Background electromagnetic noise characterization: the role of external and internal Earth sources ANNALS OF GEOPHYSICS, 58, 3, 2015, G0330; doi:10.4401/ag-6745 G0330 Background electromagnetic noise characterization: the role of external and internal Earth sources Antonio Meloni, Cesidio Bianchi*, Giuliana Mele, Paolo Palangio Istituto Nazionale di Geofisica e Vulcanologia, Rome, Italy ABSTRACT The Earth is surrounded by the ionosphere and magnetosphere that can roughly be seen schematically as two concentric shells. These two com- posed and inhomogeneous structured shells around the Earth selectively affect electromagnetic (EM) waves propagation. Both ionosphere and magnetosphere interact also with particles and waves coming from ex- ternal sources, generating electromagnetic phenomena that in turn might become sources of EM waves. Conversely, EM waves generated inside the ionosphere remain confined at various altitudes in this region, up to a so- called critical frequency limit, depending on frequency, EM waves can es- cape out of the ionosphere and magnetosphere or get through. The EM waves generated inside the magnetospheric cavity mainly originate as a result of the electrical activity in the atmosphere. It is well known that also man-made sources, now widely spread on Earth, are a fundamental source of EM waves; however, excluding certain frequencies employed in power distribution and communication, man-made noise can be domi- nant only at local scale, near their source. According to recent studies, EM waves are also generated in the Earth’s lithosphere; these waves were sometimes associated with earthquake activity showing, on the Earth’s surface, intensities that are generally orders of magnitude below the back- ground EM noise. In this review paper, we illustrate EM waves of natu- ral origin and discuss their characterization in order to try discriminate those of lithospheric origin detectable at or near the Earth’s surface. 1. Introduction The electromagnetic (EM) spectrum is subdivided into regions, according to wavelength (Table 1). The Earth’s atmosphere prevents most of the electromag- netic radiation coming from outer space from reaching the Earth’s surface; in fact, only portions of radio and visible light reach the ground. The natural electromagnetic background noise in the Earth’s environment originates from different phe- nomena that involve both Earth’s external and internal sources and their possible interaction with the ionos- pheric and magnetospheric cavities. External sources bring EM radiations inside the magnetospheric cavity but they can also be generated inside the cavity, for ex- ample by energetic particles acting as forcing mecha- nisms that perturb and modify the ionospheric and magnetospheric configuration. The EM Power Spectral Density (PSD) at the Earth’s surface was deeply stud- ied and reported since the first extensive measurements were available, as discussed for example by Lanzerotti [1978] and Lanzerotti et al. [1990]. The most intensive near Earth’s sources are associated with the atmos- pheric electrical activity, while the external sources mainly originate on the Sun. Relevant information about the external EM noise can be found in the CCIR/ITU (Comité Consultatif International pour la Radio; now also International Telecommunication Union) Radio- communication Sector (ITU-R) reports, published in 1964, 1988, and 1990 [CCIR/ITU 1960, 1988, 1990], as well as in Davies [1990], Helliwell [2006], and Bianchi and Meloni [2007]. Sudden fluxes of particles that enter the magne- tosphere generate magneto-hydro-dynamic (MHD) waves and other relevant phenomena, especially in the polar regions (e.g., auroral displays, magnetic sub- storms, Polar Caps Absorption) and also wave-parti- cle interactions (whistlers, chorus, hiss). According to Lan-zerotti et al. [1990], PSD varies of about 16 or- ders of magnitude in the frequency band 10-5 - 105 Hz, with values ranging from ≈10-1 to 10-17 W m-2 Hz-1. For frequencies up to the GHz, the PSD falls to ≈10-21 W m-2 Hz-1. The spectrum of the EM pertur- bations inside the ionosphere and magnetosphere cavities is mostly studied for the ULF (Ultra Low Fre- quencies) up to the HF (High Frequency) bands. Highest frequencies will not be considered in this paper since, starting from HF onwards, i.e. corre- sponding to waves able to escape the ionosphere, the background noise decreases accordingly. In Table 1 acronyms, frequency, wavelength and PSD approxi- Article history Received January 29, 2015; accepted June 9, 2015. Subject classification: Radio noise characterization, Ionosphere-magnetosphere, Background EM noise, Crustal/lithospheric EM sources. mate values of the natural radio noise sources in the magnetosphere and ionosphere, are reported. As briefly reported above, the background EM noise be- comes less important at higher frequency while the cosmic noise is prevalent starting from VHF-UHF fre- quency band. At lower frequency a primary source of noise is the interaction between particles and waves coming from outer space and the magnetosphere, while a secondary source originates inside the ionos- pheric cavity: in this case atmospheric lightning dis- charges produce several remarkable phenomena such as sferics, tweeks and so on [e.g., Wait 1982, Helliwell 2006, Bianchi and Meloni 2007, and references therein]. Over the range ≈10-4 to ≈104 Hz the ampli- tude spectra (fT/√Hz) fall between the two straight lines f -1.5 and f -1 with an average half line slope that rolls off as f -1.25 (Figure 1). As a natural trend, the spectral density decreases considerably from lower to higher frequency, since contributions to background noise from sources internal to the cavities are less im- portant and cosmic noise becomes prevalent up to the millimetric wavelength. The f -1 decay is a natural trend, being a constant when considering a confined environment [Füllekrug and Fraser-Smith 2011]. The purpose of this work is to describe the EM background noise in the Earth’s environment and to compare the natural external signals with the internal Earth’s crust, or more generally lithospheric, sources. These are supposed to be weaker and local and to have peculiar characteristics that can be used to discriminate the internal from the external contributions (see a re- view by Hayakawa and Hobara [2010]). EM emissions radiated from the lithosphere were at times related to earthquakes occurrence. In particular, many authors have hypothesized that EM waves can be generated at the hypocenter during the earthquake or its preparation phase [e.g., Hayakawa et al. 1996, Johnston 2002, Teys- seire and Ernst 2002, Uyeda et al. 2009]; reported EM phenomena take place in a wide frequency range [see for example: Hayakawa and Fujinawa 1994, Hayakawa and Molchanov 2002]. Some reported ionospheric phe- nomena, were related to earthquake occurrence [Pu- linets et al. 2007], but were criticized. In fact Thomas et al. [2012] had reproduced the same time series re- sults, in which the authors identified an anomalous sig- nal, and found it was normal solar and geomagnetic activity that were, in their opinion, unrelated to the earthquake. In the following, the most relevant sources of nat- ural EM waves, in the various frequency ranges, will be treated separately. 2. Background noise in the ULF range Geomagnetic pulsations, also called micropulsa- tions (or simply pulsations) are the ground signature of ultra low frequency (ULF) hydromagnetic waves prop- MELONI ET AL. 2 Figure 1. EM natural noise PSD. In the frequency interval from 10-4 to 104 Hz, the noise falls between the two straight lines trending as f -1 and f -1.5 (redrawn after Lanzerotti et al. [1990]). Sources of natural EM radio noise Frequency Wavelength (m) Spectral power (W m-2 Hz-1) Physical environment resonance in the magnetospheric cavity, interaction with particles of solar origin and radiative pressure with the magnetosphere ULF 0.1 ÷ 3,000 mHz 3·1011 ÷ 108 10-1 ÷ 10-10 magnetosphere resonance in the ionospheric cavity ELF 3 ÷ 3,000 Hz 108 ÷ 105 10-10 ÷ 10-12 ionosphere propagation in the ionospheric cavity of the radiating atmospheric discharge VLF 3 ÷ 30 kHz 105 ÷ 104 10-12 ÷ 10-14 ionosphere atmospheric noise LF-MF-HF 30 kHz ÷ 30 MHz 104 ÷ 102 10-14 ÷ 10-17 ionosphere Table 1. Main physical characteristics of the natural EM radio noise sources in the magnetosphere and ionosphere. 3 agating in the Earth’s magnetosphere [Anderson 1993, 1994]. Their period varies from less than 1 s to 10 min, while their amplitude varies from tens to hundreds nT, increasing with increasing period and magnetic latitude. Traditionally, micropulsations are classified as continuous (Pc) and irregular (Pi) as reported in Table 2. Most of the ULF waves detected at the Earth’s surface originate outside the magnetosphere [Yumoto 1988]. The solar wind, the magnetospheric foreshock, the magnetos- pheric bow shock, and the magnetopause, are sources of ULF waves. Many of these waves pass through the magnetopause and propagate through the magnetos- phere interacting with waveguides, magnetic field lines and the ionosphere generating the pulsations detected at the Earth’s surface. Externally generated ULF waves detected at the ground level are not identical to waves that enter the magnetosphere through the solar wind, since their en- ergy is transformed and amplified inside the magnetos- phere [see for example McPherron 1991]. Schematically, regular pulsations from Pc3 to Pc5 are due to the so called Kelvin-Helmholtz instability generated at the magnetopause by the differential flow of the solar wind on either side of the magnetosphere. Another mechanism of pulsations generation, mainly in the bands Pc3-Pc4, is given by the penetration of waves generated upstream from the bow shock region, called ‘upstream waves’ into the magnetosphere. Pc1 and Pc2 have periods comparable with cyclotronic res- onances in the magnetosphere and are associated with the magnetospheric plasma, in particular drift of pro- tons and heavier ions. Pulsations amplitude is normally not larger than a few nT in the Pc1 and Pc2 range and a few tens of nT at longer periods; in exceptional cases pulsations can reach very large amplitudes (100 nT and more). Short period pulsations Pc1, Pc2 and Pi1 are travel- ling waves, while long period pulsations are mainly standing waves inside the magnetospheric cavity. The same mechanism can be observed in the ionospheric cavity for low frequency longitudinal and transversal standing waves. The short period pulsations are re- lated to the Kp magnetic activity index; they show both frequency and amplitude modulation and are more sporadic than the longer period pulsations [Kangas et al. 1998]. The understanding of the pulsation generation mechanisms requires a detailed knowledge of the mag- netosphere structure and dynamics that is beyond the purpose of this work. Long period pulsations (Pc5 and Pc4) are the magnetic signatures of plasma waves with wavelengths comparable to the magnetospheric size. These large-amplitude (up to hundreds of nT), low-fre- quency continuous pulsations are much stronger in the auroral zone [Samson and Rostoker 1972]. In the vicin- ity of the resonance shell, a portion of the energy of these waves is transferred, through a coupling phe- nomenon, to Alfven oscillations and then to Pc5 and Pc4 pulsation standing waves [Kivelson et al. 1984, Surkov and Hayakawa 2008]. EARTH’S BACKGROUND EM NOISE CHARACTERIZATION Continuous pulsations Pc1 Pc2 Pc3 Pc4 Pc5 Period (s) 0.2 - 5 5 - 10 10 - 45 45 - 180 150 - 600 Frequency (mHz) 5,000 - 200 200 - 100 100 - 22 22 - 7 7 - 2 Maximum intensity (nT) 1 3 10 < 300 300 Source EM ion-cyclotron instability in the equatorial magnetosphere EM ion-cyclotron instability in the equatorial magnetosphere wave-particle interaction in the bow shock region drift of protons from the night side magnetopause instability Impulsive pulsations Incoherent noise Pi1 Pi2 Period (s) 1 - 40 40 - 150 1 - 1,000 Frequency (mHz) 1,000 - 25 25 - 2 1,000 - 1 Maximum intensity (nT) 10 100 - Source modulation of charged particles sub-storms ionospheric electric currents Table 2. IAGA classification and standard nomenclature of the most relevant continuous and impulsive pulsations in the ULF band. 3. Background noise in ELF-VLF range Broad-band ELF-VLF spectrum measurements made in typical low noise environments, show a uni- form structure all over the Earth [Lanzerotti et al. 1990]; a more intense radio noise is present at high lat- itude in the frequency band between 0.5 and 1,500 Hz, where so called polar hiss/chorus emissions can be relevant [Golden et al. 2009]. Typical background noise at mid latitude in the VLF-ELF band is reported in Figure 2 [Fraser-Smith 2007]. During daytime, the background noise is particularly attenuated in the fre- quency window 1-5 kHz [Garcia and Jones 2002]. This frequency interval is called “dead band” and is nearly coincident with the transversal resonances of the ionospheric cavity. The EM noise background fills the space around the Earth interacting in a complex way with the mag- neto-ionosphere boundary shells at various altitudes and latitudes. The EM background noise due the su- perimposition of numerous sources distributed inside the two cavities [Meloni et al. 1992] is the result of all the existing phenomena, such as sferics, tweeks, whistlers and others [e.g., Wait 1982, Helliwell 2006, Bianchi and Mel- oni 2007], and of the complex interaction of the EM fields with the approximate reference surfaces delimit- ing the cavities. In general, as mentioned before, the largest the cavity volume, the lowest the resonant fre- quencies and the highest the intensity of the fields con- fined in the cavity. The background noise in the ionospheric cavity is sustained by a range of 60-100 natural lightning elec- trical discharges per second, especially generated in the tropics, with a wide spectrum ranging from few Hz to hundreds of MHz. This spectrum is strongly influenced by the ionospheric cavity shape and, as a result, only signals in the spectral windows ranging from about 8 to 45 Hz, survive. These are the Schumann longitudinal resonances [Bliokh et al. 1980, Sentman 1987, Sentman and Fraser 1991] and related harmonics. All possible EM frequencies are given by the relation: where a is the Earth’s radius, c is the velocity of light in vacuum and n is any integer number. Moreover the conductive layer of the ionosphere, at about 100 km of height, gives the highest amplitude resonant frequency and related harmonics, not on a global scale, in the 900- 5,000 Hz band (transversal resonances). In the frequency band from 5 to 30 kHz, the dom- inant signals are the sferics (also called statics), that are frequently audible in the radio equipments, as well as the tweeks (dispersed sferics) are. Taking into consid- eration a single lightning discharge, the observed spec- trum depends on the source-observer distance and is called source spectrum. The convolution of the source function with the cavity transfer function becomes more and more important at increasing distance. As a consequence the nearest observer detects a maximum intensity around 300 kHz, the farthest observer experi- ences a maximum intensity around 7 kHz. Table 3 shows the main EM sources in the magne- tosphere and ionosphere. In the frequency range from ELF/VLF to LF, the PSD varies roughly as the inverse squared frequency [Fraser-Smith 1995]. 4. Possible origin and mechanisms for internal Earth generated EM waves The description of the ionospheric and magnetos- pheric signals, in the frequency range from ULF to VLF, above presented, attains to a consolidated investigation field, following several decades of radio wave investi- gations. In the lasts decades several authors have re- ported on the existence of internally generated EM waves that are reported as weaker and local. Detection of EM waves of internal Earth origin is made with direct ob- servation of electromagnetic emissions from the litho- sphere. We should mention however that some authors also operate detecting signals originated in the litho- sphere, indirectly. This is done investigating on the prop- agation anomaly of a pre-existing transmitter signal; the anomaly could be due to a mechanism of lithosphere– atmosphere–ionosphere coupling [see: Uyeda et al. 2009, Haykawa and Hobara 2010, and references therein]. The characterization of EM waves generated by internal Earth sources, is done in terms of their range of fre- quency and amplitude, wave attenuation, crustal or sub- soil mode of propagation, that allow to detect this kind of EM emission. Moreover the type of emitting process encompasses a range of uncertainty that is still domi- r2 a c n n 1n= +f ^^ h MELONI ET AL. 4 Figure 2. PSD of the ELF-VLF background mid-latitude noise. The blue and red lines indicate minima and maxima, respectively. (1) 5 nant in the achieved findings [Dea et al. 1993, Johnston 2002, Teysseire and Ernst 2002, Palangio et al. 2007]. All authors generally reported that rocks modifi- cation or fracture in the Earth’s crust, could produce EM waves. The fracture in brittle materials, such as solid rocks, can be envisaged as the result of a series of different phases taking place prior to earthquakes, sum- marized in the so called dilatancy theory [Stuart 1974]. This theory is based on the idea that a rock, when stressed, begins to increase its volume or to dilate; this is caused by the opening of micro-cracks and fractures in the rock. This theory was revitalized several times for earthquake prediction, but also severely criticized [Geller et al. 1997]. The sequence of events is following: 1) ac- cumulation of stress; 2) micro-fracturing; 3) dilatancy; 4) macroscopic rupture. This last rupture occurs with release of stress and emission of seismic waves. In the Earth’s crust the first phase can last tens of years, cen- turies, or even more, and the stress can increase continu- ously or with intermittence. The macroscopic rupture of rocks consists in a fracturing at the microscopic scale, radiating elastic energy in a manner analogous to earth- quakes [Scholz 1968, Kato et al. 1994]. When micro-fracturing occurs in certain rocks, it can cause piezoelectric effects that create packets of EM waves of different frequency capable of travelling sig- nificant distances through the Earth’s crust before reach- ing the surface. This is one of the mechanisms that could account for generation of EM waves before earthquakes [Johnston 2002, and references therein]. Micro-fractur- ing occurs when a rock is roughly half a way towards its breaking point. Several laboratory experiments were undertaken to detect these waves in a controlled envi- ronment [e.g., Freund 2008, 2009]. The change in the rock size may lead to the uplift of ground surface or to a variation in the groundwater pressure and level. Bolt [1988] has shown that, as crustal rocks break up under the tectonic stress preceding an earthquake, the variation in the pore pressure inside the rock causes a temporary decrease of the seismic P waves velocity. Unfortunately, the time interval between the P-wave velocity change and the earthquake, might last months or years; the greater the magnitude of the earth- quake, the greater the lapse time. Much of earthquake predictions was based on the changes that the rocks un- dergo when they are subject to all these phases. Rocks change physically transmitting seismic waves at chang- ing speeds, magnetic properties can be altered and elec- trical resistance will also vary. Famous example is the Guam earthquake with magnitude 8.0 occurred in 1993 [Hayakawa et al.1996]. Thomas et al. [2009] report how- ever, that the anomalous magnetic noise identified by Fraser-Smith et al. is not related to the Loma Prieta earthquake but, according to their work, was an arti- fact of sensor-system malfunction. Also in the case of Guam 1993 earthquake Thomas et al. [2009] reported that the changes in polarization was part of normal global magnetic activity and unrelated to the earth- quake. So the above mentioned results and their inter- pretation are still awaiting for confirmation. For what concerns EM waves generation process, piezo-magnetism, but also other physical mechanisms, have been proposed to explain the appearance of mag- netic waves prior to earthquakes. Other possible EM waves generation mechanisms include electro-kinetic [Fenoglio et al. 1995, Simpson and Taflove 2005] and EARTH’S BACKGROUND EM NOISE CHARACTERIZATION (a) Magnetospheric sources Hz Free oscillations in the magnetospheric cavity 3·10-3 - 102 Resonance in the magnetospheric cavity 0.1 - 10 Magneto-hydrodynamic waves in the magnetospheric cavity 101 - 2·104 Resonance wave-particles (transversal and longitudinal) 3·104 - 2·104 Emission due to plasma instabilities 3·10-3 - 105 (b) Ionospheric sources Hz Emission due to ablation of meteorites 3·102 - 3·104 Sferics 1·103 - 1·105 Sprite, Blue Jet, Red Flash, Elves 2·103 - 6·104 Longitudinal resonance in the ionospheric cavity 7.8 - 48 Transversal resonance in the ionospheric cavity 8·102 - 5·103 Atmospheric gamma rays 6·103 - 6·104 Table 3. (a) Magnetospheric and (b) ionospheric EM sources. magneto-hydrodynamic effects [Surkov and Pilipenko 2014], resulting from fluid circulation through the crust, stress-induced increase in local crustal electrical conductivity [Merzer and Klemperer 1997], micro-frac- turing, thermal magnetization/demagnetization effects and so on [Johnston 2002, Uyeda et al. 2009]. Freund [2008, 2009] has suggested a semiconduc- tor model of rocks to describe unipolar magnetic pulses, as several times was observed prior to earthquakes. The suggested mechanism is based on coupling a semicon- ductor drift-diffusion model to a magnetic field to de- scribe the electromagnetic effects associated with electrical currents flowing within rocks [see also Scoville et al. 2014]. However for fluid saturated samples, ac- cording to Dahlgren et al. [2014], observation of sig- nificant electrical charge buildup is not expected during the observed slow stress accumulation prior to earth- quakes, or during any slow precursory stress release that may occur in the region of earthquake nucleation. 5. Characterization of internal Earth sources For what concerns the Earth’s crust, or more gen- erally the lithosphere, among the less known quanti- ties, are the EM properties. Once the variability range of electrical conductivity (v) and electrical permittivity (f) is established, other features of the EM internal sig- nals that enable to discriminate them from external sources, can be inferred within an acceptable level of accuracy. Except for frequency and intensity, other sig- nal electromagnetic characteristics like velocity, propa- gation mode, polarization, impulsiveness, coherence and spectral content, strongly depend on the media electrical properties. Most of the published works deal with the ULF and ELF part of the electromagnetic spectrum. In fact, electromagnetic signals produced in the seismogenic layer (for example between 10 and 20 km of depth, Chiarabba et al. [2005]), that can reach the Earth’s surface with little attenuation [Molchanov and Hayakawa 1995], fall in the ULF-ELF bands. In this framework the most probable internal sig- nals should have the following characteristics: - frequency range 0.01-1,000 Hz - magnetic intensity range 1-100 fT - depending on v, the diffuse magnetic field veloc- ity range 1,000-10,000 km/s - polarization, depending on sources pattern and nature - impulsiveness, following estimate of the proba- bility density function - coherence, depending on the extent of the source. Sources can be magnetic, electric and hybrid, with the spectral content depending on the medium charac- teristics and the statistical behavior depending on the emission mechanism. In Table 4 some of the expected features of the internal sources are listed. Depending on distance, the EM fields emitted by a source, or scattered by a reflector, are always charac- terized by an inductive, so-called near field, and a prop- agating so-called far field. Here, ‘near’ and ‘far’, are referred with respect to: a distance d from the source, the wavelength m, the extent and the coherence of the source D physical dimension. Through the above quan- tities, four spatial regions are defined: (2) reactive field region, (3) I Fresnel’s region, (4) II Fresnel’s region, and (5) Fraunhofer’s region: The EM near field contains useful information on the nature of the source, but the way to resolve this matter can be rather complex. In the near field the sig- nal’s power falls as r-5 with the distance r from the source and there is no propagating field except for a small frac- tion that will sustain the far field. In the above defined condition (a) the E/H ratio gives indication about the magnetic/electric nature of the source [Palangio et al. 2008]. In the regions (b), (c) and (d) the power falls as 2 2 4 4 d d D D d D d D 2 2 1 1 1 1 1 2 r m r m m m m m 2 2 2 2 MELONI ET AL. 6 Table 4. Some expected features of the EM wave internal sources. (a) Mechanism of emission Point source elemental dipole Extended elemental dipole Point source hybrid multipole Extended hybrid multipole (c) Statistical behavior Gaussian Isotropic Stationary Impulsive sources Homogeneous sources (b) Spectral content and trend Thermic (white) Colored Complex (2) (3) (4) (5) 7 r-4, r-3 and r-2 respectively, while the wave impedance is defined only in (d). These considerations are essential both for signal measurement purpose and analysis, in order to characterize the observed signal properties. In the terrestrial environment quasi non radiating (QNR) magnetic sources, whose fields remain confined in the same volume of the source, are quite common. Typical examples are the toroidal sources e.g., dynamo in the Earth’s core and in the ionosphere. These sources are generally complex, with toroidal components that are not observable far from the source, while the sys- tem energetic balance cannot be resolved by measure- ments made externally to the system. Only the sources poloidal components can be measured. Very extended sources, larger than the dominant observed signal wavelength, emit electric and magnetic fields whose space-time correlation length reflects coherence prop- erties that depend on the nature of the emitting source (see general Geomagnetism textbooks, e.g.: Backus et al. [1996], Campbell [2001]). 6. Conclusions In this paper, we started describing the terrestrial natural EM noise in the frequency band from 10-5 to 105 Hz. Sections 2 and 3 were devoted specifically to the main natural sources of EM waves in the ionospheric and magnetospheric cavities, in the frequency ranges from ULF to VLF. The following sections dealt with EM sources that could be of Earth’s internal origin. We have shown some typical characteristics of nat- ural sources in the ionosphere and magnetosphere cavi- ties and some of their relevant properties. We have shown that the EM natural background radio noise fol- lows some general physical laws. For example the back- ground noise generated externally to the Earth’s surface, shows a PSD that follows a typical decreasing trend, as shown in Figures 1 and 2, exhibiting a repeating pattern that displays at every scale; this occurrence is typical of fractals, as it happens in many other natural phenomena. We have not dealt here in detail with the EM man- made radio noise sources and their characteristics; we know however that this noise can be compared with natural noise, and that only in particular in the ELF band, the two contributions can be easily mixed up. Moreover in the past decades, the widespread EM field measurements eliminated some uncertainty concern- ing natural noise parameterization in terms of time- space-frequency. Electromagnetic signals existing in the ionospheric and magnetospheric cavities are not observed only as primary fields but also as secondary. In fact they can penetrate the Earth and propagate at various depths with different attenuations depending on the frequency and on the electric characteristics of the media. This issue is dealt with in Magnetotelluric methods that try to characterize the internal Earth electrical conductiv- ity by means of a combination of electric and magnetic measurements on the Earth’s surface [Garcia and Jones 2002] and in some cases by means of magnetic field time variations transfer functions only [Egbert and Booker 1986]. These induced signals can also be reflected at the interfaces and interfere with the primary inducing sig- nals. This leads to a representation of the distribution of electromagnetic fields in the time-space and fre- quency domain. In case of possible lithospheric sources, the emit- ted signal’s energy is not yet characterized and is rea- sonably dependent on the distance. While for ionospheric or magnetospheric sources, the emitted energy is better characterized. It has been reported also that the signal polarization, the ratio of magnetic vertical to horizontal component, is relatively small for the plasma waves coming from the ionosphere/mag- netosphere, while this ratio is considerably enhanced, for lithospheric origin emissions. Moreover the signal impulsiveness can be important to discriminate the in- ternal sources. In fact impulsiveness is a quality of in- ternal sources while if observed in case of external sources, it should subsist at a large scale, if not world- wide. If a local array network, consisting of at least three stations can be used in the field, the so called ‘di- rection finding’ radio technique can be used to infer the signal internal origin. The operating method can be go- niometric or gradiometric. In the Izu and Chiba penin- sulas ( Japan), by measuring the gradient of horizontal and vertical components of the magnetic field at dif- ferent frequencies, Kopytenko et al. [2002] and Is- maguilov et al. [2002] have computed the azimuth of the incoming waves from the normal to the observed gradient, leading to an internal source. In case of in- ternal Earth generated signals no significant correlation between EM signals and geomagnetic activity, ex- pressed also with geomagnetic indices, should be found. Possible relations of EM waves propagating in the ionospheric-magnetospheric cavity and earthquakes, as we have also mentioned before, cannot be excluded ‘a priori’. We note however that the mechanism of litho- sphere–atmosphere–ionosphere coupling, that should be relevant as the final goal of seismo-electromagnetics, still needs to be deeply investigated. As we have reported here the separation between the various external sources of EM signals and the Earth internal sources, is on the other hand the primary problem to be solved and many questions still remain unanswered on the separa- tion of the two contributions. However some elements EARTH’S BACKGROUND EM NOISE CHARACTERIZATION concerning the internal EM source are now considered important, for example the knowledge of electrical con- ductivity, permittivity, temperature, chemical composi- tion of the crust and its layered structure, with possible travel paths, might play an important role in this re- search field. Moreover the EM probable frequency band for these phenomena should fall in the ULF-ELF bands. For this reason a detailed monitoring and knowledge of the background EM noise is still needed. References Anderson, B.J. (1993). Statistical studies of Pc 3-5 pul- sations and their relevance for possible source mech- anisms of ULF waves, Ann. Geophysicae, 11, 128-143. Anderson, B.J. (1994). An overview of spacecraft ob- servations of 10 s to 600 s period magnetic pulsations in the Earth’s magnetosphere, In: M.J. Engebretson, K. Takahashi and M. Scholer Solar (eds.), Wind Sources of Magnetospheric Ultra-Low-Frequency Waves, AGU Geophysical Monograph, 81, 25-43. Backus, G., R. Parker and C. Constable (1996). Foun- dations of geomagnetism. Cambridge University Press, Cambridge, 369 pp. Bianchi, C., and A. Meloni (2007). Terrestrial natural and man-made electromagnetic noise: an outlook, Annals of Geophysics, 50 (3), 435-445. Bliokh, P.V., A.P. Nikolaenko and Y.F. Filippov (1980). Schumann Resonances in the Earth-Ionosphere Cavity, Peter Peregrinus Ltd., London. Bolt, B. (1988). Earthquakes, W.H. Freeman, 282 pp. Campbell, W.H. (2001). Earth magnetism. A guided tour through magnetic fields. Harcourt Academic Press, San Diego, CA, USA, 151 pp. CCIR/ITU (1964). World distribution and characteris- tics of atmospheric radio noise, Rep. 322, Int. Radio Consultative Comm., Int. Telecommun. Union, Geneva, Switzerland. CCIR/ITU (1988). Characteristics and applications of atmospheric radio noise data, Rep. 322-3, Int. Radio Consultative Comm., Int. Telecommun. Union, Geneva, Switzerland. CCIR/ITU (1990). Man-made radio noise, Rep. 258-5, Int. Radio Consultative Comm., Int. Telecommun. Union, Geneva, Switzerland. Chiarabba, C., L. Jovane and R. Di Stefano (2005). A new view of Italian seismicity using 20 years of in- strumental recordings, Tectonophysics, 395 (3-4), 251-268; doi:10.1016/j.tecto.2004.09.013. Dahlgren, R.P., M.J.S. Johnston, V.C. Vanderbilt and R.N. Nakaba (2014). Comparison of the Stress Stim- ulated Current of Dry and Fluid Saturated Gabbro Samples, B. Seismol. Soc. Am., 104, 2662-2672; doi:10. 1785/0120140144. Davies, K. (1990). Ionospheric Radio (Peter Peregrinus Ltd., London, U.K.). IEE Electromagnetic Waves Ser. 31, 580 pp. Dea, J.Y., P.M. Hansen and W.M. Boerner (1993). Long- term ELF background noise measurements, the ex- istence of window regions, and applications to earth- quake precursor emission studies. Phys. Earth Planet. Interior, 77, 109-125. Egbert, G.D., and J.R. Booker (1986). Robust estimation of geomagnetic transfer functions, Geophys. J. R. Astron. Soc., 87, 173-194. Fenoglio, M.A., M.J.S. Johnston and J.D. Byerlee (1995). Magnetic and electric fields associated with changes in high pore pressure in fault zones: application to the Loma Prieta earthquake, J. Geophys. Res., 100 (B7), 12951-12958. Fraser-Smith, A.C. (1995). Low-frequency radio noise, In: H. Volland (ed.), Handbook of Atmospheric Electrodynamics, 1, 297-310, CRC Press. Fraser-Smith, A.C. (2007). The effective antenna noise figure Fa for a vertical loop antenna and its applica- tion to extremely low frequency/very low frequency atmospheric noise, Radio Sci., 42, RS4026; doi:10.10 29/2005RS003387. Freund, F.T. (2008). Earthquake probabilities and pre- earthquake signals, Current Sci., 94, 1-2. Freund, F.T. (2009). Stress-activated positive hole charge carriers in rocks and the generation of pre-earthquake signals, In: M. Hayakawa (ed.), Electromagnetic Phenomena Associated with Earthquakes, Research Signpost, India, 41-96; ISBN 978-81-7895-297-0. Füllekrug, M., and A.C. Fraser-Smith (2011). The Earth’s electromagnetic environment, Geophys. Res. Lett., 38, L21807; doi:10.1029/2011GL049572. Garcia, X., and A.G. Jones (2002). Atmospheric sources for audio-magnetotelluric (AMT) sounding, Geo- physics, 67 (2), 448-458; doi:10.1190/1.1468604. Geller, R.J., D.D. Jackson, Y.Y. Kagan and F. Mulargia (1997) Earthquakes Cannot Be Predicted, Science , 275 (14), 1616-1617. Golden, D.I., M. Spasojevic and U.S. Inan (2009). Diur- nal dependence of ELF/VLF hiss and its relation to chorus at L=2.4, J. Geophys. Res., 114, A05212; doi:10.1029/2008ja013946. Hayakawa, M., and Y. Fujinawa (1994). Electromagnetic Phenomena Related to Earthquake Prediction, Terra Scientific, Tokyo. Hayakawa, M., R. Kawate, O.A. Molchanov and K. Yu- moto (1996). Results of ultra-low frequency magnetic field measurements during the Guam earthquake of 8 August 1993, Geophys. Res. Lett., 23, 241-244. Hayakawa, M., and O.A. Molchanov (2002). Seismo Elec- MELONI ET AL. 8 9 tromagnetics: Lithosphere–atmosphere–ionosphere coupling, Terra Scientific, Tokyo. Hayakawa, M., and Y. Hobara (2010) Current status of seismo-electromagnetics for short-term earthquake prediction, Geomatics, Natural Hazards and Risk, 1/2, 115-155. Helliwell, R.A. (2006). Whistlers and Related Ionos- pheric Phenomena, Dover Publications, Inc., ISBN 0-486-44572-0; originally published by Stanford Uni- versity Press, Stanford, California, 1965. Ismaguilov, V.S., Y.A. Kopytenko, K. Hattori and M. Hayakawa (2002). Variations of phase velocity and gradient values of ULF geomagnetic disturbances connected with the Izu strong earthquakes, Nat. Hazards Earth Syst. Sci., 20, 1-5. Johnston, M.J.S. (2002). Electromagnetic fields gener- ated by Earthquakes, International Handbook of Earthquakes and Engineering Seismology, 81A, El- sevier Science Ltd. Kangas, J., A. Guglielmi and O. Pokhotelov (1998). Morphology and physics of short period magnetic pulsations, Space Sci. Rev., 83, 435-512. Kato, N., K. Yamamoto and T. Hirasawa (1994). Mi- crofracture processes in the breakdown zone during dynamic shear rupture inferred from laboratory ob- servation of near fault high frequency strong mo- tion, Pageoph, 142, 713-734. Kivelson, M.G., J. Etcheto and J.G. Trotignon (1984). Global compressional oscillations of the terrestrial magnetosphere: The evidence and a model, J. Geo- phys. Res., 89 (11), 9851-9856. Kopytenko, Y.A., V.S. Ismaguilov, K. Hattori and M. Hayakawa (2002). Monitoring of the ULF electro- magnetic disturbances at the station network before EQ in seismic zones of Izu and Chiba peninsulas, In: M. Hayakawa and O.A. Molchanov (eds.), Seismo Electromagnetics: Lithosphere–atmosphere–iono- sphere coupling, 11-18, Terra Scientific, Tokyo. Lanzerotti, L.J. (1978). Studies of geomagnetic pulsa- tions, In: L.J. Lanzerotti and C.G. Park (eds.), Upper Atmosphere Research in Antarctica, 130, AGU, Wash- ington D.C. Lanzerotti, L.J., C.G. Maclennan and A.C. Fraser-Smith (1990). Background magnetic spectra: ~10-5 to ~105 Hz, Geophys. Res. Lett., 17, 1593-1596. McPherron, R.L. (1991). Physical Processes Producing Magnetospheric Substorms and Magnetic Storms, In: J. Jacobs (ed.), Geomagnetism, 593-739, Aca- demic Press, London. Meloni, A., P. Palangio and A.C. Fraser-Smith (1992). Some characteristics of the ELF/VLF radio noise measured near L’Aquila, Italy, IEEE T. Ant. Propag., 40 (2), 12-18. Merzer, M., and S.L. Klemperer (1997). Modeling Low- frequency Magnetic-field Precursors to the Loma Prieta Earthquake with a Precursory Increase in Fault-zone Conductivity, Pure Appl. Geophys., 150, 217-248. Mochanov, O.A., and M. Hayakawa (1995). Generation of ULF electromagnetic emissions by microfractur- ing, Geophys. Res. Lett., 22, 3091-3094. Palangio, P., C. Di Lorenzo, F. Masci and M. Di Persio (2007). The study of the electromagnetic anomalies linked with the Earth’s crustal activity in the fre- quency band [0.001 Hz–100 kHz], Nat. Hazards Earth. Syst. Sci., 7, 507-511. Palangio, P., A. Lozito, A. Meloni and C. Bianchi (2008). Electromagnetic Monitoring MEM Project, Quaderni di Geofisica, 53, INGV (in Italian). Pulinets, S.A., N. Kotsarenko, L. Ciraolo and I.A. Pu- linets (2007), Special case of ionospheric day-to-day variability associated with earthquake preparation, Adv. Space Res., 39, 970-977; doi:10.1016/j.asr.2006. 04.032. Samson, J.C., and G. Rostoker (1972). Latitude-depen- dent characteristics of high latitude Pc4 and Pc5 mi- cropulsations, J. Geophys. Res, 77, 6133. Scholz, C.H. (1968). Microfracturing and the inelastic deformation of rock in compression, J. Geophys. Res., 73 (4), 1417-1432; doi:10.1029/JB073i004p01417. Scoville, J., J. Heraud and F. Freund (2014). Pre-earth- quake magnetic pulses, Nat. Hazards Earth Syst. Sci. Discuss., 2, 7367-7381; doi:10.5194/nhessd-2-7367- 2014. Sentman, D.D. (1987). Magnetic polarization of Schu- mann resonances, Radio Sci., 22, 595-606. Sentman, D.D., and B.J. Fraser (1991). Simultaneous ob- servations of Schumann resonances in California and Australia: evidence for intensity modulation by the local height of the D-region, J. Geophys. Res., 96, 15973-15984. Simpson, J.J., and A. Taflove (2005). Electrokinetic ef- fect of the Loma Prieta earthquake calculated by an entire Earth FDTD solution of Maxwell’s equations, Geophys. Res. Lett., 32; doi:10.1029/2005GL022601. Stuart, W.D. (1974). Diffusionless dilatancy model for earthquake precursors, Geophys. Res. Lett., 1, 261- 264; doi:10.1029/GL001i006p00261. Surkov, V.V., and M. Hayakawa (2008). Natural electro- magnetic ULF noise due to fluctuations of ionos- pheric currents, J. Geophys. Res., 113, A11310; doi:10.1029/2008JA013196. Surkov, V.V., and V.A. Pilipenko (2014). Estimate of ULF electromagnetic noise caused by a fluid flow during seismic or volcano activity, Nat. Hazards Earth Syst. Sci., Discuss., 2, 6475-6497. EARTH’S BACKGROUND EM NOISE CHARACTERIZATION Teisseyre, R., and T. Ernst (2002). Electromagnetic ra- diation related to dislocation dynamics in a seismic preparation zone, Annals of Geophysics, 45 (2), 393-399. Thomas, J.N., J.J. Love and M.J.S. Johnston (2009). On the reported magnetic precursor of the 1989 Loma Prieta earthquake, Phys. Earth Planet. Int., 173, 207- 215; doi:10.1016/j.pepi.2008.11.014. Thomas, J.N., J.J. Love and M.J.S. Johnston and K. Yu- moto (2009). On the reported magnetic precursor of the 1993 Guam earthquake, Geophys. Res. Lett., 36, L16301; doi:10.1029/2009GL039020. Thomas, J.N., J.J. Love, A. Komjathy, O.P. Verkho- glyadova, M. Butala and N. Rivera (2012). On the re- ported ionospheric precursor of the 1999 Hector Mine, California earthquake, Geophys. Res. Lett., 39, L06302; doi:10.1029/2012GL051022. Uyeda, S., T. Nagao and M. Kamogawa (2009). Short- term earthquake prediction: current status of seismo- electromagnetics, Tectonophysics, 470, 205-213. Wait, J.R. (1982). Geo-electromagnetism, Academic Press Inc., 265 pp. Warwick, J.W., C. Stoker and T.R. Meyer (1982). Radio emission associated with rock fracture: possible ap- plication to the great Chilean earthquake of May 22, 1960, J. Geophys. Res., 87, 2851-2859. Yumoto, K. (1988). External and internal sources of low-frequency MHD waves in the magnetosphere: A review, J. Geomag. Geoelectr., 40, 293-311. Corresponding author: Cesidio Bianchi, Istituto Nazionale di Geofisica e Vulcanologia, Rome, Italy; email: cesidio.bianchi@ingv.it. © 2015 by the Istituto Nazionale di Geofisica e Vulcanologia. All rights reserved. MELONI ET AL. 10 << /ASCII85EncodePages false /AllowTransparency false /AutoPositionEPSFiles false /AutoRotatePages /None /Binding /Left /CalGrayProfile (Dot Gain 20%) /CalRGBProfile (sRGB IEC61966-2.1) /CalCMYKProfile (U.S. Web Coated \050SWOP\051 v2) /sRGBProfile (sRGB IEC61966-2.1) /CannotEmbedFontPolicy /Warning /CompatibilityLevel 1.3 /CompressObjects /Tags /CompressPages true /ConvertImagesToIndexed true /PassThroughJPEGImages true /CreateJobTicket false /DefaultRenderingIntent /Default /DetectBlends true /DetectCurves 0.1000 /ColorConversionStrategy /LeaveColorUnchanged /DoThumbnails false /EmbedAllFonts true /EmbedOpenType false /ParseICCProfilesInComments true /EmbedJobOptions true /DSCReportingLevel 0 /EmitDSCWarnings false /EndPage -1 /ImageMemory 1048576 /LockDistillerParams true /MaxSubsetPct 100 /Optimize false /OPM 1 /ParseDSCComments true /ParseDSCCommentsForDocInfo true /PreserveCopyPage true /PreserveDICMYKValues true /PreserveEPSInfo true /PreserveFlatness true /PreserveHalftoneInfo false /PreserveOPIComments false /PreserveOverprintSettings true /StartPage 1 /SubsetFonts true /TransferFunctionInfo /Apply /UCRandBGInfo /Preserve /UsePrologue false /ColorSettingsFile (None) /AlwaysEmbed [ true /AndaleMono /Apple-Chancery /Arial-Black /Arial-BoldItalicMT /Arial-BoldMT /Arial-ItalicMT /ArialMT /CapitalsRegular /Charcoal /Chicago /ComicSansMS /ComicSansMS-Bold /Courier /Courier-Bold /CourierNewPS-BoldItalicMT /CourierNewPS-BoldMT /CourierNewPS-ItalicMT /CourierNewPSMT /GadgetRegular /Geneva /Georgia /Georgia-Bold /Georgia-BoldItalic /Georgia-Italic /Helvetica /Helvetica-Bold /HelveticaInserat-Roman /HoeflerText-Black /HoeflerText-BlackItalic /HoeflerText-Italic /HoeflerText-Ornaments /HoeflerText-Regular /Impact /Monaco /NewYork /Palatino-Bold /Palatino-BoldItalic /Palatino-Italic /Palatino-Roman /SandRegular /Skia-Regular /Symbol /TechnoRegular /TextileRegular /Times-Bold /Times-BoldItalic /Times-Italic /Times-Roman /TimesNewRomanPS-BoldItalicMT /TimesNewRomanPS-BoldMT /TimesNewRomanPS-ItalicMT /TimesNewRomanPSMT /Trebuchet-BoldItalic /TrebuchetMS /TrebuchetMS-Bold /TrebuchetMS-Italic /Verdana /Verdana-Bold /Verdana-BoldItalic /Verdana-Italic /Webdings ] /NeverEmbed [ true ] /AntiAliasColorImages false /CropColorImages true /ColorImageMinResolution 150 /ColorImageMinResolutionPolicy /OK /DownsampleColorImages true /ColorImageDownsampleType /Bicubic /ColorImageResolution 300 /ColorImageDepth -1 /ColorImageMinDownsampleDepth 1 /ColorImageDownsampleThreshold 1.10000 /EncodeColorImages true /ColorImageFilter /DCTEncode /AutoFilterColorImages true /ColorImageAutoFilterStrategy /JPEG /ColorACSImageDict << /QFactor 0.15 /HSamples [1 1 1 1] /VSamples [1 1 1 1] >> /ColorImageDict << /QFactor 0.15 /HSamples [1 1 1 1] /VSamples [1 1 1 1] >> /JPEG2000ColorACSImageDict << /TileWidth 256 /TileHeight 256 /Quality 30 >> /JPEG2000ColorImageDict << /TileWidth 256 /TileHeight 256 /Quality 30 >> /AntiAliasGrayImages false /CropGrayImages true /GrayImageMinResolution 150 /GrayImageMinResolutionPolicy /OK /DownsampleGrayImages true /GrayImageDownsampleType /Bicubic /GrayImageResolution 300 /GrayImageDepth -1 /GrayImageMinDownsampleDepth 2 /GrayImageDownsampleThreshold 1.10000 /EncodeGrayImages true /GrayImageFilter /DCTEncode /AutoFilterGrayImages true /GrayImageAutoFilterStrategy /JPEG /GrayACSImageDict << /QFactor 0.15 /HSamples [1 1 1 1] /VSamples [1 1 1 1] >> /GrayImageDict << /QFactor 0.15 /HSamples [1 1 1 1] /VSamples [1 1 1 1] >> /JPEG2000GrayACSImageDict << /TileWidth 256 /TileHeight 256 /Quality 30 >> /JPEG2000GrayImageDict << /TileWidth 256 /TileHeight 256 /Quality 30 >> /AntiAliasMonoImages false /CropMonoImages true /MonoImageMinResolution 1200 /MonoImageMinResolutionPolicy /OK /DownsampleMonoImages true /MonoImageDownsampleType /Bicubic /MonoImageResolution 1200 /MonoImageDepth -1 /MonoImageDownsampleThreshold 1.08250 /EncodeMonoImages true /MonoImageFilter /CCITTFaxEncode /MonoImageDict << /K -1 >> /AllowPSXObjects false /CheckCompliance [ /None ] /PDFX1aCheck false /PDFX3Check false /PDFXCompliantPDFOnly false /PDFXNoTrimBoxError true /PDFXTrimBoxToMediaBoxOffset [ 0.00000 0.00000 0.00000 0.00000 ] /PDFXSetBleedBoxToMediaBox true /PDFXBleedBoxToTrimBoxOffset [ 0.00000 0.00000 0.00000 0.00000 ] /PDFXOutputIntentProfile (None) /PDFXOutputConditionIdentifier () /PDFXOutputCondition () /PDFXRegistryName (http://www.color.org) /PDFXTrapped /Unknown /CreateJDFFile false /SyntheticBoldness 1.000000 /Description << /ENU (Use these settings to create PDF documents with higher image resolution for high quality pre-press printing. The PDF documents can be opened with Acrobat and Reader 5.0 and later. These settings require font embedding.) /JPN /FRA /DEU /PTB /DAN /NLD /ESP /SUO /NOR /SVE /KOR /CHS /CHT /ITA >> >> setdistillerparams << /HWResolution [2400 2400] /PageSize [595.000 842.000] >> setpagedevice