler velocity ( c ) as a fu n ction of the radar range of scattering and t ime over the period from S e ptember 04 to 06, 2016 al o ng beam 7.

The Fi g ure depicts a large spatial r egion occupied by the radio a u rora, and high equivalent Doppler vel o cities of irregularities exceeding 200 m/s. A signal sca t tered by such ir r egularities is very comple x and has a si g nificant temporal variability, which is confirmed b y the large spect r al width.

Notice that both electron densit y and electric field variations c an cause changes in the D oppler freq u ency shift of a r eceived signal. In this cas e , the Doppl e r frequency shift of a signal scattered b y magneticall y oriented irre g ularities is modulated [B l and et al., 2 014]. These effe c ts, observed with the EK B radar, have been analyzed i n [Mager et al., 2015; Chel p anov et al., 2 016].

Examples of such oscillations are shown in Fig u re 7. The Figur e shows the power of a sc a ttered signal ( a , b , d ) and the Doppler velocity ( b , d , f ) . This Figure indicates that t h ere are radio-aurora regi o ns where the Doppler freque n cy shift reverses sign (ma r ked with an o val). Such quas i -periodic sign reversals can be either long-period (Fi g ure 7, a , b ) or short-period (Figure 7, c – f ) .

Giv e n that plas m a at large h e ights can be considered magnetized, the Doppler vel o city of irre g ularities is dete r mined by t h e E × B drif t , and Dopp l er velocity osci l lations can b e explained by the elect r ic field rotati o n both due t o a change in the structure of current s in the E la y er [Chisham et al., 2007] and due to pro p agation of w aves of va r ious types t hrough the magnetosphere [ C helpanov et al., 2016].

CONCLUSION

T his paper is an attempt t o review effe c ts of space wea t her disturbances on opera ti on of radio d evices. The emp h asis is on s p ace weather i mpact on pr o pagation of HF r adio waves. T he ISTP SB RAS EKB r a dar data are use d to demonstr a te some man i festations of this impact. Exa m ples are given of change s in the grou p delay of a sign a l scattered b y Earth's sur fa ce, which ar e associated wit h the dynamics of the bac k ground elect r on density. Abs o rption of su c h a scattered signal durin g solar flares is al s o exemplifie d . The effect s of large-scale irregularities on the group delay and s t ructure of a signal scattere d from Earth' s surface, as w ell as medi u m-scale irreg u larities on th e group delay,, and the Doppler shift of the s cattered signal frequency a re demonstrated. Examples are g iven of the a p pearance of a signal, scattered by magneto-oriented irre g ularities, and variations in the Doppler freq u ency shift, w hich are m o st often ass o ciated with

Time (UT,hours)

Figure 7 . Power of a scattered signal ( a , c , e ) and Do p pler velocity ( b , d , f ) during t he observation of ULF oscilla t ions of radioaurora char a cteristics on December 04, 2 0 14, December 25, 2014, and December 24, 2014. Ovals i n dicate regions of oscillation observation s

magnetosp h eric waves. Thus, it ha s been show n that pulsed dec a meter radars, including S u perDARN r a dars, are multifunction sensitive devices w hich are af f ected by various space weather factors an d allow moni t oring of space w e ather effects in the ionosp h ere.

The author is grateful to NOAA for providing GOES data [], to A.V. Tashchilin, S.B. Lunyushkin, and M.V. Uspensky for useful and fruitful discussions. Experimental data were obtained with the EKB radar of ISTP SB RAS under the project II.12.2.3. The work was funded by RAS Presidium Program 7.