Earthquakes Precursors

The study of earthquakes plays a key role in order to minimize human and material losses when they inevitably occur. Chapters in this book will be devoted to various aspects of earthquake research and analysis. The different sections present in the book span from statistical seismology studies, the latest techniques and advances on earthquake precursors and forecasting, as well as, new methods for early detection, data acquisition and interpretation. The topics are tackled from theoretical advances to practical applications.


Introduction
Strong earthquake of magnitude 7 or more (on the Richter scale) strikes about once a year somewhere in the world and, several times triggers a cascade of follow-on events, such as tsunamis, floods, landslides, nuclear power plant crisis and public health catastrophes in the affected regions. Thus, during the 2004 Sumatra-Andaman earthquake and Indian Ocean tsunami nearly 230,000 people were killed and more than one million people were left homeless in 13 countries surrounding the Indian Ocean. The May 12th, 2008 earthquake in Western Sichuan, China and January 8th, 2010 earthquake in Haiti caused a death toll well over 75,000 and 320,000 people, respectively. The latest M9 Tohoku earthquake of March 11th 2011 in Japan was the biggest recorded earthquake ever to hit Japan. The earthquake triggered extremely destructive tsunami waves of up to 10 meters that struck Japan minutes after the quakes and caused about 26,000 deaths and 3000 injured. Recent catastrophic earthquakes (2004Recent catastrophic earthquakes ( -2011 occurred in Asia, Europe and America have provided and renewed interest in question of the existence of precursory signals related to earthquakes. In these circumstances, the science community is struggling on how to provide early information related to the occurrence time of such events in order to reduce the loss of human life and property. Previous studies (Gotoh et al., 2002;Fraser-Smith et al., 1990;Freund et al., 1999;Hattori et al., 2006;Hayakawa & Fujinawa, 1994;Hayakawa & Molchanov, 2002;Kopytenko et al., 1994;Liu et al., 2004;Ouzounov et al., 2006;Parrot et al., 2007;Pulinets et al., 2004;Stanica & M. Stanica, 2007;Stanica & D.A. Stanica, 2010;Tramutoli et al., 2005;Tronin et al., 2004;Varotsos, 2005) have shown that there were precursory signals observed on the ground and in space associated with several earthquakes. In the last 10 years, the interdisciplinary group for Electromagnetic Study of Earthquakes and Volcanoes (EMSEV) have demonstrated that the existence of the electromagnetic earthquake precursors by terrestrial and satellite observations is not trivial, and it is necessary a wide international cooperation and several more years of research with primary focus in the following directions: (i) what is the possible generation mechanisms of the electromagnetic phenomena; and (ii) whether electromagnetic precursors systematically precede earthquakes. In this respect, taking into account that the seismic-active Vrancea zone, Romania is one of the "hot" subjects in the Eastern Europe, this paper is focused on the specific methodology able to emphasize the short-term electromagnetic (EM) precursory parameters, associated to intermediate depth earthquakes . We consider that one of the realistic mechanisms for triggering such events in the seismogenic volume can be the dehydration of rocks which make fluid-assisted faulting possible. The changes of electrical conductivity occurred before an earthquake, as a sequence of geodynamic processes associated with fluid migration through faulting system developed into and in the close vicinity of the seismogenic volume, could be detected by means of the anomalous behavior of the Bzn parameter taken throughout the frequency range less than 1.66E-2 Hz (Stanica & M. Stanica, 2007;Stanica & D.A. Stanica, 2010). According to the electromagnetic information acquired in 2009-2010 years correlated with seismic events, it is relieved that some days before an EQ occurred, the daily mean variation of the Bzn parameter have an anomalous behavior marked by a significant increase versus its normal distribution identified in non seismic conditions.

Geodynamic models and possible mechanisms of intermediate depth earthquakes
The seismic active Vrancea zone is situated at the curvature of the Carpathians and it is bounded on the north-east by the East European Platform, to southward by the Moesian Platform, and on the north-west by Transylvanian Basin (Fig.1). The hypocenters of the intermediate depth earthquakes are concentrated within a very small seismogenic volume and they are much denser than any other mantle events of intracontinental origin known in the world. In the past century, 4 large seismic events have   (in 1940with moment magnitude Mw7.7, in 1977with Mw7.4, in 1986with Mw7.1, and in 1990) and all of them cause destruction in Bucharest, the capital city of Romania, and shake central and eastern European cities several hundred kilometers away from their epicenters. Several geodynamic models related to the triggering mechanism of the intermediate depth earthquakes have been elaborated in this area. Oncescu (1984) and Oncescu at al., (1984) proposed a double subduction model on the basis of 3-D seismic tomographic images: in their interpretation, the intermediate-depth earthquakes are generated within a vertical surface separating the sinking slab from stable lithosphere. Trifu & Radulian (1989), analyzing the seismic behavior of the Vrancea zone, proposed a model based on the existence of two active zones located at depths of 80-110 km and 120-170 km. Both zones are characterized by local stress inhomogeneities capable of generating large earthquakes. Khain & Lobkosky (1994) suggest that the Vrancea zone results from delamination processes occurred during continental collision and lithosphere sinking into the mantle. Linzer (1996) explains the nearly vertical position of the Vrancea slab as the final rollback stage of a small fragment of oceanic crust.  Stanica et al., 1999); green line is Peceneaga-Camena fault (P-C fault); pink arrows show the direction of the asthenospheric currents; red arrow shows the direction of the torsion process of the relic slab www.intechopen.com Sperner, B., the Collaborative Research Center [CRC] 461 Team, (2005), taking into consideration that the geometry of the subduction zone was not unequivocally defined, proposed four possible configurations for the Vrancea zone: (i) subduction beneath the suture zone; (ii) subduction beneath the fore deep area; (iii) two interacting subduction zones, and (iv) subduction beneath the suture, followed by delamination. Various types of slab detachment or delamination have been proposed to explain the present-day seismic images of the descending slab (Girbacea & Frisch, 1998;Gvirtzman, 2002;Sperner et al., 2001;Wortel & Spakman, 2000). Viscous flows due to the sinking seismogenic slab together with dehydration-induced faulting can be considered as possible triggering mechanism explaining the intermediatedepth seismicity in Vrancea (Ismail-Zadeh et al., 2000) Stanica et al., (2004 show, on the base of the three-dimensional (3D) resistivity tomographic image carried out using magnetotelluric data, that the possible triggering mechanism of the intermediate-depth earthquakes in the Vrancea zone may be the rock response to the active torsion processes sustained by the descending asthenospheric currents and the irregular shape of the relic slab. In their opinion, this torque effect may generate the increase shear stress and drive faulting process within the rigid slab (Fig.2).

Possible generation mechanisms of electromagnetic precursors
This paper is not intended to present an exhaustive analysis of all of literature published in this field. Instead, we have tried to provide only some representative hypothetic mechanisms related to the electromagnetic precursors to earthquakes. The theory of semiconductors launched by Freund (2000) is considered the most comprehensive one. According to this theory, the solid rocks (e.g. granites) begin to crack under the action of a stress exceeding their elasticity limit, what leads to a release of electric charges. These charges, carried by water moving through the rocks fissures, generate currents of high amperage that, in their turn, create disturbances of the magnetic field and, also, infrared signature in the bands of 8μm and 11μm, when the charges are neutralized at the surface of the Earth. In accordance with this theory, to have seismo-electric signals it is necessary that a few compulsory conditions to be fulfilled: -Presence of major tectonic accidents able to produce an important stress; -Existence of quartz or rocks with a rich content of quartz; -Fluid carrying electric charges through porous rock what generates electric currents that lead to disturbances of the magnetic field. The resistivity of porous rocks is changed as a function of compression and shearing (Brace et al., 1965) and may be measured by using passive experiments (magnetotelluric method), or experiments with active methods (Park et al., 1993). Revol et al., (1977) have shown that the magnetic properties of rocks are changed depending on the applied stress and are associated to the changes of stress emphasized on fault rupture, what produces oscillations of the magnetic field of a few nT (Johnston, 1978(Johnston, , 1997; this mechanism is due to the piezomagnetic effect. When a conductive fluid is forced to flow close to a surface with stationary electric charges, an electrokinetic effect appears generating currents that start to flow either through fluid, or through the surrounding rock, what, in conditions imposed by real crustal parameters, may create surface magnetic fields of a few nT (Fenoglio et al., 1995). Another theory supposes the generation of the magnetic signal either by conductive fluid flowing in presence of the magnetic field of the Earth, or by magnetohydrodynamic conversion of the seismic signal into an electric signal during the propagation through a conductive medium (Molchanov et al., 2001). While these mechanisms were proved in laboratory conditions, it is unclear, yet, how this process takes place in conditions rather similar to those specific to the Earth, owing to the lack of measurements in active fault zones.

Theoretical base of the electromagnetic precursors
At the Earth surface the vertical geomagnetic component (Bz) is entirely secondary field and its existence is an immediate indicator of lateral inhomogeneity. For a two-dimensional (2D) structure, the vertical geomagnetic component (Bz) is produced essentially by the horizontal geomagnetic component perpendicular (B  ) to geoelectric structure orientation and, consequently, the normalized Bzn function defined as: should be time invariant in non geodynamic conditions (Ward et al., 1970), but it becomes unstable due to the geodynamic processes and, therefore, it could be used as a precursory parameter of the intermediate depth seismic activity (Stanica and D.A. Stanica, 2010). In order to explain cause (earthquake) -effect (anomalous Bzn) relationship, we introduce the following equations: where: z is vertical resistivity [m = VmA -1 ], f is the frequency [Hz] and the E ║ is the electric field parallel to strike [Vm -1 ], Bz is the vertical component of the magnetic induction [Tesla (T) = V s m -2 ]. Also, it is possible to write the relation: where:  ║ is the resistivity parallel to strike[m], B  is the component of the magnetic induction perpendicular to strike [Tesla (T) = V s m -2 ]. From the relations (2) and (3) we may estimate the normalized function Bzn (f), in terms of resistivities as follows: The relation (4) demonstrates that normalized function Bzn could be linked to the resistivity/conductivity variation along the faulting systems acting as high sensitive path (represented by the Carpathian electrical conductivity anomaly) through the lithosphere and its right part lead to the normalized resistivity defined as: Approximate field solutions were computed for two simple 2D geoelectric structures to illustrate the robustness of the relation (1). Solutions for the sloping interface and vertical contact models were obtained using finite element code (Wannamaker et al., 1986) and the results are presented in Fig. 3. These models represent two extremes in dipping angle of the interface and similarity in the properties of the normalized function Bzn obtained for the www.intechopen.com model (Fig.3, b.) is of interest in selecting the site (GOPS) for continuous monitoring of the geomagnetic field. The normalized frequency scale () is proportional to distance along the x-axis and inverse-proportional to penetration depth (δ 1 ) in medium of resistivity  1 . The electromagnetic skin depth or penetration depth is given by:

Electromagnetic data collection
As we have seen in relation (1), Bzn could be used as precursory parameter of seismic event by measuring the vertical geomagnetic component (Bz) and horizontal component perpendicular to the strike (B  ) which have been collected at the Geodynamic Observatory Provita de Sus (GOPS), placed on the Carpathian electrical conductivity anomaly (CECA). This anomaly is delineated by the Wiese induction arrows, and it can represent a zone of partial melting or of hot highly-mineralized fluids in sedimentary layers, formed at the collisional limit between the both platforms (East European and Moesian) with Carpathian Alpine structures (Fig.4). It is also quit possible that these two varieties of fluid anomalies to co-exist and gradually flow one into another, as indicated by the fact that geoelectric parameters remain fairly constant throughout its entire length (Pinna et al., 1993, Rokityansky & Ingerov, 1999. Induction arrows are vector representations of the ratio of vertical to horizontal magnetic field components. Since vertical magnetic fields are generated by lateral conductivity gradients, induction arrows map can be used to infer the presence, or absence of lateral variation of conductivity/resistivity. In the Wiese convection (Wiese, 1962) the vectors point away from the conductivity anomaly generated by anomalous internal concentrations of current, while in the Parkinson convection (Parkinson, 1959), the vectors point towards anomalous internal concentrations of current. Thus, insulator-conductor boundaries extended through a 2D geoelectrical structure (like CECA) give rise to induction arrows that orientate perpendicular to their geoelectrical strike, and have magnitude proportional to the intensity of anomalous current concentration (Jones & Price, 1970), which are in turn determined by the magnitude of conductivity gradient. In our methodology, it was also supposed that pre-seismic conductivity changes, due to the fluid migration through faulting system, may generate changes of the normalized function Bzn, having magnitude proportional to the intensity of anomalous current concentrations through CECA. The Geodynamic Observatory Provita de Sus (Fig.1) is located at about 100 km towards south-west of seismic active Vrancea zone and the criteria of selection as monitoring site are: -Existence of logistic base able to supply optimal EM data; -Placement on the Carpathian electrical conductivity anomaly where, ideally, the condition for a 2D type geoelectric structure is fulfilled (Fig. 4); -Real time wireless data transfer to the central office (Institute of Geodynamics, Bucharest). In order to select the frequency range where the relation (1) is valid (i.e., existence of a 2D geoelectrical structure and its strike orientation), as a first step in our EM methodology, at the GOPS we made a magnetotelluric sounding using the magnetotelluric (MT) equipment www.intechopen.com GMS-06 (METRONIX -Germany). This geophysical system has 5 channels (two electric Ex, Ey and three magnetic Bx, By, Bz components), 24 bit resolution, GPS, two frequency ranges (LF: 4096sec.-1kH; HF=0.5kH-10kH) and for data processing "MAPROS" software packages. Fig. 4. Carpathian electrical conductivity anomaly and Wiese induction arrows on a tectonic sketch map: 1) main boundaries and fractures (over thrusts) of regional structures; 2) Neogene volcanic rocks; 3) Carpathian fore-deep; 4) Carpathian flysch nape system; 5) Piena and Marmarosh cliff belt; 6) Carpathian electrical conductivity anomaly (CECA); 7) Wiese induction vectors magnitude; 8) seismic active Vrancea zone (intermediate-depth earthquakes); 9) Geodynamic Observatory Provita de Sus (GOPS) used for the electromagnetic data collection (Modified after Rokityansky & Ingerov, 1999) It is well known that the magnetotelluric (MT) method is a passive technique that involves measuring fluctuations of natural electric (E) and magnetic (B) fields in orthogonal directions at the surface of the Earth (Kaufman, & Keller, 1981). The orthogonal components of the horizontal electric (Ex, Ey) and magnetic (Bx, By) fields are related by the complex impedance tensor, Z: Using single site magnetotelluric impedance tensor decomposition technique (Bahr, 1988), it was possible to identify the following two parameters: skewness and strike orientation. The skewness is a dimensionality parameter of the impedance tensor, defined as: Zxx + Zyy Skew = Zxy + Zyx (9) This parameter should be less than 0.3 to interpret the structure as 2D. The tensor impedance from relation (7) can be rotated to obtain the strike orientation of the 2D geoelectrical structure using the relation: 2Re(Zxy +Zyx)(Zxx -Zyy) 1 α =a r c t a n 22 4 (Zxx-Zyy -Z x x+Z yx) Where:  (strike) is rotation angle [ 0 ]. The MAPROS software packages include all these mathematical relations presented above and have been applied to MT sounding carried out at GOPS. This program performs the following basic tasks: www.intechopen.com -Real time data acquisition and processing; -Robust estimation of the magnetotelluric transfer functions; -Real time display of time series and all important electromagnetic parameters (  , ║ , skewness and strike, etc). Thus, on the base of MT results, a 2D geoelectrical structure has been identified on the frequency range less than 1.66 E-2 Hz where skewness < 0.3 (Fig.5) and average strike orientation is N96 0 E (Fig.6). This frequency range is also associated with the intermediatedepth earthquakes interval (70-180km) where EM precursors are generated. These results confirm, once more, that the CECA's geoelectrical structure is of 2D type with strike orientation approximately east-west, and forms not only a tectonic boundary between Moesian Platform and Carpathian Alpine structures, but also represents a peculiar conducting channel extended to the seismic active Vrancea zone (Fig.4). The next step in our study was to realize a continuous monitoring of the geomagnetic components (B  , B ║ , Bz) using the acquisition module MAG-03 DAM (Bartington-England), with 6 channels, 24 bit resolution and three axis magnetic field sensor MAG-03 MSL (frequency range: DC -1kHz). In order to obtain B  component of the geomagnetic field, one of the horizontal components of the three axis magnetic sensor must be orientated perpendicular to strike. The parameters of the data acquisition card are under software control and additional program collects information at each five seconds and stored them, every 60 seconds (Table.1), on the PC HD. Using the wireless connection, all the data are transferred from GOPS to the central unit, placed at the Institute of Geodynamics in Bucharest, for real-time data processing and analysis (Fig.7).
www.intechopen.com  The geomagnetic time series recorded at the GOPS for 7 days interval (April19-April 25, 2009), including the occurrence time of the earthquake of M5.0 (April 25), are presented in Fig. 8. The pre-seismic disturbance of the vertical component (Bz) occurred about 6 days before earthquake. But, as we have seen later on, this disturbance i s m a s k e d b y t h e superposition effect started on March 9, 2009.

Results
In this paper, daily mean distribution of the normalized function Bzn and its standard deviation are performed in the frequency range less than 1.666E-2 Hz, where 2D structural condition is fulfilled. The concept of this analysis is based on the idea that signal associated with solar-terrestrial origin is constant, according to relation (1), while lithospheric origin signal from the underground current flowing along the CECA is considered to have a vertical component (see fig.8). With the other words, the normalized function Bzn shows a small and certain value for its normal trend (in non seismic condition) and increased values in pre-seismic conditions.
To The second anomalous domain, with an average value of Bzn of about 1.855, is extended on the interval July 14 -August 7 and reflects also superposition effect of the two earthquakes of magnitude 5.1 and 5.2, occurred on July 24 and August 5, respectively. The pre-seismic superposition effect of the Bzn started on July 15 and is developed in 9 days interval up to The local variation of earthquakes energy (Es) carried out for the analyzed interval in 2009 year is shown on Fig. 12. The relationship between earthquake magnitude and energy in foci is: logEs = 11.8 + 1.5 M Where: E S is energy [Erg], M is earthquake magnitude. It is quite obvious that the pre-seismic anomalous behaviour of Bzn observed on March 9 ( Fig.9) and July 18 ( Fig.10) respectively, may be correlated with the increased seismic energy reflected by Fig. 13, as follows: more than 15 days interval for the earthquake of M4.0 (March 21) and, about 7 days for earthquakes of M5.1 (July 24). This seismic activity support a possible generation mechanism of the electromagnetic precursors based on the stress generation followed by dehydration of rocks and fluid migration through the faulting system, which may produce concentration of induced currents in highly conductive elongated structure such as Carpathian electrical conductivity anomaly. These induced currents may cause distortion of the vertical geomagnetic component (Fig. 8) which is reflected by increased values of Bzn (Fig.9). The pre-seismic lead time of the normalized function Bzn is between 7days and 15 days for all the data presented in this paper.

Conclusions
The results carried out in this paper are based on the hypothesis according to the preseismic conductivity changes, due to the fluid migration through faulting system, may generate increased values of the normalized function Bzn proportionally with the intensity of anomalous current concentrations through CECA.
The normalized function Bzn carried out at GOPS has been analyzed in order to detect its pre-seismic anomalous behaviour related to the intermediate depth earthquakes with M 4. Before all the earthquakes of M  4 the Bzn distribution exhibit significant enhancements from the normal trend and the pre-seismic lead time is about 7-15 days. The Bzn average value of 1.842 (Fig.9), associated with earthquakes of M≤3.3 occurred on the period January 16 -March 7, represents the threshold limit between the normal trend and its pre-seismic anomalous behaviour, taken as possible earthquake precursor. When anomalous behavior and normal trend domains are much closed, as a multitude of earthquakes of different magnitude occurred at short time intervals, then a superposition effect has been observed. For the Vrancea zones, two correlations between the magnitude of seismic events and Bzn are highlighted (Fig.14): i. earthquake of M  4 is expected to occur when Bzn  1.847; ii. anomalous behaviour of Bzn  1.854 may be use as pre-seismic value for an earthquake of M  5. Sometimes, superposition effects may generate exceptions to the above rules. As this methodology allows us to know always the structure changes after any seismic event (on the base of dimensionality parameters), what permit to use further on the most adequate electromagnetic techniques, it becomes an interesting subject of studying the earthquake mechanism and the associated electromagnetic precursors.