Space Research Institute, Graz, Austria  Workshop on ‘Earthquakes: Ground-based and Space Observations‘

1-2 June 2007

Investigation of electromagnetic ULF/ELF-phenomena possibly related to seismic events in south Europe during 2004 and 2005 using South European Ground Magnetometer (SEGMA) and DEMETER data.(1) G. Prattes/ (2) K. Schwingenschuh/ (3) W. Magnes/ (4) M. Stachel/ (5) M. Boudjada/ (6) M. Horn/ (7) M. Vellante.(2,3,4,5) Space Research Institute, Austrian Academy of Sciences, Schmiedlstrasse 6, A-8042 Graz, Austria(1) Department of Communication and Wave Propagation, technical University Graz, Austria.(6) Institute of Physics, Department of Geophysics, Astrophysics and Meteorology (IGAM), Karl-Franzens-University Graz, Austria.(7) Department of Physics, University of l’Aquila (Italy).

IntroductionSeismic events can possibly produce electromagnetic waves in the ULF/ELF frequency range. Two seismic events were investigated: 1. magnitude 5.5, depth=15km, date=july 10 2005, region of Podgorica Montenegro.2. magnitude=5.4, depth=5.1km, date=july 12 2004, region of Bovec Slowenia.In the frame of the DEMETER and the SEGMA projects these phenomena are jointly investigated by ground based and space observations. ULF/ELF magnetic field data in mid and south Europe are provided by the South European Ground Magnetometer (SEGMA) chain. The sampling rate of the data analysed is between 1 and 16 Hz. The main problem occurring during analysing the data is to separate magnetic field disturbances which are caused by seismic events and geomagnetic activity or man made noise. One important parameter for the present analysis is the ratio of the vertical to the horizontal component (polarization) of the magnetic field. In order to differentiate man made noise from seismomagnetic events a quality figure of the stations has been determined. The global geomagnetic variations have been sorted out by using geomagnetic indices. ULF/ELF magnetic field data in Nagycenk (Hungary), L’Aquila (Italy) and Ranchio (Italy) have been correlated with the seismic events. Comparisons between results at 1Hz sampling frequency and 16Hz sampling frequency were made.

SEGMA stationsMagnetic field data is measured at the following locations signed with yellow markers in the map. Data from the stations in Ranchio(RNC), Nagycenk(NCK), Castello Tesino(CST) and L’Aquila(LAQ) were analysed.

StationGeographic Coordinates

Corrected Geomagnetic Coordinates

Name Code Lat.[°N]

Long.[°E]

Lat.[°N]

Long.[°E]

Castello Tesino

CST 46.0 11.7 40.7 87.0

Nagycenk NCK 47.6 16.7 42.6 91.7

Ranchio RNC 43.97 12.08 38.22 86.71

L’Aquila AQU 42.38 13.32 36.30 87.35

Panagyurishte PAG 42.51 24.18 36.98 97.21

Event locations 2004Information about the seismic event was supplied by DEMETER and ZAMG. It is designed to investigate disturbances in the Earth's ionosphere due to seismic and volcanic activities. Earthquakes in Austria and global heavy earthquakes are registered by ZAMG. The seismic event is depicted with the red marker.

Date Time Geographiccoordinates

Magnitude

Depth

2004-07-12 13:04 46.31°N 13.61°E 5.1 5.4km

Distances to stations [km]

CST NCK RNC LAQ

154 275 291 446

Date Geographiccoordinates

Intensity* Distances [km]

CST NCK RNC

2004-06-18 47.48°N 13.36°E

5 215 251 400

2004-06-28 47.60°N 16.46°E

4-5 407 20 526

2004-06-30 47.45°N 13.21°E

4-5 200 260 390

2004-06-30 47.45°N 13.21°E

3 200 260 390

2004-07-12 46.31°N 13.61°E

6 154 275 291

2004-07-14 South Austria 3-4

2004-07-22 47.40°N 12.05°E

4 160 350 380

2004-08-21 46.52°N 14.30°E

4-5 212 214 330

2004-08-22 46.52°N 14.30°E

3-4 212 214 330

ZAMG registered the following seismic events in Austria from June to August 2004.

Event locations 2005Date Time Geographic

coordinatesMagnitude Depth Distances [km]

CST NCK RNC

LAQ

2005-07-10 13:10 42.39°N

19.8°E 5.5 10km 750 620 650 530ZAMG registered the following seismic events in Austria from June to July 2005.

Date Geographiccoordinates

Intensity* Distances [km]

CST NCK RNC

2005-06-06 47.18°N 10.81°E 3 158 450 371

2005-06-17 48.16°N 15.54°E 4 385 109 542

2005-06-22 47.57°N 12.56°E 4 190 310 400

2005-06-23 47.55°N 13.70°E 4 230 226 415

2005-06-24 47.24°N 10.93°E 4 155 435 375

2005-07-05 47.23°N 10.50°E 4 170 466 380

2005-07-08 46.52°N 14.66°E 4 230 196 349

2005-07-25 47.82°N 16.23°E 6 405 45 533

2005-07-31 47.62°N 14.33°E 3-4 277 180 440

*Intensity scale: “Europäische Makroseismische Skala, 1998 (EMS 98)”

Measuring system

22 YXH

The components of the measured Earth’s magnetic field are X,Y and Z. The X component is along the geographic meridian (positive northward), the Y component is along the geographic latitude (positive eastward) and the Z component is vertical and positive towards the Earth’s centre.The horizontal component of the magnetic field was calculated as

The CHIMAG fluxgate magnetometer Measurement range: +/- 512nT Maximum compensation field: +60000nT in X and Z

+/- 30000nT in Y direction Accuracy: 8pT Sampling frequency: 64Hz (highest possible) Frequency resolution: 1Hz.

Seismic waves

At s-waves or transversal waves the particle movement is across the direction of propagation. The different wave forms are explained in the figure above.

Types of stress: •tensional •compressional •shear stress.

Produced elastic waves: •body waves •surface waves

Concerning body waves we distinguish p-waves and s-waves. At p-waves or longitudinal waves the particle movement is along the direction of propagation. The elastic body wave passes through the medium as series of dilatations and compressions.

7.13 s

p

v

vVelocity of elastic waves VP ~ 5 – 7 km/s in typical Earth’s crust.

Electromagnetic wavesThere are two possible ways to explain the emission of ULF perturbations.1. ULF emission due to the mechanism of microfracture electrification.

This effect appears as fast fluctuations of charge and electromagnetic field from an ensemble of opening microfractures, so called cracks (Molchanov and Hayakawa 1996). Dimension: 10-4m-10-1m Time scale: 10-4s - 10-7s. The velocity of the opening cracks is in the range of the seismic velocity of p- and s- waves~103m/s.If the rate of microcracks is rather high, the process will lead to origination of wideband electromagnetic noise. This noise might dissipate outside the source region and might produce noise like ULF emissions on the ground surface.

2. The same assumption of microfracturing, but the electromagnetic perturbation is caused by a magneto-hydro-dynamic (MHD) or inductive effect from a propagating seismic wave. An inductive seismo-magnetic effect from a solitary seismic wave was assumed in relation to the observation of so called magnetic forerunners or coseismic electric signals were observed as well (Molchanov, Surkov, Hayakawa, Kulchitsky (2001)).

Electromagnetic waves

- - - -

y

+ + + +

z

x

The figure shows the microcurrents ic(t) during a crack opening. The crack has got an elliptical cross section in the x - z plane and circular cross section in the z - y plane. The point x0 is a centre of an ellipsoidal crack. The circular lines show the conductive currents which destroy the charge fluctuation Q0 within the opening crack. The equivalent linear current is shown below by the solid arrow. It is supposed that some charge production Q0 exists inside the crack and on its boundary. A well known charge production is the piezoelectric effect.

The charge

Sc(t) … lateral surface of crack q0 …surface density of charge on the crack boundary.

00 qtStQ c

Why ULF

20

where σ the conductivity and μ is the magnetic constant. The skin depth

According to K. Simonyi the attenuation coefficient alpha for low frequencies is given as

0

21

Am

Vs70 104

Material Conductivity σ δ10mHz δ100mHz δ1Hz

Igneous rock ~ 10-4 S/m 5.033*105m 1.591*105m 5.033*104m

Metamorphous/Limestone

~ 10-3 S/m 1.591*105m 5.033*104m 1.591*104m

Sediment ~ 0.01 S/m 5.033*104m 1.591*104m 5.032*103m

Sea water ~ 5 S/m 2.250*103m 711.76m 225.08m

Graphite ~ 100 S/m 503.29m 159.15m 50.33m

Data processing methods

To estimate the wave intensity monthly mean and standard deviation were calculated. Comparison between the ULF wave activity and ΣKp helps to distinguish between space geomagnetic pulsations and emissions of pulsations which are not emitted in space. To decide whether the emission of pulsations is under the ground or not, the ratio of Z/H, so called polarization ratio is of essential importance. Data on day are much more variable than at night, as described in many publications, e.g. Hayakawa. So the night time period from L.T. 22.00h – 02.00h was chosen for detailed analysis.The wave forms of the field components in the H and Z direction during each 30 minutes interval are subjected to a FFT analysis. The data for one day consist of 8 such 30 minutes intervals. Further the power spectral density was calculated. The upper analyzable frequency is about 0.45Hz. The bandwidth from zero to 0.45Hz was divided into three parts from 10mHz – 50mHz, 50mHz – 100mHz and from 100mHz – 450mHz.

Geomagnetic activityGeomagnetic activity is expressed by the ΣKp. High ULF activity during the intervals with high ΣKp values is clearly associated with geomagnetic activity. Possible ULF waves in the frequency range of Pi2 and Pc4 waves appear around midnight of days with high geomagnetic activity. These pulsations are usually H-polarized causing a small polarization ratio on geomagnetic disturbed days.

Quiet day: Station: NCKdate: 10.4.2005 (22.00-02.00)H-componentKp = 4o,(Q1)

Disturbed day: Station: NCKdate: 5.4.2005(22.00–02.00) H- componentKp = 36o, (D1)

Observations 2004The general situation is depicted for the stations in Castello Tesino, Nagycenk and Ranchio below. The bar plot shows the polarization in the low frequency band (10mHz-50mHz) after one day averaging from the 1st of June to the 31st of August 2004. The vertical full lines show the seismic events.

Date Geographiccoordinates

Intensity*

Distances [km]

CST NCK RNC

2004-06-28 47.60°N

16.46°E

4-5 407 20 526

2004-06-30 47.45°N

13.21°E 4-5 200 260 390

2004-06-30 47.45°N

13.21°E 3 200 260 390

2004-07-12 46.31°N

13.61°E

6 154 275 291

2004-07-14 South Austria 3-4

2004-07-22 47.40°N

12.05°E 4 160 350 380

2004-08-21 46.52°N

14.30°E 4-5 212 214 330

2004-08-22 46.52°N

14.30°E 3-4 212 214 330

Observations 2004The three bar plots compare the polarization of a geomagnetic quiet day (10.4.2005) with a disturbed day (5.4.2005) measured in Nagycenk and one day 19 days before an earthquake (23.6.2004) measured in Castello Tesino. The polarization on the quiet day is higher than on the disturbed day. The highest ratios are on the 23rd of June 2004 which was the day before the event. The time period from the 22nd to the 27th of June was geomagnetical quiet.

Polarization mean values:Quiet day 0.3126Disturbed day 0.0755Day before the event 1.094

If the source of emission is under the ground the ratio is expected to be higher, even greater than 1 during concerned half hour intervals.

Observations 2004The Z magnetic field component was analysed before the event happened at 13:04 on 12th of July 2004 using the IMF- toolbox. The magnetic field data from the stations in CST, NCK and RNC were Pc4 filtered and compared on the left.

CST_maximum 3.7267nT CST_minimum -3.3021nT NCK_maximum 0.1655nT NCK_minimum -0.1950nT

RNC_maximum 0.5071nT RNC_minimum -0.5072nT

The upper panel on the right side shows the Z component of the magnetic field, in the lower panel the standard deviation calculated at 16Hz sampling frequency is depicted for the three stations.

Future outlook/ReferencesThe correlation between the so far achieved results and the earthquake is slightly positive. The aim in future is to find more reliable mathematical methodologies to find out whether the emission of ULF perturbations is related to seismic activity or not. In this context the generation mechanisms of ULF seismogenic electromagnetic waves will be further investigated. Referring to wave propagation, skin depths and attenuation of electromagnetic waves in the lithosphere contacts to magnetotelluric working groups and geophysical departments are planned.ReferencesHayakawa M., Ryusuke K., Molchanov O. and Kiyohumi Y.: Results of ultra-low-frequency magnetic field measurements during the Guam earthquake of 8 August 1993. Geophysical Research letters, Vol. 23, No.3, pages 241-244,February 1,1996.Molchanov O., Hayakawa M.: On the generation mechanism of ULF seismogenic electromagnetic emission. Physics Of The Earth and Planetary Interiors 105(1998)201-210.Molchanov O., Kulchitsky A., Hayakawa M.:Inductive seismo-electromagnetic effect in relation to seismogenic ULF emission. Natural Hazards and Earth System Science(2001)1:61-67.Kushwah V.K., Singh B.: Initial results of ultra low frequency magnetic field observations at Agra and their relation with seismic activities Current Science Vol. 87, No.3,10 August 2004.Kotsarenko A., Perez Enriquez R., Lopez Cruz-Abeyro J.A., Koshevaya S., Grimalsky V., Zuniga F.R.: Analysis of the ULF electromagnetic emission related to seismic activity, Teoloyucan geomagnetic station, 1998-2001. Natural Hazards and Earth System

Science(2004)4:679-684.