MEMFIS – MULTIPLE ELECTROMAGNETIC FIELD AND

MEMFIS – MULTIPLE ELECTROMAGNETIC FIELD AND INFRASOUND MONITORING NETWORK

I.A. MOLDOVAN1, A.S. MOLDOVAN2, C. IONESCU1, C.G. PANAIOTU2 1National Institute for Earth Physics, P.O. Box MG-2, RO-077125 Bucharest-Magurele, Romania

E-mail: [email protected]; E-mail: [email protected] 2AZEL Designing Group Srl, P.O. Box MG-2, RO-077125 Bucharest-Magurele, Romania

E-mail: [email protected] 3Bucharest University, Faculty of Physics, RO-077125 Bucharest-Magurele, Romania

E-mail: [email protected]

Received May 22, 2009

The paper presents a complex geophysical monitoring and recording system, deployed in Romania, at Plostina observatory (4 sites: PLO1, PLO2, PLO3 and PLO4) and at Surlari site (SULR – Bucharest University). Plostina is located at 45.8512 N latitude and 26.6499 E longitude in the Vrancea (Romania) epicentral zone and is one of the most modern monitoring sites under the administration of the National Institute for Earth Physics (NIEP), Romania. Surlari is located at 44.6798 N latitude and 26.2543 E longitude, outside of the Vrancea epicentral zone, being under the administration of the Bucharest University (BU) and NIEP. Starting with July 2006, NIEP, AZEL – Designing Group S.R.L. and BU joined in a research consortium who’s project – “Complex Multidisciplinary Research System On Precursory Phenomena Associated With Strong Intermediate Vrancea Earthquakes, In Conformity With The Latest International Approaches – MEMFIS” – was financed by the Romanian Ministry of Research and Education, through the Programme “Excellency Research” and had as a final purpose a new and modern geophysical monitoring network, that uses specific instruments providing information on acoustic (both earth’s seismic and atmosphere’s infrasonic activities), electric, magnetic and electromagnetic fields. The main goal is to find the correlations between monitored fields and the preparatory stage of strong intermediate earthquakes in Vrancea zone.

Key words: Earthquake forecast, electromagnetic and infrasonic monitoring, seismic precursors.

1. INTRODUCTION

During the last 20 years there have been reported facts that confirm the interrelations between the tectonic activity and the anomalous changes of the geophysical and geochemical parameters of lithosphere [1]. These changes may reflect modifications in the crustal and subcrustal state of stress and may indicate when a critical stage is reached. As a consequence, different methods of geosciences are

Rom. Journ. Phys., Vol. 55, Nos. 7–8, P. 841–851, Bucharest, 2010

I.A. Moldovan et al. 2

842

trying to monitor tectonic activity, by detecting the regions and phases of critical stress using precursory phenomena connected with earthquakes. We have adopted a system composed by multiple investigation methods intended to detect any anomaly of precursors for the Vrancea (Romania) earthquakes. Electric, magnetic and infrasonic field measurements are recorded by a complex monitoring system of the seismic area.

In some cases, earthquakes generate variations < 10 nT in the local geomagnetic field, mainly by two different phenomena: the piezo-magnetic effect, resulting from variations of the rocks magnetization induced by mechanical or thermal stress and associated to slow variations (from weeks to months) and electro-kinetic effects due to the presence of electric currents in the crust associated to rapid variations from seconds to days [2]. This is the first investigation method used in our monitoring system.

The second parameter that is monitored intends to highlight abnormal infrasonic acoustic waves. This method was tested under laboratory conditions [3]. The pattern of the interchange oscillatory energy between Earth’ surface and atmosphere can warn on the processes of large earthquake preparations. Changes in the infrasound spectrum caused by lithospheric processes have been found. During 5–10 days prior to large earthquakes, the spectrum of infrasonic waves in atmosphere suffers essential changes, this fact providing a proper method for earthquakes prediction [2].

The third parameter used in our system is the vertical component of the atmospheric electric field, which indicate variations of electrical properties of the near-ground air, possibly caused by radon and aerosols with a high content of volatile metals, such as Hg, As and Sb [1]. [4] observed a significant increase of the vertical electric field with an amount of 770V/m from the normal value of 100 V/m one day before 30.08.1986 Vrancea earthquake, Mw = 6.9.

A specific geophysical method associated with the seismic activity and considered to be promising for earthquake prediction is based on the electro-magnetic field monitoring [2]. This can take place both at the ground and (sub) ionospheric level, in different frequency ranges [5–6]. Our working group will improve the geophysical network with a VLF/LF receiver in the near future.

2. THE REASONS

The National Institute for Earth Physics has monitored geomagnetic field in relation with seismicity in the last 10 years, in only one site (Muntele Rosu – MLR – Table 1 – Fig. 1) situated at the extremity of Vrancea epicentral zone.

3 MEMFIS – multiple electromagnetic field

843

Table 1

The location of the magnetic observatories mentioned in this paper

Magnetic Observatory Code Country Latitude Longitude Altitude (m)

MLR Romania 45.49N 25.95E 1360 SUA (INTERMAGNET) Romania 44.68N 26.12E 84

SULR Romania 44.68N 26.25E 97 PLOR2 Romania 45.8502N 26.6438E 694 PLOR3 Romania 45.8539N 26.6455E 708 PLOR4 Romania 45.8512N 26.6499E 656

THY (INTERMAGNET) Hungary 43.10N 17.54E 187

In the last 10 years only 3 Vrancea events exceeded Mw=5.0, one in May 1999 (Mw=5.3), the second in May 2005 (Mw=5.2) and the third in October 27, 2004 (Mw=6.0).

The latter event provided the first opportunity to investigate possible connections between geomagnetic field and local seismicity.

Fig. 1 – The Romanian seismic (black triangles) and geophysic network (gray and light gray triangles). On the figure are also marked the seismic stations that ensures the real time seismic international data exchange and the THY Intermagnet station. In the upper right corner is presented the new Plostina geophysical network comprising seismic, magnetic, electric and infrasonic sensors. The Geomagnetic MLR station is marked with blue and also the National Data Center from Magurele.


844

The magnetic data as well as the global geomagnetic indexes have proved that October 2004 was a quiet month, with a magnetic index Kp < 5, where K variations are all irregular disturbances of the geomagnetic field caused by solar particle radiation within a 3 h interval concerned. Between 26–28 October the solar-terrestrial perturbations were extremely small (Kp < 1), providing a very good medium to observe tectonomagnetic signals. Nevertheless, no anomalies preceding this earthquake were observed on the Bz (vertical magnetic component) and Bx (NS horizontal magnetic component) diagrams as someone could expect, but we had a very large one only on the By – EV horizontal magnetic component (Fig. 2). The

1.10.2004

27.10.20

04

h=98

.6Km

Mw=6

.0

Time (days)

∆Bx (nT)

20

60

100

-20

-60

-100

Bx(SUA)=22.600 nTBx(MLR)=23.200 nT Bx(THY)=21.450 nT

1.10.2004

27.10.20

04

h=98

.6Km

Mw=6

.0

Time (days)

∆By (nT)

20

60

100

-20

-60

-100

By(SUA)= 1670 nTBy(MLR)= 220 nTBy(THY)= 1100 nT

1.10.2004

27.10.20

04

h=98

.6Km

Mw=6

.0

Time (days)

∆Bz (nT)

20

60

100

-20

-60

-100

Bz(SUA)=42.300 nTBz(MLR)=43.390 nT Bz(THY)=42.860 nT

Fig. 2 – Comparison of the magnetic field variation at MLR (blue), THY (black) and SUA (red)

observatories on horizontal (Bx and By) and vertical (Bz) magnetic components.

Bx

By

Bz


845

literature from the last decade [2], [3] and [4] has reported anomalies < 10 nT for earthquakes with Mw > 6.0 and epicentral distances not larger than 100 km. Because the Vrancea region presents intermediate earthquakes, with depths beyond 60 km and the Muntele Rosu Observatory is located at the „border“ of the seimogenic area, such that the hypocentral distances are at the limit of those 130km (and even beyond it) we didn’t expected to record anomalies larger than 10nT.

Despite of this fact, starting with 10th of October 2004, the eastern component (By) of local geomagnetic field recorded at this observatory “left” the general pattern (recorded by other observatories – SUA (SULR) and THY – Table 1 and Fig. 1) and started to decrease (Fig. 2). After a relative steep decaying it reached a low peak of about –40nT in respect with the mean value (220nT), which is far over the expected values of an anomaly from a hypocentral distance of about 122 km. After this, the value of the eastern component (By) has begun to increase slowly towards a normal, mean value. The earthquake occured when the value of this component returned to its mean value.

It has to be remarked that two days after the anomaly started (on 12th of October) the k indexes show higher values which denote solar storms and these are easily visible on the recordings. But this is still “after”. We have tried to imagine what could be the determining factor of this anomaly, recorded only at MLR and only on the horizontal East component. The result is that, at present, we still don’t know if the anomaly has a tectonic origin or not.

This fact has triggered the need to develop a complex monitoring network, offering complete geophysical information and decreasing the uncertainties related to the recording devices.

3. THE SEISMIC ENVIRONMENT AND THE OLD GEOPHYSICAL DATA RECORDING SYSTEM

In Romania, the connection between geoelectromagnetic anomalies and Vrancea earthquakes 3.7 ≤ Mw ≤ 5 was first surveied using seismic and electro-agnetic data collected in the 1998 – 2003 time interval [7–9]. This finding was then extended in 2004 to a broader magnitude range 3.7 ≤ Mw ≤ 6.0. Vrancea epicentral region [45.3N–46.0N and 26.0E–27.3] is the most active intermediate seismic areas in Romania (Fig. 1), where the last devastating earthquake with Mw=7.4 and h=94km occurred in March, 1977. Since then, only three earthquakes with Mw > 6.0, in August 1986, Mw=7.1 and in May 1990, Mw=6.9 and Mw=6.4 occurred. During the geomagnetic monitoring period (1998–2009), only two earthquakes with 5.0 < Mw < 5.4, in April 1999, Mw=5.3 and in May 2005, Mw=5.2, h=147km took place. Until now, in the Vrancea epicentral area only one magnetic observatory was operating, at Muntele Rosu (MLR – Table 1 – Fig. 1 and Fig. 2), in an L-shaped tunnel, situated in a steep part of the southern slope of the Carpathians. At the MLR


846

Observatory, a triaxial Bartington fluxgate magnetometer and a high resolution acquisition system are used to measure and record the geomagnetic field fluctuations. The data from MLR observatory were and are transmitted weekly by e-mail at the National Data Center of the National Institute for Earth Physics from Bucharest (Fig. 1). In order to discriminate local and global phenomena, the MLR geomagnetic data were compared with data provided by the international INTERMAGNET organization (www.intermagnet.org), from stations situated outside the epicentral region and related to the magnetic global activity Kp index, computed for each station.

4. THE MEMFIS MONITORING NETWORK

Previous studies of Vrancea seismogenic zone indicate that observable precursory anomalies in the geomagnetic impedance might precede intermediate Vrancea earthquakes with moment magnitudes Mw≥4.0 [8–9]. This encouraged us to improve the existing geomagnetic monitoring system, by installing 3 new triaxial fluxgate magnetometers, two of them in the epicentral zone (Plostina) and one outside, near Bucharest in the Surlari site. In this way the observations will be improved by differential measuring methods involving simultaneous, data aquisition from sensors located far from the epicentral zone the ULF and sub-ULF bands, from DC to 5Hz.

4.1. EPICENTRAL ZONE MONITORING

To sustain the geomagnetic method, NIEP and AZEL have developed in the epicentral zone a triangle-shaped array that consists in three independent data collecting points (PLO2, PLO3 and PLO4; PLO1 is still under construction), equipped with triaxial fluxgate magnetometers, seismic sensors, infrasound stations (microbarometers) and electrometers.

The monitoring network represents an important step toward a rithmical and sustained observation activity. The three locations (PLO2, PLO3 and PLO4) are surrounding the central point (PLO1) (Figs. 1 and 3).

The distance between the triangle corners is about 500m. Equipment list of each location is as follows: PLO2 – seismic sensors with Quanterra Q330 digitizer, triaxial magnetometer MAG-03MS (+/–100µT) and microbarometer MBAZEL2007 (+/–50Pa) with Hi6 digitizer; PLO3 – seismic sensors with Quanterra Q330 digitizer, triaxial magnetometer MAG-03MS (+/–100µT) and microbarometer MBAZEL2007 (+/–50Pa) with Hi6 digitizer; PLO4 – seismic sensors with Quanterra Q330 digitizer, vertical atmospheric electric field monitor EFM–100 (+/–20KV/m) and microbarometer MBAZEL2007 (+/–50Pa) with Hi6 digitizer and a weather station WS-3600.


847

Fig. 3 – Block diagram of the monitoring network installed at Plostina – Vrancea. At this time, PLO1

location is still under construction.

Each location from Plostina array is connected to a local network by an optic fiber link and a media converter. The local power supply consists in a backup accumulator, permanently charged from the mains power supply (220V/50Hz). The local acquisition system within each location consists in a Hi6 digitizer (6 channels of 24 bits, zero skew-time on three synchronous channels, 2...50SPS, 135dB dynamic range), a TCP/IP server, the power supply unit for the magnetometers (where is the case). For time stamping of the data, the Hi6 digitizer is equipped with a GPS module (Fig. 4).

At the main station (PLO4), a local data acquisition system built around a mini-PC which runs a LabView client/server application is installed (Fig. 5).

Data from the three acquisition points PLO2, PLO3 and PLO4 are gathered here, stored on the hard disk, compressed and then retransmitted to the connected clients. Data are stored in separate files, having a name composed of station’s name and the date of creation. To increase the amount of data that can be stored, files are automatically zipped after closing (at 00:00:00 UTC). The local mini-PC can be remotely operated and managed as long as it uses the Remote Desktop (Windows) application.


848

Fig. 4 – An outter image of PLO3 site and some of the equipments that are involved in the monitoring process: Data Acquisition System, Microbaro-meter MBAZEL2007, Triaxial Fluxgate Magnetometer MAG-03MS. The Electric Field Monitor EFM-100 is installed at PLO4, in the vicinity of the Weather Station WS-3600.

Fig. 5 – A snapshot of the Client Application’s screen, as it can be seen by the final user.


849

Fig. 6 – PLO3 station. Geomagnetic anomalies (down) are shown in the lower pannel; in

the upper pannel, infrasonic acoustic data (up).

Fig. 7 – PLO2 station. Geomagnetic anomalies (down) are shown in the lower pannel in the same time interval like in Fig. 6; in the upper pannel, infrasonic acoustic data (up).

Data transmission is made through the network of a mobile telephony operator, using a 3G/2G firewalled router with a static IP address. The active redundancy of the data transmission is ensured by the using of dual-channel communication: both within NIEP internal network and through the GSM network. The great advantage of using the Internet as a communication medium is that data can be seen from anywhere, using a client application (Fig. 5).

Usage of multiple-point monitoring gives the advantage of an accurate observation: short-range anomalies can be filtered using a validation method that correlates one-station information with the information from the other two stations. Data recorded before, during and after the crustal seismic event that occurred on 06.09.2008 (M=4.4, h=13km) revealed obvious anomalies of the geomagnetic and acoustic fields in the extremely low frequency range (ULF). More clear indications


850

of the correlation between the geomagnetic activity and the infrasonic one can be seen. At PLO2 and PLO3 both geomagnetic and infrasonic anomalies were recorded. The correlation between the signals at these two stations is evident (Figs. 6 and 7). The plots represent the filtered data between 18:00:00.000 and 22:00:00.000 UTC. A 2nd order Butterworth filter was applied. This filter has a band-pass characteristic with cut-off frequencies of 0.01Hz and 0.1Hz, respectively. Because in this extremely low frequency band the seismic event is not visible a large band filter (0.1Hz–4Hz) was applied on the acoustic recording channel. On this plot, the moment of the seismic movement recorded by the microbarometers (at 19:48:04.800 UTC) is marked and clearly visible (Fig. 8).

Fig. 8 – PLO2 station. Infrasonic acoustic data in the frequency range 0.1Hz–4Hz (upper panel)

reveals the moment of the seismic motion arrival. The geomagnetic anomalies are shown in the lower pannel.

4.2. MONITORING AT A REFERENCE STATION

The location from the Surlari Research Facility (SULR) of the University of Bucharest was carefully selected considering several site characteristics that include relatively flat topography away from local villages and also away from large main roads. The research facility is located in the middle of a forest and the nearest village is at 3.5 km. The surface rocks are formed by loess and river sediments with modest magnetic properties. From tectonic point of view it is located in the Moesian Platform not far from the Intramoesian fault. This fault is characterized by sporadic and moderate seismicity (Mw < 5). The distance to the Vrancea seismic area is around 130 km. The sensor house is located at 50 m from the amagnetic building where the recording equipment and the GPS sensor are installed. The Surlari Research Facility is in the nearby vicinity of the Surlari National Geomagnetic Observatory of Romania (SUA). This will allow the calibration of the MEMFIS epicentral network with respect to the national geomagnetic observatory.


851

5. CONCLUSIONS

The multiple-points monitoring system is an important stage of the monitoring process as it increases the confidence in observation data. A number of uncertainties (regarding the accuracy of the data recorded until now) should be removed using the differential measuring system. As we are exploring different parameters (geomagnetic field, vertical component of the atmospheric electric field, infrasonic activity and, in the very next future, the total electrons content – TEC – of the ionosphere above the epicentral area and ionospheric influences on propagation paths of electromagnetic waves at very low frequencies) we are expecting some conclusions regarding the correlation of their anomalies with the seismic activity. Like never before, there is a chance to observe the epicentral area from few different points of view, at different time-scales and from different causal perspectives.

Our expectations regard the chance to take a snapshot of the geophysical medium before, during and after a significant earthquake occurrence and to reveal if there was or wasn’t a noticeable trace of the preparatory stage of it.

REFERENCES

1. P. Pulinets, P.K. Boyarchuk, Ionospheric precursors of earthquakes, Springer-Verlag, Berlin, 2004.

2. M. Hayakawa, Y. Fujinawa, Electromagnetic Phenomena Related to Earthquake Prediction, Terra Scientific Publishing Company, Japan, 1994.

3. J.Z. Li, Z.Q. Bai, W.S. Chen, Y.Q. Xia, Y.R. Liu, Z.Q. Ren, Strong earthquakes can be predicted: a multidisciplinary method for strong earthquake prediction, Natural Hazards and Earth System Sciences 3, 703–712 (2003).

4. M. Hayakawa, R. Kawate, O.A. Molchanov, K. Yumoto, Results of ultra-low-frequency magnetic field measurements during Guam earthquake of August, 1993, Geophys. Res. Lett., 23, 241–244 (1996).

5. M. Hayakawa, Atmospheric and Ionospheric Electromagnetic Phenomena Associated with Earthquakes, Terra Scientific Publishing Company, Tokyo, Japan, 1999.

6. P.F. Biagi, M.E. Contadakis, Seismotectonic electromagnetic effects, precursory phenomena and seismic hazard, Nat. Haz. Earth Syst. Sci., 6, 2006.

7. I.A. Moldovan, A.S. Moldovan, C.G. Panaiotu, A.O. Placinta, Gh. Marmureanu, The Geomagnetic Method On Precursory Phenomena Associated With 2004 Significant Intermediate Vrancea Seismic Activity, Romanian Journal of Physics, 54, 1–2, 249–261, (2009).

8. D. Enescu, B.D. Enescu, I.A. Moldovan, C. Chitaru, New results obtained through the electro-magnetic method for short term prediction of Vrancea earthquakes. Romanian Journal of Physics, 47, 9–10, 901–917 (2002).

9. D. Enescu, I.A. Moldovan, B.D. Enescu, Solar activity, geomagnetic perturbations and Vrancea (Romania) earthquake short-term predictability, Romanian Journal of Physics, 49, 1–2, 145–170 (2004).

Documents

MEMFIS – MULTIPLE ELECTROMAGNETIC FIELD AND