INVESTIGATION OF PRE-EARTHQUAKE IONOSPHERIC ...It has been found that solar terminator is a source of waves with periods ranging from 5 min to 1 h [27] which demonstrate high regularity

INVESTIGATION OF PRE-EARTHQUAKE IONOSPHERIC

ANOMALIES USING VLF/LF INFREP EUROPEAN AND GNSS

GLOBAL NETWORKS

CHRISTINA OIKONOMOU1, HARIS HARALAMBOUS1,2, IREN ADELINA MOLDOVAN3,

RAZVAN GRECULEASA4

1Frederick Research Center, Filokyprou St. 7, Palouriotisa, Nicosia, 1036, Cyprus

E-mail: [email protected] 2Frederick University, Y. Frederickou St. 7, Palouriotisa, Nicosia, 1036, Cyprus

E-mail: [email protected] 3National Institute for Earth Physics, PO BOX MG2, 077125, Magurele, Romania

E-mail: [email protected] 4Sabba S. Stefanescu Institute of Geodynamics,19–21 Jean-Louis Calderon St., Bucharest-37,

Romania, RO-020032, E-mail: [email protected]

Received October 24, 2016

Abstract. Ionospheric TEC (Total Electron Content) variations and Low Frequency (LF) signal amplitude data prior to three large earthquakes (M ≥ 6) in Greece were analyzed using observations from the Global Navigation Satellite System (GNSS) and the European INFREP (International Network for Frontier Research on Earthquake Precursors) networks respectively, aiming to detect potential ionospheric anomalies related to these events and describe their characteristics. For this, spectral analysis on TEC data and terminator time method on LF data were applied. It was found that TEC perturbations appeared few days (1–7) up to few hours before the events lasting around 2–3 hours, with periods 20 and 3–5 minutes which could be associated with the impending earthquakes. In addition, in all three events the sunrise terminator times were delayed approximately 20–40 min few days prior and during the earthquake day.

Key words: Ionospheric earthquake precursors, Total Electron Content (TEC), Spectral Analysis, VLF, terminator time method.

1. INTRODUCTION

The understanding and interpretation of the relation between seismic activity and ionospheric disturbances has received significant attention the last three decades [1–4]. Using a multi-parameter observational dataset and model simulations, [5], examined a large number of earthquakes and proposed the Lithosphere–Atmosphere–Ionosphere Coupling (LAIC) model to describe the consecutive physical processes which lead to the geochemical, atmospheric, ionospheric and magnetospheric anomalies detected up to 12 days before the large seismic events within the earthquake preparation area. This area is designated as a circle with radius ρ = 10

0.43Μ km where M is the earthquake magnitude [6].

Romanian Journal of Physics 62, 816 (2017)

mailto:[email protected]

Article no. 816 Christina Oikonomou et al. 2

As it is described by LAIC model, large amounts of radon can be transported

from earth’s crust to the surface with the aid of various gases emanating from the

active tectonic faults. Radon discharge in the atmosphere leads to ionization of the

air and therefore to the production of huge ion clusters which affect the air

conductivity of the planetary boundary layer and the Global Electric Circuit

locally. Fluctuations of the vertical electric field over active tectonic faults can be

conveyed to the ionospheric F region without any significant attenuation due to the

equality of geomagnetic field lines [7], inducing large-scale electron content

anomalies within the F region. Acoustic gravity waves (AGWs) in regions of

highest conductivity (due to Joule heating) can be also induced which they provoke

small-scale electron concentration disturbances.

The magnetospheric tube over the preparation area is also influenced leading

to enhanced VLF diffusion into tube and to energetic electrons precipitation which

can lower the ionospheric D region resulting to abnormal propagation of VLF

waves [8, 9]. Several studies have proved that radon emission from earth’s crust is

capable of producing enormous amounts of energy release prior to earthquake [10,

11, 12]. In addition, radon release in the atmosphere induces the development of

water condensation nucleus by the produced ions [13, 14]. The latent heat produced

from the water vapor condensation can result to atmospheric thermal anomalies

prior to the seismic events.

Lately, a significant effort has been invested so as to identify possible

earthquake precursors in the ionosphere, utilizing various methodologies and

instruments such as ionosonde observations [15, 16, 17], and GNSS data [18, 19,

20]. A huge amount of studies have employed statistical analysis of GNSS TEC

time series to detect ionospheric TEC anomalies prior to the earthquake [21, 22,

23]. It has been shown that ionospheric precursors are observed between 5 or in

case of very strong earthquakes 12 days to a few hours prior to the earthquake and

that earthquakes should exceed the magnitude of 5 Richter in order to provoke

ionospheric disturbances [24].

In the present study, we follow a multi-techniques and multi-parameters

approach aiming to observe possible ionospheric precursors related to three large

(M ≥ 6) earthquakes that took place in Greece during 2011–2014. In detail, we

applied spectral analysis on GNSS TEC data, as well as the terminator time method

on LF subionospheric signal amplitude data deriving from INFREP European

network.

2. DATA AND METHODOLOGY

The main characteristics of the three seismic events under investigation were

obtained from the earthquake catalog provided by the United States Geological

Survey’s (USGS) Earthquake Hazards Program and are presented at Table 1. TEC

3 Investigation of pre-earthquake ionospheric anomalies Article no. 816

data during the interval 12 days before to the earthquake day were utilized as

derived from dual-frequency phase and code measurements made by GNSS

receivers from the European EUREF network. First we performed spectral analysis

on differential slant TEC (sTEC) defined as the difference of sTEC measurement

between two successive satellite epochs to eliminate hardware biases. The sTEC

described as the integral of the electron density over a line of sight from a ground

receiver to a satellite on the signal propagation path was calculated from the

differential delays of the pseudo-ranges and the phases using an algorithm

developed by [25] that specifies sTEC following the equations given by [26]. The

period of TEC oscillations was chosen up to 40 min so that geomagnetically

induced ionospheric disturbances to be excluded. Figure 1 shows a map of the area

of interest where the GNSS receiver stations and the epicenters of each earthquake

are depicted.

Table 1

List of the three seismic events in Greece studied here and their characteristics

Seismic

Event Mw

Date time

(UT)

R

(km)

Lat

(°)

Lon

(°)

Depth

(km) Region

1 6 4/1/2011

13:29 380 35.66 26.56 59.9

Crete Island,

Greece

2 6.2 6/15/2013

16:11 463 34.45 25.04 10

Crete Island,

Greece

3 6.9 5/24/2014

9:25 927 40.29 25.39 6.43

N. Aegean sea,

Greece

Fig. 1 – Map showing: a) the GNSS receiver stations (magenta dots) used in this study

and b) the epicenters (blue asterisks) of the three seismic events. Numbers in the map correspond to

the number of each seismic event on Table 1. DUTH station was used to study the 3rd event, while

NOA1 station was used for the examination of the 1st and 2nd events. All GNSS stations are lying

within the earthquake preparation zone of each corresponding event.


Then, we analyzed the propagation characteristics of radio signals from radio

transmitters/receivers in the LF (30–300 kHz) band of INFREP network using

terminator time (TT) method in order to identify possible perturbations due to

seismic activity. The terminator times are defined as the times of minimum in

amplitude (or phase) around sunrise and sunset and are found to shift significantly

just around the earthquake. We traced the times of occurrence of the sunrise

amplitude minima 5 days prior and 5 days after the earthquake. We did not

consider sunset minima since they are not clearly identifiable. The epicenters of the

examined earthquakes as well as the signal propagation path between transmitters

and receivers were lying within the Fresnel zone. This zone is an elliptical area for

which the VLF/LF transmitter and receiver are foci.

The data were obtained from the INFREP network which consists of 11

digital radio receivers that measure the power of the radio signals on 14

frequencies (derived from 14 radio transmitters) distributed in the VLF/LF bands,

since February 2009. In Fig. 2 the locations of the INFREP transmitters and

receivers as well as the epicenters of the three earthquakes are shown, whereas the

characteristics of the VLF/LF receivers and transmitters signals which were

recorded continuously with a time resolution of 1min are presented at Table 2. The

signal with propagation path between TRT (T1) transmitter and CIP receiver was

utilized for the examination of ionospheric anomalies prior to the 1 April 2011

earthquake, while the paths CH1 (T7)-CIP and EU1 (T9)-CIP were employed for

the earthquakes during 15 June 2013 and 24 May 2014 respectively (Fig. 2).

Fig. 2 – The locations of INFREP transmitters (T) and receivers (green triangles)

along with the corresponding Fresnel zones (ellipses).

The epicenters of the three selected earthquakes are denoted with black numbered boxes [9].


Table 2

The details of INFREP transmitters and receivers used in this study

SIGN LOCATION LAT

(°)

LON

(°)

FREQ

(Hz) LOCATION

Transmitters T1T1

T1 TRT Polatli 39.76 32.42 180000 Turkey

T7 CH1 Ouargia 31.92 5.08 198000 Algeria

T9 EU1 Felsberg-

Berus 49.28 6.68 183000 Germany

Receivers R1 CIP Nicosia 35.17 33.35 – Cyprus

R4 POR Evora 38.57 –7.9 – Portugal

3. RESULTS

3.1. SPECTRAL ANALYSIS ON TEC DATA

Inspection of TEC spectrograms few days prior to the earthquake on 1 April

2011 revealed enhanced TEC fluctuations with periods around 20 minutes 1 day

before and at the earthquake day during 4–5 UT, as well as 7 days before the event

during 11–14 UT which were most probably related to the impending earthquake.

Furthermore, persistent increased TEC fluctuations were detected at spectrograms

during 7–11 UT and 17–20 UT in all examined days. These fluctuations occurred

around sunrise and sunset respectively and are induced by solar terminator

transition. It has been found that solar terminator is a source of waves with periods

ranging from 5 min to 1 h [27] which demonstrate high regularity [28] and have

amplitudes 0.05–0.1 TECU [29]. TEC spectrograms deriving from several satellites

passing over the earthquake preparation zone and NOA1 receiver during 24 March

2011 are shown in Fig. 3 for the time-window 12–12.5 UT.

The power spectra of TEC measurements several days before the earthquake

on 15 June 2013 demonstrated intensified TEC wave-like perturbations with

periods around 20 minutes during 1–3 UT on 14 and 15 June 2013 which could be

associated to the earthquake shock. Wave-like perturbations with similar periods

were also identified during 16–18 UT in all days under investigation and were

attributed to the solar terminator transition during sunset time. In addition,

intensified TEC fluctuations with periods around 3 minutes were observed few

hours prior to the event at 1–3 UT which could be considered as possible

ionospheric precursors. Figure 4 presents spectrograms of such fluctuations with

periods around 20 and 3 minutes few hours prior to the earthquake moment.


Fig. 3 – Snapshots of TEC fluctuations (T up to 40 min) obtained from measurements of 7 satellites

(PRN) passing over the area of interest during 12–12.5 UT on 24 March 2011. The power spectra of

amplitude are also shown. Map shows the number and position of IPPs (blue asterisks),

NOA1 receiver location (pink triangle) and earthquake epicenter (green asterisk).

Fig. 4 – Same as Fig. 3 but for the earthquake day 15 June 2013 during the time-window 0–1.5 UT

and for periods up to 40 min (upper panel) and up to 5 min (lower panel).


Fig. 5 – Same as Fig. 3 but for the earthquake on 24 May 2014 at around 3 UT 1 day (upper panel)

and at around 13 UT 2 days (lower panel) prior the earthquake.

The spectral outputs of TEC observations prior to the seismic event on

24 May 2014 have shown that high TEC oscillations took place within the

earthquake preparation zone at around 3 and 13 UT one and two days before the

earthquake respectively. These fluctuations are depicted at Fig. 5 for the time

window 2.5–3.5 UT and 13–14 UT on 23 and 22 May 2014 respectively. Similarly

to the previous events, consistent TEC fluctuations were detected on a daily basis

at around 18–20 UT which were related to the solar terminator passage.

3.2. TERMINATOR TIME METHOD ON LF SIGNAL DATA

The propagation paths of LF signals for the three events were largely in the

East-West meridian plane mainly for the seismic events on 1 April 2011 and 15

June 2013, therefore, according to [30] the TT method should be effective in

identifying any seismo-ionospheric perturbation. In Fig. 6 the diurnal LF amplitude

variation for the period 5 days prior to 5 days after the earthquake on

1 April 2011 are shown in the form of 24 hours amplitude-time series. As it can be

seen, the propagation path was in complete daylight over 6–18 UT and in complete

darkness over 19–3 UT.


Fig. 6 – Diurnal variation of the TRT (18 kHz) amplitude received at Portugal receiver (PO)

during the period 27 March–6 April 2011. The red vertical line shows the sunrise minima start

and the red triangles denote the shifts in the sunrise minima times.

This daily pattern was typical from day to day. The average nighttime

amplitude was larger than the average daytime amplitude. The sunrise transition

(time during which the sunrise terminator moves between the transmitter and

receiver producing minima in the received signal amplitude) is also shown in the

same figure. One signal minimum in each day was clearly identified (red arrows).

The time of sunrise minima were shifted up to 40 min during 3 days before and 2

days after the earthquake on 1 April 2011 at Crete Island. They started to become

delayed on 29 March and shifted gradually to a maximum delay one day before the

earthquake. After the earthquake, the minima started shifting back to their normal

positions. Thus, at sunrise at which the LF signal showed minima, anomalous shifts

in TTs were observed, as if the nighttime had been prolonged up to 40 min one day

prior and during the earthquake as seen in our data.


Fig. 7 – Diurnal variation of the CH1 (T7) (19.8 kHz) amplitude received at Cyprus receiver (CIP)

during the period 10–20 June 2013. The red vertical line shows the sunrise minima start and the red

triangles denote the shifts in the sunrise minima times.

Figure 7 demonstrates the diurnal LF amplitude variation for the period 5

days prior to 5 days after the earthquake on 15 June 2013 as daily amplitude-time

series. Similar to the previous seismic event, the propagation path was in complete

daylight over 06–18 UT and in complete darkness over 19–03 UT, and the average

nighttime amplitude is larger than the average daytime amplitude as expected. The

times of sunrise minima were shifted up to 20 min during 5 days before and 1 day

after the earthquake. They started being delayed on 10 June and shifted steadily to


a maximum delay 1 day before the event, while after it the minima gradually

moved back to their regular positions. These anomalous shifts of sunrise minima

were considered as possible ionospheric precursors.

Fig. 8 – Diurnal variation of the EU1 (T9) (19.8 kHz) amplitude received at Cyprus receiver (CIP)

during the period 19–29 May 2014. The red vertical line shows the sunrise minimum start and the red

triangles denote the shifts in the sunrise minima times.

In Fig. 8 the diurnal LF amplitude variations for the period 5 days prior to

5 days after the earthquake on 24 May 2014 are depicted. Unlike the two previous


seismic events, the propagation path between T9 transmitter and CIP receiver was

shorter and thus the sunrise transition time was also short, nevertheless, one sunrise

minimum is denoted at every day. The typical daily pattern with the average

nighttime amplitude being higher than the average daytime amplitude is also

observed. The time of sunrise minima were shifted up to 20 min during 2 days

before the earthquake on 24 May 2014. The delay of the sunrise minima initiated

on 20 June and maximized at 22 May (2 days prior to the event), whereas next the

minima started moving back to their regular positions. These anomalous shifts of

sunrise minima could be described as ionospheric precursors.

4. CONCLUSIONS

The investigation of three strong crustal earthquakes that took place in

Greece by applying spectral analysis on GNSS TEC data has shown that large

ionospheric TEC anomalies can be observed 7 days up to few hours prior to the

earthquake which could be related to the impending earthquake. These anomalies

lasted approximately 2–3 hours and had periods of around 20 minutes. In addition,

similar TEC anomalies that occurred few hours prior to the earthquake presented

periods of 3–5 minutes. With the aid of spectral analysis, it was possible to

discriminate between the ionospheric TEC perturbations related to the earthquakes

and those induced by the solar terminator transition. Furthermore, we were able to

exclude TEC ionospheric anomalies produced by disturbed geomagnetic conditions

by restricting the period band of spectrograms up to 40 min.

The analysis of the same seismic events by applying termination time

method on LF signal data demonstrated that in all events the sunrise terminator

times were delayed approximately 20–40 min few days prior and during the

earthquake day.

The multi-technique and multi-parameter approach which was adopted in

this study is a requirement for the precise identification of earthquake ionospheric

precursors and is highly recommended it for ionospheric-earthquake related

studies.

Acknowledgements. This work was partially carried out within Nucleu Program, supported by

ANCSI, projects no. PN 16 35 03 01/2016, the Partnership in Priority Areas Program – PNII, under

MEN-UEFISCDI DARING Project no. 69/2014, Capacity Program, Module III – Projects supporting

Romania's participation in international research projects, Bilateral cooperation programs Romania –

Cyprus, 2014–2015, project number 759/2014, and the project Investigation of earthquake signatures

on the ionosphere over Europe – ΔΙΑΚΡΑΤΙΚΕΣ/ΚΥ–ΡΟΥ/0713/37 which is co-financed by the

Republic of Cyprus and the European Regional Development Fund (through the ΔΕΣΜΗ2009–2010

of the Cyprus Research Promotion Foundation). The radio data were obtained with the courtesy of

Prof. Pier Francesco Biagi, and INFREP project.


REFERENCES

1. S.A. Pulinets, K. Boyarchuk, Ionospheric Precursors of Earthquakes. Springer, Berlin, 2004.

2. E.L. Afraimovich, E.I. Astafieva, M.B. Gokhberg, V.M. Lapshin, V.E. Permyakova, G.M. Steblov,

S.L. Shalimov, Variations of the total electron content in the ionosphere from GPS data recorded

during the Hector Mine earthquake of October 16, 1999, California. Rus. J. Earth Sci. 6 (5), 339–

354, 2004.

3. J.W. van Dam, W. Horton, N.L. Tsintsadze, T.D. Kaladze, T.W. Garner, L.V. Tsamalashvili, Some

physical mechanisms of precursors to earthquakes, J. Plasma Fusion Res. 8 (199–202), 2009.

4. M.V. Klimenko, V.V. Klimenko, I.E. Zakharenkova, S.A. Pulinets, B. Zhao, M.N. Tsidilina,

Formation mechanism of great positive TEC disturbances prior to Wenchuan earthquake on May

12, 2008, Adv. Space Res. 48 (3), 488–499, 2011.

5. S.A. Pulinets, D. Ouzounov, Lithosphere-Atmosphere-Ionosphere Coupling (LAIC) model – An

unified concept for earthquake precursors validation, Journal of Asian Earth Sciences 41, 371–

382, 2011.

6. I.R. Dobrovolsky, S.I. Zubkov, V.I. Myachkin, Estimation of the size of earthquake preparation

zones, Pageoph. 117, 1025–1044, 1979.

7. S.A. Pulinets, Ionospheric precursors of earthquakes; recent advances in theory and practical

applications, Terrestrial Atmospheric and Oceanic Sciences, 15 (3), 413–436, 2004.

8. I.A. Moldovan, A.S. Moldovan, P.F. Biagi, A.O. Placinta, T. Maggipinto, The INFREP European

VLF/LF radio monitoring network – recent status and preliminary results of the Romanian

monitoring system, Rom. Rep. Phys., 64 (1), 263–274, 2012.

9. I.A. Moldovan, A.P. Constantin, P.F. Biagi, D. Toma-Danila, A.S. Moldovan, P. Dolea,

V.E. Toader, T. Maggipinto, The development of the Romanian VLF/LF monitoring system as

part of the international network for frontier research on earthquake precursors (INFREP), Rom.

J. Phys., 60 (7–8), 1203–1217, 2015.

10. M. Kafatos, D. Ouzounov, S. Pulinets, G. Cervone, R. Sing, Energies associated with the Sumatra

Earthquakes of December 26, 2004 and March 28, 2005, AGU 2007 Fall Meeting, San Francisco,

Paper S42B-04, 2007.

11. Y. Omori, Y. Yasuoka, H. Nagahama, Y. Kawada, T. Ishikawa, S. Tokonami, M. Shinogi,

Anomalous radon emanation linked to preseismic electromagnetic phenomena, Nat Hazards Earth

Syst Sci 7, 629–635, 2007.

12. D. Ouzounov, D. Liu, K. Chunli, G. Cervone, M. Kafatos, P. Taylor, Outgoing long wave

radiation variability from IR satellite data prior to major earthquakes, Tectonophysics, 431, 211–

220, 2007.

13. V.E. Toader, I.A. Moldovan, C. Ionescu, Complex monitoring and alert system for seismotectonic

phenomena, Rom. J. Phys., 60 (7–8), 1225–1233, 2015.

14. V.E. Toader, I.A. Moldovan, A. Marmureanu, C. Ionescu, Detection of events in a multidisciplinary network monitoring Vrancea area, Rom. J. Phys., 61 (7–8), 1437–1449, 2016.

15. V.H. Rios, V.P. Kim, V.V. Hegai, Abnormal perturbations in the F2 region ionosphere observed

prior to the great San Juan earthquake of 23 November 1977, Adv. Space Res. 33 (3), 323–327,

2004.

16. J.Y. Liu, Y.I. Chen, Y.J. Chuo, C.S. Chen, A statistical investigation of pre-earthquake

ionospheric anomaly, J. Geophys. Res., 111 (A05304), 2006.

17. Y.B. Yao, P. Chen, S. Zhang, J.J. Chen, F. Yan, W.F Peng, Analysis of pre-earthquake

ionospheric anomalies before the global M = 7.0 + earthquakes in 2010, Nat. Hazards Earth Syst.

Sci. 12 (3), 575–585, 2012.

18. B. Carter, A. Kellerman, T. Kane, P. Dyson, R. Norman, R. Zhang, Ionospheric precursors to

large earthquakes, A case study of the 2011 Japanese Tohoku Earthquake, J. Atmos. Sol-Terr.

Phys. 102, 290–297, 2013.


19. P.I. Nenovski, M. Pezzopane, L. Ciraolo, M. Vellante, U. Villante, M. De Lauretis, Local changes

in the total electron content immediately before the 2009 Abruzzo earthquake, Adv Space Res 55

(1), 243–258, 2015.

20. E.I. Nastase, C. Oikonomou, D. Toma-Danila, H. Haralambous, A. Muntean, I.A. Moldovan,

Investigation of ionospheric precursors of earthquakes in Romania using the Romanian

GNSS/GPS Network, Rom. J. Phys., 61 (7–8), 1426–1436, 2016.

21. J. Li, G. Meng, M. Wang, H. Liao, X. Shen, Investigation of ionospheric TEC changes related to

the 2008 Wenchuan earthquake based on statistic analysis and signal detection, Earthq. Sci.

22(5), 545–553, 2009.

22. M. Contadakis, D. Arabelos, C. Pikridas, S. Spatalas, Total electron content variations over

southern Europe before and during the M 6.3 Abruzzo earthquake of April 6, 2009, Annals of

Geophysics 55 (1), 2012.

23. Q. Song, F. Ding, T. Yu, W. Wan, B. Ning, L. Liu, B. Zhao, GPS detection of the coseismic

ionospheric disturbances following the 12 May 2008 M7. 9 Wenchuan earthquake in China,

Science China Earth Sciences 58 (1), 151–158, 2015.

24. S.A. Pulinets, A.D. Legen’ka, Spatial-temporal characteristics of the large scale disturbances of

electron concentration observed in the F-region of the ionosphere before strong earthquakes,

Cosmic Research 41 (3), 1–10, 2003.

25. B. Muslim, J. Efendi, D.R. Suryanal, Developing near real time TEC computation system from

GPS data for improving spatial resolution of ionospheric observation over Indonesia, Proc of 1st

International Seminar on Space Science and Technology, Serpong, Indonesia, Dec. 3, 2013.

26. Y. Gao, Z.Z. Liu, Precise ionosphere modeling using regional GPS network data, J. Global Pos.

Sys. 28–24, 2002.

27. V. Gorokhov, Studying the Effects of the Terminator and Interrelated Processes in the Mid

latitude Ionospheric E and F Regions as Inferred from the Data of the “Soika6000” Digisonde,

Ionospheric Radio Propagation 91–99, 1989.

28. I.V. Karpov, F.S. Bessarab, Model studying the effect of the solar terminator on the thermospheric

parameters, Geomagn Aeron 48, 209–219, 2008.

29. E.L. Afraimovich, First GPS-TEC evidence for the wave structure excited by the solar terminator,

Earth Planets Space 60 (8), 895–900, 2008.

30. S. Maekawa, M. Hayakawa, A statistical study on the dependence of Characteristics of VLF/LF

terminator, IEEJ Transactions on Fundamentals and Materials 126 (4), 220–226, 2006.

Documents

INVESTIGATION OF PRE-EARTHQUAKE IONOSPHERIC ...It has been found that solar terminator is a source of waves with periods ranging from 5 min to 1 h [27] which demonstrate high regularity