- Email: [email protected]
Can electromagnetic disturbances related to the recent great earthquakes be detected by satellite magnetometers?
Tectonophysics 431 (2007) 173 – 195 www.elsevier.com/locate/tecto
Can electromagnetic disturbances related to the recent great earthquakes be detected by satellite magnetometers? G. Balasis ⁎, M. Mandea GeoForschungsZentrum Potsdam, Telegrafenberg, 14473 Potsdam, Germany Received 13 January 2006; accepted 24 May 2006 Available online 8 December 2006
Abstract On December 26, 2004 the world's fourth largest earthquake since 1900 and the largest since the 1964 Prince William Sound, Alaska earthquake, occurred off the west coast of northern Sumatra with a magnitude of 9.3. On March 28, 2005 another event of magnitude 8.7 took place in the same region. The December 26, 2004 earthquake has prompted scientists to investigate possible electromagnetic signatures of this event, using ground magnetic observations. Iyemori et al. [Iyemori, T. et al., 2005. Geomagnetic pulsations caused by the Sumatra earthquake on December 26, 2004. Geophys. Res. Lett., 32, L20807, doi:10.1029/ 2005GL024083.] have suggested that a 3.6 min long geomagnetic pulsation, observed shortly after this event, was generated by the earthquake. They have speculated that a 30 s magnetic pulsation was also caused by the earthquake. Here for the first time, CHAMP satellite magnetic and electron density data are examined to find out if electromagnetic signatures which are possibly related to these recent megathrust earthquakes are observed in satellite magnetic data. We have shown that some specific features are observed after the two earthquakes, with periods of about 16 and 30 s. Our results favor an external source origin for the 30 s pulsation. Moreover, after more than 1 h, CHAMP magnetic data indicate the existence of a feature characterized by the same parameters (duration, amplitude, and frequency content), which could be associated with each earthquake, respectively. Further investigations are required in order to answer the question of whether these signals can be associated with earthquakes and to assign their possible usefulness with respect to earthquake development. © 2006 Elsevier B.V. All rights reserved. Keywords: Sumatra earthquake; CHAMP satellite; Geomagnetic field; Ionosphere; Wavelets
1. Introduction Earthquake physics involves a broad range of themes related to the Earth's system, from tectonic plates to volcanism, from microscopic processes to chemical re⁎ Corresponding author. Now at Institute for Space Applications and Remote Sensing, National Observatory of Athens, Metaxa and Vas. Pavlou, Penteli, 15236, Athens, Greece. E-mail addresses: [email protected], [email protected] (G. Balasis). 0040-1951/$ - see front matter © 2006 Elsevier B.V. All rights reserved. doi:10.1016/j.tecto.2006.05.038
actions, and from generation of electric charges to shockacoustic waves in total electron content. In this study, we are only interested in possible earthquake associated signatures, pre- or post-seismic, in satellite magnetic data. In the following we focus on electromagnetic disturbances possibly related to earthquakes. Over the last few decades, efforts have been made to improve the overall knowledge surrounding earthquakes, from physical mechanisms to pre- and postseismic signals. Recently, efforts at better understanding the physical properties of precursory ULF pre-seismic
174
G. Balasis, M. Mandea / Tectonophysics 431 (2007) 173–195
electric signals (SES) have been intensified (Varotsos and Lazaridou, 1991; Varotsos et al., 1993, 1996; Kagan and Jackson, 1996; Pham et al., 1998; Tzanis et al., 2000; Uyeda et al., 2000, 2002; Varotsos, 2005; Balasis et al., 2005a). On the other hand both acoustic as well as electromagnetic emissions in a wide frequency spectrum ranging from VLF to VHF are produced by microcracks, which can be considered as the so-called precursors of general fracture. These precursors are detectable on both a laboratory and a geological scale (Gokhberg et al., 1982; Ogawa et al., 1985; Vallianatos and Nomikos, 1998; Mavromatou and Hadjicontis, 2001; Freund et al., 2004). Several experimental results illustrating the connection between anomalous VLF–VHF electromagnetic and acoustic phenomena with earthquake preparation were presented in a rather comprehensive collection of papers in a book by Hayakawa and Molchanov (2002). Recently, kHz–MHz sequences of electromagnetic anomalies have been detected from a few days to a few hours prior to recent destructive earthquakes in Greece (Eftaxias et al., 2001, 2002). Kapiris et al. (2004a,b, 2005) and Contoyiannis et al. (2005) have advanced the hypothesis that the detected electromagnetic phenomena are probably precursors emitted from the focal area during micro-fracturing processes. Earthquakes can also produce atmospheric and ionospheric disturbances by dynamic coupling (Liu et al., 2006): vertical vibrations of the Earth's surface lead to pressure waves in the neutral atmosphere that grow in amplitude by several orders of magnitude as they attain ionospheric heights. The first papers concerning the ionospheric effects connected with earthquakes were published after the great Alaska earthquake in 1964 (see Pulinets and Boyarchuk (2004) and references therein). In principle, ionospheric perturbations after strong earthquakes occur just after the shock and are due to acoustic waves which are amplified through the atmosphere because of the decreasing atmospheric density with increasing height (Trigunait et al., 2004). Ionospheric perturbations can also be observed above a seismic zone a few days before an earthquake, but not all earthquakes produce such phenomena (Trigunait et al., 2004). In our study we focus on two great earthquakes, both related to the Sumatran region. The first, referred to hereinafter as SUM1, is the 9.3 magnitude earthquake of December 26, 2004, which occurred at 00:58:50.7 UT off the west coast of northern Sumatra, Indonesia (3.3° N, 95.9° E). The second (hereinafter SUM2) is the 8.7 magnitude earthquake of March 28, 2005 that occurred at 16:09:36.3 UT in the same region (2.074° N, 97.013° E). In this preliminary stage of our study we have chosen to concentrate only on these two events and to analyze them
carefully, considering that they might be generated by the same mechanisms. A few studies have been published related to possible electromagnetic signatures of SUM1. Röder et al. (2005) have shown that strong electrical signals, which correspond to SUM1, were recorded by an electrostatic sensor (this device works like a capacitor and detects short-term variations of the horizontal component of the Earth's electrostatic field) in Italy. These signals arrived at the station practically instantaneously with the occurrence of SUM1 and were detected up to several hours before the onset of SUM1. Iyemori et al. (2005) have argued that a geomagnetic pulsation with a period of approximately 3.6 min was observed at Phimai in Thailand, 12 min after the origin time of SUM1. At Tong Hai in China, some 10° north of Phimai, only a short period pulsation (about 30 s) was observed. At higher latitudes, no magnetic pulsations with these periods were observed. Liu et al. (2006) have reported two giant ionospheric disturbances at 01:19 and 04:10 UT on December 26, 2004 as observed by a network of digital Doppler sounders in Taiwan. The first disturbance, excited mainly by Rayleigh waves which consist of a packet of short-period Doppler shift variations, resulted in vertical ionospheric fluctuations with a displacement of about 200 m. The second disturbance is attributable to coupling of the atmospheric gravity waves excited by broad crustal uplift together with the following big tsunami waves around the earthquake source zone. The accompanying ionosonde data suggested that the gravity waves in the atmosphere may have caused the ionosphere to move up and down by about 40 km. 2. Data processing In this study we use CHAMP vector magnetic data to investigate if electromagnetic signatures could be related to the SUM1 and SUM2 events. We first describe the dataset used and then the wavelet analysis technique (Balasis et al., 2005b) applied which emphasizes high frequency signals. 2.1. Data CHAllenging Minisatellite Payload (CHAMP) is a small near polar, low altitude (∼ 400 km) satellite mission, launched in July 2000. With its highly precise magnetometer instruments, CHAMP has been generating high quality magnetic field measurements for more than 5 years. These data have been widely used to produce new models of the Earth's internal magnetic
G. Balasis, M. Mandea / Tectonophysics 431 (2007) 173–195
field (Maus et al., 2002, 2006; Olsen et al., 2006). CHAMP magnetic data have also given important hints for the geometry of the sources assumed in global electromagnetic induction studies which aim at deriving the electrical conductivity of the Earth's mantle (Balasis et al., 2004; Balasis and Egbert, 2006). The high accuracy of CHAMP magnetic measurements has led to the first detection of the magnetic field generated by the ocean tidal currents (Tyler et al., 2003). Moreover, CHAMP has offered valuable measurements of fields related to the systems of the Earth's ionospheric and magnetospheric currents (Lühr et al., 2002, 2004; Stolle et al., 2006). Recently, Purucker and Ishihara (2005) have used CHAMP data to produce magnetic images of the region surrounding SUM1 and SUM2 providing a current and historical view of subduction in the region. The CHAMP satellite provides 1 s measurements recorded by two different instruments: a three component (northern, eastern, radial) vector fluxgate magnetometer and an absolute scalar Overhauser magnetometer (the prime purpose of which is to calibrate the fluxgate measurements). Precise positioning and timing are associated with each of the measurements. Furthermore, plasma density measurements obtained by a planar Langmuir probe, onboard the CHAMP satellite, are considered. The planar Langmuir probe performs a sweep for determining the electron density every 15 s. 2.2. Analysis technique To analyze these highly accurate data, we have developed specific codes based on wavelet transforms. The wavelet transform is superior to the Fourier spectral analysis, providing excellent decompositions of even transient, non-stationary signals (Kaiser, 1994). It has the ability of providing a representation of the signal in both the time and frequency domains. In contrast to the Fourier transform, which provides the description of the overall regularity of signals, the wavelet transform identifies the temporal evolution of various frequencies. This property suits the signals under investigation, because they are not stationary by their nature, and have a time varying frequency content. Wavelet spectral analysis allows quantitative monitoring of the signal evolution by decomposing a time series into a linear superposition of predefined mathematical waveforms, each with finite duration and narrow frequency content (Kumar and Foufoula-Georgiou, 1997). Thus, the frequency range of the analyzing wavelets corresponds to the spectral content of time series components. Wavelet analysis is becoming a common tool for analyzing localized variations of power
175
within a time series. By decomposing a time series into time-frequency space, one is able to determine both the dominant modes of variability and how those modes vary in time (Alexandrescu et al., 1995). The advantage of analyzing a signal with wavelets as the analyzing kernel, is that it enables one to study features of the signal locally with a detail matched to their scale. Owing to its unique time-frequency localization, wavelet analysis is especially useful for signals that are non-stationary, have short-lived transient components, have features at different scales, or have singularities. We performed a wavelet analysis of the magnetic field magnitude data derived from the CHAMP 1 s fluxgate magnetometer measurements, available a few days before and after SUM1 and SUM2. The total field was computed from the three vector components. Since the measured magnetic field is dominated by the core magnetic field, a 32 s high-pass filter was applied to the total field. The purpose of this step of analysis is to emphasize short period signals and distinguish between artificial noise and natural source signals. To these filtered data we applied the continuous wavelet transform with the Morlet wavelet as the basis function (Torrence and Compo, 1998). Our results were checked for consistency using the Paul and DOG mother functions. We should also stress that there are several parameters of the wavelet transform (e.g., frequency range, power spectral density amplification factor, etc.) which need to be correctly adjusted in order to capture different kind of anomalous signals. This tuning of the wavelet transform is quite time consuming, but is an important step of our analysis. 3. Results In the following we present a detailed analysis of CHAMP data before and after the two events in the Sumatran region. The results are shown as successive figures, each of them containing four panels corresponding to the 1) temporal variation of the total magnetic field data after applying a 32 s high-pass filter; 2) the power spectral density of the above signal derived from the wavelet transform; 3) the electron density data, and 4) the latitudinal position of spacecraft over a track between 85° N and 85° S, except Figs. 3 and 5, which are between 45° N and 45° S, in order to better focus on the Sumatran region. 3.1. SUM1 The geomagnetic conditions represent a key parameter in interpreting the observed signals. It is well known
176
G. Balasis, M. Mandea / Tectonophysics 431 (2007) 173–195
Fig. 1. (A–I). SUM1. CHAMP satellite night (except Fig. 1E) passes from 18:00 to 06:00 UT. In Fig. 1F the time of the event is marked with a red cross. In this and next figures from top to bottom: The temporal variation of the CHAMP total magnetic field data after applying a 32 s high-pass filter; Corresponding wavelet power spectrum; The temporal variation of the CHAMP electron density data; The latitudinal position of the satellite with time. At the lowermost panel the area of the epicenter (±2°) is marked with red. The kp values corresponding to the time interval covered by Fig. 1(A–E) are 2+ and Fig. 1(F–I) are 3−.
G. Balasis, M. Mandea / Tectonophysics 431 (2007) 173–195
Fig. 1 (continued ).
177
178
G. Balasis, M. Mandea / Tectonophysics 431 (2007) 173–195
Fig. 1 (continued ).
G. Balasis, M. Mandea / Tectonophysics 431 (2007) 173–195
Fig. 1 (continued ).
179
180
G. Balasis, M. Mandea / Tectonophysics 431 (2007) 173–195
Fig. 1 (continued ).
that the maximum values for geomagnetic indices (e.g., kp) are obtained for periods characterized by active magnetic conditions. During these specific periods, contributions coming from outside the Earth, i.e., ionosphere and/or magnetosphere can contaminate variations originating from the Earth. Over the time of SUM1 the geomagnetic activity was low (kp was 3− and Dst was − 12 nT). In fact, kp was not more than 3− and Dst not less than − 25 nT even for the 12 h time interval around the event of main interest. 3.1.1. Electromagnetic disturbances before and after the event In Fig. 1(A–I) the 32 s high-pass filtered total magnetic field data, the corresponding wavelet power spectra and the ion densities are shown together for satellite passes covering the 12 h around the event. These are passes (except Fig. 1E) on the night side of the Earth, i.e., less influenced by the solar activity. Considering both the low activity geomagnetic conditions and the night-side data selection criteria, leads to a possible internal origin for the observed high frequency disturbances of the magnetic field, i.e., seismogenic in origin. The data available are continuous in time, with two exceptions (Fig. 1(E–F)), where the missing
intervals are clearly detected by wavelets as patches of high energy (dark red color). In Fig. 1(C–G) we observe the development of an ionospheric (F-region) instability in the EQ region. This anomaly is simultaneously detected in magnetic (1st panel) and ionospheric (3rd panel) time series, and gives a distinct broad frequency range signature in the wavelet power spectrum of the magnetic field (2nd panel). The key question is to find out if this instability is of either seismogenic or external origin. The answer is not obvious. On the one hand, Stolle et al. (2006) have studied the occurrence rate of CHAMP equatorial plasma bubble magnetic signatures between 2001 and 2004. They found a high occurrence rate of 80% in the Atlantic sector during the winter months, but the same region is totally void of events during summer. It is noted that there are large areas with very few equatorial plasma bubble signatures, such as the Indian sector with only some occurrences during autumn. Based on the statistical results from the Stolle et al. (2006) study, it is likely that the ionospheric anomalies found in CHAMP data around the time of the event are due to an equatorial plasma bubble. On the other hand, the fact that this instability is developing in the seismic area, could represent an excursion from the
G. Balasis, M. Mandea / Tectonophysics 431 (2007) 173–195
181
Fig. 2. (A–D). SUM1. CHAMP passes showing a Pc3 type (∼ 30 s) geomagnetic activity before and after the event. The kp values corresponding to the time interval covered by Fig. 2A is 2+ and Fig. 2(B–D) are 3−.
182
G. Balasis, M. Mandea / Tectonophysics 431 (2007) 173–195
Fig. 2 (continued ).
G. Balasis, M. Mandea / Tectonophysics 431 (2007) 173–195
183
Fig. 3. (A–F). SUM1. The closest CHAMP passes to the epicenter area on December 23, 24, 25 and 26, 2004. Fig. 3F zooms in Fig. 2C: a 16 s signal is detected accompanied by a bend in the electron density. In Fig. 3F and in the next figures the power spectral density is amplified by a factor of 8 in comparison to the rest of the figures shown in this paper. The kp values corresponding to each track are: 2+, 0o, 4o, 2−, 3o and 3−.
184
G. Balasis, M. Mandea / Tectonophysics 431 (2007) 173–195
Fig. 3 (continued ).
G. Balasis, M. Mandea / Tectonophysics 431 (2007) 173–195
Fig. 3 (continued ).
185
186
G. Balasis, M. Mandea / Tectonophysics 431 (2007) 173–195
Fig. 4. (A–G). SUM2. CHAMP satellite day passes from 11 to 21 UT. In Fig. 4D the time of the event (16:09:36.3 UT) is marked with a cross. At the lowermost panel the area of the epicenter (±2°) is also marked. The kp values corresponding to the time interval covered by Fig. 4(A–E) are 0+ and Fig. 4(F–G) are 1+.
G. Balasis, M. Mandea / Tectonophysics 431 (2007) 173–195
Fig. 4 (continued ).
187
188
G. Balasis, M. Mandea / Tectonophysics 431 (2007) 173–195
Fig. 4 (continued ).
G. Balasis, M. Mandea / Tectonophysics 431 (2007) 173–195
189
Fig. 4 (continued ).
general behavior shown by Stolle et al. (2006) and somehow be related to the changes in the region. 3.1.2. Pc3 and Pc5 type geomagnetic pulsations before and after the event In Fig. 2(A–D) pulsations with a period of ∼ 30 s are observed before but also just after the event, as Iyemori et al. (2005) also found. These passes are on the day side of the Earth and thus are more affected by solar wind activity. These perturbations characterized by a period of ∼ 30 s can be associated with Pc3 type pulsations. The fact that these pulsations are observed over all shown orbits suggests that the Pc3 pulsation reported by Iyemori et al. (2005) might have a global character, and might not be directly related to the SUM1 event. A Pc5 (∼ 3.6 min) geomagnetic pulsation was observed on December 26, 2004, just after SUM1, by Iyemori et al. (2005) in ground observatory magnetic data. CHAMP was flying, during this time interval, in a longitude almost 30° away (Fig. 2B) from the location on the Earth's surface where it was detected and no sign of these oscillations can be seen in the satellite magnetic data. To explain this pulsation Iyemori et al. (2005) have proposed an ionospheric dynamo mechanism in the ionospheric E-region at an altitude of between 100–
120 km over the epicenter. This mechanism can be generated by a vertical wind oscillation caused by the atmospheric duct resonance set up by the earthquake. At the commencement of the earthquake, a wide area at the epicenter was suddenly either lifted or depressed and as a result an atmospheric pressure variation propagated upward as an acoustic wave. Then part of the acoustic wave was reflected back to the lower thermosphere forming a duct resonance. The result is a vertical wind having a resonance frequency of 3.6 min in the ionospheric E region which generates the dynamo current. The Iyemori et al. (2005) results coupled with the fact that there is no signature in the CHAMP data, which can be related to Pc5, favors a local character for this anomaly, and a possible link between the Pc5 signal and the earthquake development. 3.1.3. The closest CHAMP passes over the epicenter area Fig. 3(A–E) shows the results corresponding to the closest CHAMP passes over the SUM1 area on December 23, 24, 25 and 26, 2004. In Fig. 3F a close up of Fig. 2C is given. Fig. 3F represents the second closest pass of the satellite over the epicenter area (within 10° longitude) on December 26, 2004, and it is
190
G. Balasis, M. Mandea / Tectonophysics 431 (2007) 173–195
Fig. 5. (A–F). SUM2. The closest CHAMP passes to the epicenter area on March 26, 27 and 28, 2005. The kp values corresponding to the time interval covered by Fig. 5(A–C) are 3+, 3o and 4−, respectively, and Fig. 5(D–F) are 1+.
G. Balasis, M. Mandea / Tectonophysics 431 (2007) 173–195
Fig. 5 (continued ).
191
192
G. Balasis, M. Mandea / Tectonophysics 431 (2007) 173–195
Fig. 5 (continued ).
G. Balasis, M. Mandea / Tectonophysics 431 (2007) 173–195
only 2 h after the event. In this plot an interesting feature (high power signal accompanied by a bend in the electron density diagram) can be observed immediately after 03:03 UT, with a period of about 16 s. This feature is well developed over some 700 km in the earthquake area. Since equatorial plasma bubbles are primarily confined on the night side of the Earth (Stolle et al., 2006) it is possible that this anomaly is not related to an ionospheric source and it might be a result of seismogenic processes. An explanation of this possible 16 s aftershock signal could be provided by Pulinets and Boyarchuk (2004): “… precipitation of energetic particles observed on satellites before strong earthquakes is caused by effective interaction of VLF noises trapped in irregularity ducts, which are created by anomalous electric field penetrating from the ground into the ionosphere and magnetosphere, with energetic particles of radiation belts …”. The physical processes involved are summarized in a 5 step procedure (Pulinets and Boyarchuk, 2004): (i) penetration of the electric field of seismic source into the plasmasphere; (ii) formation of a VLF duct in the plasmasphere; (iii) cyclotron–resonance interaction between waves and particles inside the duct in the radiation belt region; (iv) precipitation of energetic electrons into the lower atmosphere; (v) increase in electron density in the ionospheric D-region. 3.2. SUM2 The geomagnetic activity during this event was low (kp was 0+ and Dst was − 16 nT). It is noteworthy that the geomagnetic conditions were quiet (kp never exceeds 2− and Dst never falls below − 29 nT) for the whole day. 3.2.1. Electromagnetic disturbances before and after the event Fig. 4(A–G) shows magnetic data, its wavelet power spectrum and the corresponding electron density measurements a few hours around the event. These figures represent passes on the day side of the Earth, therefore, are more influenced by solar activity. In Fig. 4 (A–C) pulsations with a period of ∼ 30 s are clearly observed before the event. In Fig. 4(D–E) high power signal patches (after 15:54 UT and 17:24 UT, respectively) have developed over the area of the earthquake. 3.2.2. The closest CHAMP passes over the epicenter area The closest CHAMP passes over the epicenter area on March 26, 27, 28, 2005 are shown in Fig. 5(A–F).
193
The first three tracks contain no specific features which could be related to a seismic event. On March 28, 2004 two tracks are plotted: the one before the earthquake during the day side and the one after the earthquake during the night side. The day track is very much disturbed by external contributions. On the night side a single perturbation is observed, however, it is too far away from the area of interest to make any speculation as to its origin. 4. Discussion and conclusions In this study we analyzed CHAMP magnetic data using wavelets tools in order to determine if signals observed over the Sumatran region could be associated with the earthquakes of either December 26, 2004 or March, 28 2005. Our results can be summarized as follows: 1) Geomagnetic perturbations of Pc3 type pulsations were observed before and after the SUM1 earthquake, suggesting an external source for this signal. This result is different from the one published by Iyemori et al. (2005), where a Pc3 pulsation observed on December 26, 2004 is correlated with the great Sumatra earthquake. In order to better characterize these kinds of perturbations, resolve their origin, and determine if they are due to external sources in space or Earth fracture, ground observatory magnetic recordings should be combined with satellite magnetic measurements. 2) Comparing Fig. 1F with Fig. 4D, which are both passes at the time of SUM2 and SUM1, respectively, and Fig. 1G with Fig. 4E, which are the next ones, on the same side of the Earth, a common feature is to be underlined. Indeed, there is an interesting signature in the wavelet spectra in both cases, consisting in a gradual enhancement of short period fluctuations. More interesting is that for both earthquakes these specific signals occurred some 70–80 min after their respective events. 3) The closest CHAMP passes over the epicenter area show a clear feature with a period of about 16 s developed over 700 km in the SUM1 earthquake region. The same feature is not observed when CHAMP data corresponding to the SUM2 occurrence time are analyzed. Although the two events have been carefully investigated, it is difficult to connect the observed perturbations to earthquake mechanisms. The ionosphere itself is a very complicated system, highly dynamic, and mainly under the control of solar activity. Only a detailed
194
G. Balasis, M. Mandea / Tectonophysics 431 (2007) 173–195
statistical analysis using the results obtained from all the large magnitude earthquakes during the CHAMP satellite mission will be able to support these first findings. Finally, let us note DEMETER (Detection of ElectroMagnetic Emissions Transmitted from Earthquake Regions), a satellite mission which is devoted to the investigation of the ionospheric perturbations due to seismic activity, as well as to the global study of the Earth's electromagnetic environment (Parrot, 2002). This mission has already acquired a large amount of data, however for the time being only a few satellite observations related to seismo-electromagnetic events are available (Parrot et al., 2006). We should stress, however, the serious difficulties that exist in detecting such signatures in satellite data; despite the efforts of numerous scientists in the field no widely accepted theories have been reached or new methods developed with respect to this topic. Additionally, the foreseen Swarm mission (a constellation of three satellites, providing precise simultaneous measurements of the magnetic field over different regions of the Earth) could offer a new and unique opportunity to develop tools for distinguishing seismogenic emissions from non-seismic external electromagnetic signals. Acknowledgments Constructive remarks and suggestions from I.A. Daglis and A. De Santis, as well as helpful discussions with H. Lühr and R. Schachtschneider are gratefully acknowledged. G. Balasis acknowledges support from the Greek State Scholarship Foundation (IKY) and General Secretariat for Research and Technology project 210-c, in the frame of bilateral Science and Technology Cooperation between the Hellenic Republic and Czech Republic. References Alexandrescu, M., Gibert, D., Hulot, G., Le Mouel, J.-L., Saracco, G., 1995. Detection of geomagnetic jerks using wavelet analysis. J. Geophys. Res. 100, 12557–12572. Balasis, G., Egbert, G.D., 2006. Empirical orthogonal function analysis of magnetic observatory data: further evidence for nonaxisymmetric magnetospheric sources for satellite induction studies. Geophys. Res. Lett. 33 (11), L11311. Balasis, G., Egbert, G.D., Maus, S., 2004. Local time effects in satellite estimates of electromagnetic induction transfer functions. Geophys. Res. Lett. 31. doi:10.1029/2004GL020147. Balasis, G., Bedrosian, P.A., Eftaxias, K., 2005a. A magnetotelluric study of the sensitivity of an area to seismoelectric signals. Nat. Hazard Earth Syst. Sci. 5, 931–946. Balasis, G., Maus, S., Lühr, H., Rother, M., 2005b. Wavelet analysis of CHAMP fluxgate magnetometer data. In: Reigber, C., Lühr, H.,
Schwintzer, P., Wickert, J. (Eds.), Earth Observation with CHAMP. Springer, New York, pp. 347–352. Contoyiannis, Y., Kapiris, P., Eftaxias, K., 2005. Monitoring of a preseismic phase from its electromagnetic precursors. Phys. Rev., E Stat. Phys. Plasmas Fluids Relat. Interdiscip. Topics 71, 066123-1–066123-14. Eftaxias, K., Kapiris, P., Polygiannakis, J., Bogris, N., Kopanas, J., Antonopoulos, G., Peratzakis, A., Hadjicontis, V., 2001. Signatures of pending earthquake from electromagnetic anomalies. Geophys. Res. Lett. 28, 3321–3324. Eftaxias, K., Kapiris, P., Dologlou, E., Kopanas, J., Bogris, N., Antonopoulos, G., Peratzakis, A., Hadjicontis, V., 2002. Electromagnetic anomalies before the Kozani earthquake: a study of their behavior through laboratory experiments. Geophys. Res. Lett. 29, 69/1–69/4. Freund, F.T., Takeuchi, A., Lau, B.W.S., Post, R., Keefner, J., Mellon, J., Al-Manaseer, A., 2004. Stress-induced changes in the electrical conductivity of igneous rocks and the generation of ground currents. Terr. Atmos. Ocean. Sci. 1, 437–467. Gokhberg, M., Morgunov, V., Tomizawa, I., 1982. Experimental measurements of electromagnetic emissions possibly related to earthquake in Japan. J. Geophys. Res. 87 (B9), 7824–7828. Hayakawa, M., Molchanov, O., 2002. Seismo Electromagnetics. Terrapub, Tokyo. Iyemori, T., et al., 2005. Geomagnetic pulsations caused by the Sumatra earthquake on December 26, 2004. Geophys. Res. Lett. 32, L20807. doi:10.1029/2005GL024083. Kagan, Y.Y., Jackson, D.D., 1996. Statistical tests of VAN earthquake predictions: comments and reflections. Geophys. Res. Lett. 23, 1433–1436. doi:10.1029/95GL03786. Kaiser, G., 1994. A Friendly Guide to Wavelets. Birkhauser. Kapiris, P., Balasis, G., Kopanas, J., Antonopoulos, G., Peratzakis, A., Eftaxias, K., 2004a. Scaling similarities of multiple fracturing of solid materials. Nonlinear Process. Geophys. 11, 137–151. Kapiris, P., Eftaxias, K., Chelidze, T., 2004b. The electromagnetic signature of prefracture criticality in heterogeneous media. Phys. Rev. Lett. 92 (6) (065702/1–4). Kapiris, P., Polygiannakis, J., Li, X., Yao, X., Eftaxias, K., 2005. Similarities in precursory features in seismic shocks and epileptic seizures. Europhys. Lett. 69, 657–663. Kumar, P., Foufoula-Georgiou, E., 1997. Wavelet analysis for geophysical applications. Rev. Geophys. 35, 385–412. Liu, J.Y., Tsai, Y.B., Chen, S.W., Lee, C.P., Chen, Y.C., Yen, H.Y., Chang, W.Y., Liu, C., 2006. Giant ionospheric disturbances excited by the M9.3 Sumatra earthquake of 26 December 2004. Geophys. Res. Lett. 33, L02103. doi:10.1029/2005GL023963. Lühr, H., Maus, S., Rother, M., Cooke, D., 2002. First in-situ observation of night-time F region currents with the CHAMP satellite. Geophys. Res. Lett. 29. doi:10.1029/2001GL013845. Lühr, H., Maus, S., Rother, M., 2004. The noon-time equatorial electrojet, its spatial features as determined by the CHAMP satellite. J. Geophys. Res. 109, A01306. doi:10.1029/2002JA009656. Maus, S., Rother, M., Holme, R., Lühr, H., Olsen, N., Haak, V., 2002. First scalar magnetic anomaly map from CHAMP satellite data indicates weak lithospheric field. Geophys. Res. Lett. 29. doi:10.1029/2001GL013685. Maus, S., Rother, M., Hemant, K., Stolle, C., Lühr, H., Kuvshinov, A., Olsen, N., 2006. Earth's lithospheric magnetic field determined to spherical harmonic degree 90 from CHAMP satellite measurements. Geophys. J. Int. doi:10.1111/j.1365-246X.2005.02833.x. Mavromatou, C., Hadjicontis, V., 2001. Laboratory investigation of the electric signals preceding the fracture of crystalline insulators.
G. Balasis, M. Mandea / Tectonophysics 431 (2007) 173–195 In: Teisseyre, R., Majewski, E. (Eds.), Earthquake Thermodynamics and Phase Transformations in the Earth's Interior. Academic Press, pp. 501–517. Ogawa, T., Oike, K., Miura, T., 1985. Electromagnetic radiations from rocks. J. Geophys. Res. 90, 6245–6249. Olsen, N., Haagmans, R., Sabaka, T.J., Kuvshinov, A., Maus, S., Purucker, M.E., Rother, M., Lesur, V., Mandea, M., 2006. The swarm end-to-end mission simulator study: a demonstration of separating the various contributions to Earth's magnetic field using synthetic data. Earth Planets Space. 58 (4), 359–370. Parrot, M., 2002. The micro-satellite DEMETER. J. Geodyn. 33, 535–541. Parrot, M., Berthelier, J.J., Lebreton, J.P., Sauvaud, J.A., Santolik, O., Blecki, J., 2006. Examples of unusual ionospheric observations made by the DEMETER satellite over seismic regions. Phys. Chem. Earth. 31, 486–495. Pham, V.N., Boyer, D., Chouliaras, G., Le Mouel, J.L., Rossignol, J.C., Stavrakakis, G.N., 1998. Characteristics of electromagnetic noise in the Ioannina region (Greece); a possible origin for so called Seismic Electric Signal (SES). Geophys. Res. Lett. 25, 2229–2232. doi:10.1029/98GL01593. Pulinets, S.A., Boyarchuk, K.A., 2004. Ionospheric Precursors of Earthquakes. Springer, Berlin, Germany, p. 315. Purucker, M., Ishihara, T., 2005. Magnetic images of the Sumatra region crust. EOS Trans. 86 (10), 101–102. Röder, H., Braun, T., Schumann, W., Boschi, E., Buettner, R., Zimanowski, B., 2005. Great Sumatra earthquake registers on electrostatic sensor. EOS Trans. 86 (45), 445. Stolle, C., Lühr, H., Rother, M., Balasis, G., 2006. Magnetic signatures of equatorial spread F, as observed by the CHAMP satellite. J. Geophys. Res. 111, A02304. doi:10.1029/2005JA011184. Torrence, C., Compo, G.P., 1998. A practical guide to wavelet analysis. Bull. Am. Meteorol. Soc. 79, 61–78.
195
Trigunait, A., Parrot, M., Pulinets, S., Li, F., 2004. Variations of the ionospheric electron density during the Bhuj seismic event. Ann. Geophys. 22, 4123–4131. Tyler, R., Maus, S., Lühr, H., 2003. Satellite observations of magnetic fields due to ocean tidal flow. Science 299, 239–241. Tzanis, A., Vallianatos, F., Cruszow, S., 2000. Identification and discrimination of transient electrical earthquake precursors: fact, fiction and some possibilities. Phys. Earth Planet. Inter. 121, 223–248. Uyeda, S., Nagao, T., Orihara, Y., Yamaguchi, T., Takahashi, I., 2000. Geoelectric potential changes: possible precursors to earthquakes in Japan. PNAS 97, 4561–4566. Uyeda, S., Hayakawa, M., Nagao, T., Molchanov, O., Hattori, K., Orihara, Y., Gotoh, K., Akinaga, Y., Tanaka, H., 2002. Electric and magnetic phenomena observed before the volcano-seismic activity in 2000 in the Izu Island Region, Japan. PNAS 99, 7352–7355. doi:10.1073/pnas.072208499. Vallianatos, F., Nomikos, K., 1998. Seismogenic Radioemissions as Earthquake Precursors in Greece. Phys. Chem. Earth 23 (9), 953–957. Varotsos, P., Lazaridou, M., 1991. Latest aspects of earthquake prediction in Greece based on Seismic Electric Signals. Tectonophysics 188, 321–347. Varotsos, P., Alexopoulos, K., Lazaridou, M., 1993. Latest aspects of earthquake prediction in Greece based on Seismic Electric Signals, II. Tectonophysics 224, 1–37. Varotsos, P., Lazaridou, M., Eftaxias, K., Antonopoulos, G., Makris, J., Kopanas, J., 1996. Short term earthquake prediction in Greece by Seismic Electric Signals. In: Sir Lighthill, J. (Ed.), The Critical Review of VAN: Earthquake Prediction from Seismic Electric Signals. World Scientific Publishing Co., Singapore, pp. 29–76. Varotsos, P., 2005. The Physics of Seismic Electric Signals. Terra Sci. Pub, Tokyo, p. 338.
Can electromagnetic disturbances related to the recent great earthquakes be detected by satellite magnetometers? G. Balasis ⁎, M. Mandea GeoForschungsZentrum Potsdam, Telegrafenberg, 14473 Potsdam, Germany Received 13 January 2006; accepted 24 May 2006 Available online 8 December 2006
Abstract On December 26, 2004 the world's fourth largest earthquake since 1900 and the largest since the 1964 Prince William Sound, Alaska earthquake, occurred off the west coast of northern Sumatra with a magnitude of 9.3. On March 28, 2005 another event of magnitude 8.7 took place in the same region. The December 26, 2004 earthquake has prompted scientists to investigate possible electromagnetic signatures of this event, using ground magnetic observations. Iyemori et al. [Iyemori, T. et al., 2005. Geomagnetic pulsations caused by the Sumatra earthquake on December 26, 2004. Geophys. Res. Lett., 32, L20807, doi:10.1029/ 2005GL024083.] have suggested that a 3.6 min long geomagnetic pulsation, observed shortly after this event, was generated by the earthquake. They have speculated that a 30 s magnetic pulsation was also caused by the earthquake. Here for the first time, CHAMP satellite magnetic and electron density data are examined to find out if electromagnetic signatures which are possibly related to these recent megathrust earthquakes are observed in satellite magnetic data. We have shown that some specific features are observed after the two earthquakes, with periods of about 16 and 30 s. Our results favor an external source origin for the 30 s pulsation. Moreover, after more than 1 h, CHAMP magnetic data indicate the existence of a feature characterized by the same parameters (duration, amplitude, and frequency content), which could be associated with each earthquake, respectively. Further investigations are required in order to answer the question of whether these signals can be associated with earthquakes and to assign their possible usefulness with respect to earthquake development. © 2006 Elsevier B.V. All rights reserved. Keywords: Sumatra earthquake; CHAMP satellite; Geomagnetic field; Ionosphere; Wavelets
1. Introduction Earthquake physics involves a broad range of themes related to the Earth's system, from tectonic plates to volcanism, from microscopic processes to chemical re⁎ Corresponding author. Now at Institute for Space Applications and Remote Sensing, National Observatory of Athens, Metaxa and Vas. Pavlou, Penteli, 15236, Athens, Greece. E-mail addresses: [email protected], [email protected] (G. Balasis). 0040-1951/$ - see front matter © 2006 Elsevier B.V. All rights reserved. doi:10.1016/j.tecto.2006.05.038
actions, and from generation of electric charges to shockacoustic waves in total electron content. In this study, we are only interested in possible earthquake associated signatures, pre- or post-seismic, in satellite magnetic data. In the following we focus on electromagnetic disturbances possibly related to earthquakes. Over the last few decades, efforts have been made to improve the overall knowledge surrounding earthquakes, from physical mechanisms to pre- and postseismic signals. Recently, efforts at better understanding the physical properties of precursory ULF pre-seismic
174
G. Balasis, M. Mandea / Tectonophysics 431 (2007) 173–195
electric signals (SES) have been intensified (Varotsos and Lazaridou, 1991; Varotsos et al., 1993, 1996; Kagan and Jackson, 1996; Pham et al., 1998; Tzanis et al., 2000; Uyeda et al., 2000, 2002; Varotsos, 2005; Balasis et al., 2005a). On the other hand both acoustic as well as electromagnetic emissions in a wide frequency spectrum ranging from VLF to VHF are produced by microcracks, which can be considered as the so-called precursors of general fracture. These precursors are detectable on both a laboratory and a geological scale (Gokhberg et al., 1982; Ogawa et al., 1985; Vallianatos and Nomikos, 1998; Mavromatou and Hadjicontis, 2001; Freund et al., 2004). Several experimental results illustrating the connection between anomalous VLF–VHF electromagnetic and acoustic phenomena with earthquake preparation were presented in a rather comprehensive collection of papers in a book by Hayakawa and Molchanov (2002). Recently, kHz–MHz sequences of electromagnetic anomalies have been detected from a few days to a few hours prior to recent destructive earthquakes in Greece (Eftaxias et al., 2001, 2002). Kapiris et al. (2004a,b, 2005) and Contoyiannis et al. (2005) have advanced the hypothesis that the detected electromagnetic phenomena are probably precursors emitted from the focal area during micro-fracturing processes. Earthquakes can also produce atmospheric and ionospheric disturbances by dynamic coupling (Liu et al., 2006): vertical vibrations of the Earth's surface lead to pressure waves in the neutral atmosphere that grow in amplitude by several orders of magnitude as they attain ionospheric heights. The first papers concerning the ionospheric effects connected with earthquakes were published after the great Alaska earthquake in 1964 (see Pulinets and Boyarchuk (2004) and references therein). In principle, ionospheric perturbations after strong earthquakes occur just after the shock and are due to acoustic waves which are amplified through the atmosphere because of the decreasing atmospheric density with increasing height (Trigunait et al., 2004). Ionospheric perturbations can also be observed above a seismic zone a few days before an earthquake, but not all earthquakes produce such phenomena (Trigunait et al., 2004). In our study we focus on two great earthquakes, both related to the Sumatran region. The first, referred to hereinafter as SUM1, is the 9.3 magnitude earthquake of December 26, 2004, which occurred at 00:58:50.7 UT off the west coast of northern Sumatra, Indonesia (3.3° N, 95.9° E). The second (hereinafter SUM2) is the 8.7 magnitude earthquake of March 28, 2005 that occurred at 16:09:36.3 UT in the same region (2.074° N, 97.013° E). In this preliminary stage of our study we have chosen to concentrate only on these two events and to analyze them
carefully, considering that they might be generated by the same mechanisms. A few studies have been published related to possible electromagnetic signatures of SUM1. Röder et al. (2005) have shown that strong electrical signals, which correspond to SUM1, were recorded by an electrostatic sensor (this device works like a capacitor and detects short-term variations of the horizontal component of the Earth's electrostatic field) in Italy. These signals arrived at the station practically instantaneously with the occurrence of SUM1 and were detected up to several hours before the onset of SUM1. Iyemori et al. (2005) have argued that a geomagnetic pulsation with a period of approximately 3.6 min was observed at Phimai in Thailand, 12 min after the origin time of SUM1. At Tong Hai in China, some 10° north of Phimai, only a short period pulsation (about 30 s) was observed. At higher latitudes, no magnetic pulsations with these periods were observed. Liu et al. (2006) have reported two giant ionospheric disturbances at 01:19 and 04:10 UT on December 26, 2004 as observed by a network of digital Doppler sounders in Taiwan. The first disturbance, excited mainly by Rayleigh waves which consist of a packet of short-period Doppler shift variations, resulted in vertical ionospheric fluctuations with a displacement of about 200 m. The second disturbance is attributable to coupling of the atmospheric gravity waves excited by broad crustal uplift together with the following big tsunami waves around the earthquake source zone. The accompanying ionosonde data suggested that the gravity waves in the atmosphere may have caused the ionosphere to move up and down by about 40 km. 2. Data processing In this study we use CHAMP vector magnetic data to investigate if electromagnetic signatures could be related to the SUM1 and SUM2 events. We first describe the dataset used and then the wavelet analysis technique (Balasis et al., 2005b) applied which emphasizes high frequency signals. 2.1. Data CHAllenging Minisatellite Payload (CHAMP) is a small near polar, low altitude (∼ 400 km) satellite mission, launched in July 2000. With its highly precise magnetometer instruments, CHAMP has been generating high quality magnetic field measurements for more than 5 years. These data have been widely used to produce new models of the Earth's internal magnetic
G. Balasis, M. Mandea / Tectonophysics 431 (2007) 173–195
field (Maus et al., 2002, 2006; Olsen et al., 2006). CHAMP magnetic data have also given important hints for the geometry of the sources assumed in global electromagnetic induction studies which aim at deriving the electrical conductivity of the Earth's mantle (Balasis et al., 2004; Balasis and Egbert, 2006). The high accuracy of CHAMP magnetic measurements has led to the first detection of the magnetic field generated by the ocean tidal currents (Tyler et al., 2003). Moreover, CHAMP has offered valuable measurements of fields related to the systems of the Earth's ionospheric and magnetospheric currents (Lühr et al., 2002, 2004; Stolle et al., 2006). Recently, Purucker and Ishihara (2005) have used CHAMP data to produce magnetic images of the region surrounding SUM1 and SUM2 providing a current and historical view of subduction in the region. The CHAMP satellite provides 1 s measurements recorded by two different instruments: a three component (northern, eastern, radial) vector fluxgate magnetometer and an absolute scalar Overhauser magnetometer (the prime purpose of which is to calibrate the fluxgate measurements). Precise positioning and timing are associated with each of the measurements. Furthermore, plasma density measurements obtained by a planar Langmuir probe, onboard the CHAMP satellite, are considered. The planar Langmuir probe performs a sweep for determining the electron density every 15 s. 2.2. Analysis technique To analyze these highly accurate data, we have developed specific codes based on wavelet transforms. The wavelet transform is superior to the Fourier spectral analysis, providing excellent decompositions of even transient, non-stationary signals (Kaiser, 1994). It has the ability of providing a representation of the signal in both the time and frequency domains. In contrast to the Fourier transform, which provides the description of the overall regularity of signals, the wavelet transform identifies the temporal evolution of various frequencies. This property suits the signals under investigation, because they are not stationary by their nature, and have a time varying frequency content. Wavelet spectral analysis allows quantitative monitoring of the signal evolution by decomposing a time series into a linear superposition of predefined mathematical waveforms, each with finite duration and narrow frequency content (Kumar and Foufoula-Georgiou, 1997). Thus, the frequency range of the analyzing wavelets corresponds to the spectral content of time series components. Wavelet analysis is becoming a common tool for analyzing localized variations of power
175
within a time series. By decomposing a time series into time-frequency space, one is able to determine both the dominant modes of variability and how those modes vary in time (Alexandrescu et al., 1995). The advantage of analyzing a signal with wavelets as the analyzing kernel, is that it enables one to study features of the signal locally with a detail matched to their scale. Owing to its unique time-frequency localization, wavelet analysis is especially useful for signals that are non-stationary, have short-lived transient components, have features at different scales, or have singularities. We performed a wavelet analysis of the magnetic field magnitude data derived from the CHAMP 1 s fluxgate magnetometer measurements, available a few days before and after SUM1 and SUM2. The total field was computed from the three vector components. Since the measured magnetic field is dominated by the core magnetic field, a 32 s high-pass filter was applied to the total field. The purpose of this step of analysis is to emphasize short period signals and distinguish between artificial noise and natural source signals. To these filtered data we applied the continuous wavelet transform with the Morlet wavelet as the basis function (Torrence and Compo, 1998). Our results were checked for consistency using the Paul and DOG mother functions. We should also stress that there are several parameters of the wavelet transform (e.g., frequency range, power spectral density amplification factor, etc.) which need to be correctly adjusted in order to capture different kind of anomalous signals. This tuning of the wavelet transform is quite time consuming, but is an important step of our analysis. 3. Results In the following we present a detailed analysis of CHAMP data before and after the two events in the Sumatran region. The results are shown as successive figures, each of them containing four panels corresponding to the 1) temporal variation of the total magnetic field data after applying a 32 s high-pass filter; 2) the power spectral density of the above signal derived from the wavelet transform; 3) the electron density data, and 4) the latitudinal position of spacecraft over a track between 85° N and 85° S, except Figs. 3 and 5, which are between 45° N and 45° S, in order to better focus on the Sumatran region. 3.1. SUM1 The geomagnetic conditions represent a key parameter in interpreting the observed signals. It is well known
176
G. Balasis, M. Mandea / Tectonophysics 431 (2007) 173–195
Fig. 1. (A–I). SUM1. CHAMP satellite night (except Fig. 1E) passes from 18:00 to 06:00 UT. In Fig. 1F the time of the event is marked with a red cross. In this and next figures from top to bottom: The temporal variation of the CHAMP total magnetic field data after applying a 32 s high-pass filter; Corresponding wavelet power spectrum; The temporal variation of the CHAMP electron density data; The latitudinal position of the satellite with time. At the lowermost panel the area of the epicenter (±2°) is marked with red. The kp values corresponding to the time interval covered by Fig. 1(A–E) are 2+ and Fig. 1(F–I) are 3−.
G. Balasis, M. Mandea / Tectonophysics 431 (2007) 173–195
Fig. 1 (continued ).
177
178
G. Balasis, M. Mandea / Tectonophysics 431 (2007) 173–195
Fig. 1 (continued ).
G. Balasis, M. Mandea / Tectonophysics 431 (2007) 173–195
Fig. 1 (continued ).
179
180
G. Balasis, M. Mandea / Tectonophysics 431 (2007) 173–195
Fig. 1 (continued ).
that the maximum values for geomagnetic indices (e.g., kp) are obtained for periods characterized by active magnetic conditions. During these specific periods, contributions coming from outside the Earth, i.e., ionosphere and/or magnetosphere can contaminate variations originating from the Earth. Over the time of SUM1 the geomagnetic activity was low (kp was 3− and Dst was − 12 nT). In fact, kp was not more than 3− and Dst not less than − 25 nT even for the 12 h time interval around the event of main interest. 3.1.1. Electromagnetic disturbances before and after the event In Fig. 1(A–I) the 32 s high-pass filtered total magnetic field data, the corresponding wavelet power spectra and the ion densities are shown together for satellite passes covering the 12 h around the event. These are passes (except Fig. 1E) on the night side of the Earth, i.e., less influenced by the solar activity. Considering both the low activity geomagnetic conditions and the night-side data selection criteria, leads to a possible internal origin for the observed high frequency disturbances of the magnetic field, i.e., seismogenic in origin. The data available are continuous in time, with two exceptions (Fig. 1(E–F)), where the missing
intervals are clearly detected by wavelets as patches of high energy (dark red color). In Fig. 1(C–G) we observe the development of an ionospheric (F-region) instability in the EQ region. This anomaly is simultaneously detected in magnetic (1st panel) and ionospheric (3rd panel) time series, and gives a distinct broad frequency range signature in the wavelet power spectrum of the magnetic field (2nd panel). The key question is to find out if this instability is of either seismogenic or external origin. The answer is not obvious. On the one hand, Stolle et al. (2006) have studied the occurrence rate of CHAMP equatorial plasma bubble magnetic signatures between 2001 and 2004. They found a high occurrence rate of 80% in the Atlantic sector during the winter months, but the same region is totally void of events during summer. It is noted that there are large areas with very few equatorial plasma bubble signatures, such as the Indian sector with only some occurrences during autumn. Based on the statistical results from the Stolle et al. (2006) study, it is likely that the ionospheric anomalies found in CHAMP data around the time of the event are due to an equatorial plasma bubble. On the other hand, the fact that this instability is developing in the seismic area, could represent an excursion from the
G. Balasis, M. Mandea / Tectonophysics 431 (2007) 173–195
181
Fig. 2. (A–D). SUM1. CHAMP passes showing a Pc3 type (∼ 30 s) geomagnetic activity before and after the event. The kp values corresponding to the time interval covered by Fig. 2A is 2+ and Fig. 2(B–D) are 3−.
182
G. Balasis, M. Mandea / Tectonophysics 431 (2007) 173–195
Fig. 2 (continued ).
G. Balasis, M. Mandea / Tectonophysics 431 (2007) 173–195
183
Fig. 3. (A–F). SUM1. The closest CHAMP passes to the epicenter area on December 23, 24, 25 and 26, 2004. Fig. 3F zooms in Fig. 2C: a 16 s signal is detected accompanied by a bend in the electron density. In Fig. 3F and in the next figures the power spectral density is amplified by a factor of 8 in comparison to the rest of the figures shown in this paper. The kp values corresponding to each track are: 2+, 0o, 4o, 2−, 3o and 3−.
184
G. Balasis, M. Mandea / Tectonophysics 431 (2007) 173–195
Fig. 3 (continued ).
G. Balasis, M. Mandea / Tectonophysics 431 (2007) 173–195
Fig. 3 (continued ).
185
186
G. Balasis, M. Mandea / Tectonophysics 431 (2007) 173–195
Fig. 4. (A–G). SUM2. CHAMP satellite day passes from 11 to 21 UT. In Fig. 4D the time of the event (16:09:36.3 UT) is marked with a cross. At the lowermost panel the area of the epicenter (±2°) is also marked. The kp values corresponding to the time interval covered by Fig. 4(A–E) are 0+ and Fig. 4(F–G) are 1+.
G. Balasis, M. Mandea / Tectonophysics 431 (2007) 173–195
Fig. 4 (continued ).
187
188
G. Balasis, M. Mandea / Tectonophysics 431 (2007) 173–195
Fig. 4 (continued ).
G. Balasis, M. Mandea / Tectonophysics 431 (2007) 173–195
189
Fig. 4 (continued ).
general behavior shown by Stolle et al. (2006) and somehow be related to the changes in the region. 3.1.2. Pc3 and Pc5 type geomagnetic pulsations before and after the event In Fig. 2(A–D) pulsations with a period of ∼ 30 s are observed before but also just after the event, as Iyemori et al. (2005) also found. These passes are on the day side of the Earth and thus are more affected by solar wind activity. These perturbations characterized by a period of ∼ 30 s can be associated with Pc3 type pulsations. The fact that these pulsations are observed over all shown orbits suggests that the Pc3 pulsation reported by Iyemori et al. (2005) might have a global character, and might not be directly related to the SUM1 event. A Pc5 (∼ 3.6 min) geomagnetic pulsation was observed on December 26, 2004, just after SUM1, by Iyemori et al. (2005) in ground observatory magnetic data. CHAMP was flying, during this time interval, in a longitude almost 30° away (Fig. 2B) from the location on the Earth's surface where it was detected and no sign of these oscillations can be seen in the satellite magnetic data. To explain this pulsation Iyemori et al. (2005) have proposed an ionospheric dynamo mechanism in the ionospheric E-region at an altitude of between 100–
120 km over the epicenter. This mechanism can be generated by a vertical wind oscillation caused by the atmospheric duct resonance set up by the earthquake. At the commencement of the earthquake, a wide area at the epicenter was suddenly either lifted or depressed and as a result an atmospheric pressure variation propagated upward as an acoustic wave. Then part of the acoustic wave was reflected back to the lower thermosphere forming a duct resonance. The result is a vertical wind having a resonance frequency of 3.6 min in the ionospheric E region which generates the dynamo current. The Iyemori et al. (2005) results coupled with the fact that there is no signature in the CHAMP data, which can be related to Pc5, favors a local character for this anomaly, and a possible link between the Pc5 signal and the earthquake development. 3.1.3. The closest CHAMP passes over the epicenter area Fig. 3(A–E) shows the results corresponding to the closest CHAMP passes over the SUM1 area on December 23, 24, 25 and 26, 2004. In Fig. 3F a close up of Fig. 2C is given. Fig. 3F represents the second closest pass of the satellite over the epicenter area (within 10° longitude) on December 26, 2004, and it is
190
G. Balasis, M. Mandea / Tectonophysics 431 (2007) 173–195
Fig. 5. (A–F). SUM2. The closest CHAMP passes to the epicenter area on March 26, 27 and 28, 2005. The kp values corresponding to the time interval covered by Fig. 5(A–C) are 3+, 3o and 4−, respectively, and Fig. 5(D–F) are 1+.
G. Balasis, M. Mandea / Tectonophysics 431 (2007) 173–195
Fig. 5 (continued ).
191
192
G. Balasis, M. Mandea / Tectonophysics 431 (2007) 173–195
Fig. 5 (continued ).
G. Balasis, M. Mandea / Tectonophysics 431 (2007) 173–195
only 2 h after the event. In this plot an interesting feature (high power signal accompanied by a bend in the electron density diagram) can be observed immediately after 03:03 UT, with a period of about 16 s. This feature is well developed over some 700 km in the earthquake area. Since equatorial plasma bubbles are primarily confined on the night side of the Earth (Stolle et al., 2006) it is possible that this anomaly is not related to an ionospheric source and it might be a result of seismogenic processes. An explanation of this possible 16 s aftershock signal could be provided by Pulinets and Boyarchuk (2004): “… precipitation of energetic particles observed on satellites before strong earthquakes is caused by effective interaction of VLF noises trapped in irregularity ducts, which are created by anomalous electric field penetrating from the ground into the ionosphere and magnetosphere, with energetic particles of radiation belts …”. The physical processes involved are summarized in a 5 step procedure (Pulinets and Boyarchuk, 2004): (i) penetration of the electric field of seismic source into the plasmasphere; (ii) formation of a VLF duct in the plasmasphere; (iii) cyclotron–resonance interaction between waves and particles inside the duct in the radiation belt region; (iv) precipitation of energetic electrons into the lower atmosphere; (v) increase in electron density in the ionospheric D-region. 3.2. SUM2 The geomagnetic activity during this event was low (kp was 0+ and Dst was − 16 nT). It is noteworthy that the geomagnetic conditions were quiet (kp never exceeds 2− and Dst never falls below − 29 nT) for the whole day. 3.2.1. Electromagnetic disturbances before and after the event Fig. 4(A–G) shows magnetic data, its wavelet power spectrum and the corresponding electron density measurements a few hours around the event. These figures represent passes on the day side of the Earth, therefore, are more influenced by solar activity. In Fig. 4 (A–C) pulsations with a period of ∼ 30 s are clearly observed before the event. In Fig. 4(D–E) high power signal patches (after 15:54 UT and 17:24 UT, respectively) have developed over the area of the earthquake. 3.2.2. The closest CHAMP passes over the epicenter area The closest CHAMP passes over the epicenter area on March 26, 27, 28, 2005 are shown in Fig. 5(A–F).
193
The first three tracks contain no specific features which could be related to a seismic event. On March 28, 2004 two tracks are plotted: the one before the earthquake during the day side and the one after the earthquake during the night side. The day track is very much disturbed by external contributions. On the night side a single perturbation is observed, however, it is too far away from the area of interest to make any speculation as to its origin. 4. Discussion and conclusions In this study we analyzed CHAMP magnetic data using wavelets tools in order to determine if signals observed over the Sumatran region could be associated with the earthquakes of either December 26, 2004 or March, 28 2005. Our results can be summarized as follows: 1) Geomagnetic perturbations of Pc3 type pulsations were observed before and after the SUM1 earthquake, suggesting an external source for this signal. This result is different from the one published by Iyemori et al. (2005), where a Pc3 pulsation observed on December 26, 2004 is correlated with the great Sumatra earthquake. In order to better characterize these kinds of perturbations, resolve their origin, and determine if they are due to external sources in space or Earth fracture, ground observatory magnetic recordings should be combined with satellite magnetic measurements. 2) Comparing Fig. 1F with Fig. 4D, which are both passes at the time of SUM2 and SUM1, respectively, and Fig. 1G with Fig. 4E, which are the next ones, on the same side of the Earth, a common feature is to be underlined. Indeed, there is an interesting signature in the wavelet spectra in both cases, consisting in a gradual enhancement of short period fluctuations. More interesting is that for both earthquakes these specific signals occurred some 70–80 min after their respective events. 3) The closest CHAMP passes over the epicenter area show a clear feature with a period of about 16 s developed over 700 km in the SUM1 earthquake region. The same feature is not observed when CHAMP data corresponding to the SUM2 occurrence time are analyzed. Although the two events have been carefully investigated, it is difficult to connect the observed perturbations to earthquake mechanisms. The ionosphere itself is a very complicated system, highly dynamic, and mainly under the control of solar activity. Only a detailed
194
G. Balasis, M. Mandea / Tectonophysics 431 (2007) 173–195
statistical analysis using the results obtained from all the large magnitude earthquakes during the CHAMP satellite mission will be able to support these first findings. Finally, let us note DEMETER (Detection of ElectroMagnetic Emissions Transmitted from Earthquake Regions), a satellite mission which is devoted to the investigation of the ionospheric perturbations due to seismic activity, as well as to the global study of the Earth's electromagnetic environment (Parrot, 2002). This mission has already acquired a large amount of data, however for the time being only a few satellite observations related to seismo-electromagnetic events are available (Parrot et al., 2006). We should stress, however, the serious difficulties that exist in detecting such signatures in satellite data; despite the efforts of numerous scientists in the field no widely accepted theories have been reached or new methods developed with respect to this topic. Additionally, the foreseen Swarm mission (a constellation of three satellites, providing precise simultaneous measurements of the magnetic field over different regions of the Earth) could offer a new and unique opportunity to develop tools for distinguishing seismogenic emissions from non-seismic external electromagnetic signals. Acknowledgments Constructive remarks and suggestions from I.A. Daglis and A. De Santis, as well as helpful discussions with H. Lühr and R. Schachtschneider are gratefully acknowledged. G. Balasis acknowledges support from the Greek State Scholarship Foundation (IKY) and General Secretariat for Research and Technology project 210-c, in the frame of bilateral Science and Technology Cooperation between the Hellenic Republic and Czech Republic. References Alexandrescu, M., Gibert, D., Hulot, G., Le Mouel, J.-L., Saracco, G., 1995. Detection of geomagnetic jerks using wavelet analysis. J. Geophys. Res. 100, 12557–12572. Balasis, G., Egbert, G.D., 2006. Empirical orthogonal function analysis of magnetic observatory data: further evidence for nonaxisymmetric magnetospheric sources for satellite induction studies. Geophys. Res. Lett. 33 (11), L11311. Balasis, G., Egbert, G.D., Maus, S., 2004. Local time effects in satellite estimates of electromagnetic induction transfer functions. Geophys. Res. Lett. 31. doi:10.1029/2004GL020147. Balasis, G., Bedrosian, P.A., Eftaxias, K., 2005a. A magnetotelluric study of the sensitivity of an area to seismoelectric signals. Nat. Hazard Earth Syst. Sci. 5, 931–946. Balasis, G., Maus, S., Lühr, H., Rother, M., 2005b. Wavelet analysis of CHAMP fluxgate magnetometer data. In: Reigber, C., Lühr, H.,
Schwintzer, P., Wickert, J. (Eds.), Earth Observation with CHAMP. Springer, New York, pp. 347–352. Contoyiannis, Y., Kapiris, P., Eftaxias, K., 2005. Monitoring of a preseismic phase from its electromagnetic precursors. Phys. Rev., E Stat. Phys. Plasmas Fluids Relat. Interdiscip. Topics 71, 066123-1–066123-14. Eftaxias, K., Kapiris, P., Polygiannakis, J., Bogris, N., Kopanas, J., Antonopoulos, G., Peratzakis, A., Hadjicontis, V., 2001. Signatures of pending earthquake from electromagnetic anomalies. Geophys. Res. Lett. 28, 3321–3324. Eftaxias, K., Kapiris, P., Dologlou, E., Kopanas, J., Bogris, N., Antonopoulos, G., Peratzakis, A., Hadjicontis, V., 2002. Electromagnetic anomalies before the Kozani earthquake: a study of their behavior through laboratory experiments. Geophys. Res. Lett. 29, 69/1–69/4. Freund, F.T., Takeuchi, A., Lau, B.W.S., Post, R., Keefner, J., Mellon, J., Al-Manaseer, A., 2004. Stress-induced changes in the electrical conductivity of igneous rocks and the generation of ground currents. Terr. Atmos. Ocean. Sci. 1, 437–467. Gokhberg, M., Morgunov, V., Tomizawa, I., 1982. Experimental measurements of electromagnetic emissions possibly related to earthquake in Japan. J. Geophys. Res. 87 (B9), 7824–7828. Hayakawa, M., Molchanov, O., 2002. Seismo Electromagnetics. Terrapub, Tokyo. Iyemori, T., et al., 2005. Geomagnetic pulsations caused by the Sumatra earthquake on December 26, 2004. Geophys. Res. Lett. 32, L20807. doi:10.1029/2005GL024083. Kagan, Y.Y., Jackson, D.D., 1996. Statistical tests of VAN earthquake predictions: comments and reflections. Geophys. Res. Lett. 23, 1433–1436. doi:10.1029/95GL03786. Kaiser, G., 1994. A Friendly Guide to Wavelets. Birkhauser. Kapiris, P., Balasis, G., Kopanas, J., Antonopoulos, G., Peratzakis, A., Eftaxias, K., 2004a. Scaling similarities of multiple fracturing of solid materials. Nonlinear Process. Geophys. 11, 137–151. Kapiris, P., Eftaxias, K., Chelidze, T., 2004b. The electromagnetic signature of prefracture criticality in heterogeneous media. Phys. Rev. Lett. 92 (6) (065702/1–4). Kapiris, P., Polygiannakis, J., Li, X., Yao, X., Eftaxias, K., 2005. Similarities in precursory features in seismic shocks and epileptic seizures. Europhys. Lett. 69, 657–663. Kumar, P., Foufoula-Georgiou, E., 1997. Wavelet analysis for geophysical applications. Rev. Geophys. 35, 385–412. Liu, J.Y., Tsai, Y.B., Chen, S.W., Lee, C.P., Chen, Y.C., Yen, H.Y., Chang, W.Y., Liu, C., 2006. Giant ionospheric disturbances excited by the M9.3 Sumatra earthquake of 26 December 2004. Geophys. Res. Lett. 33, L02103. doi:10.1029/2005GL023963. Lühr, H., Maus, S., Rother, M., Cooke, D., 2002. First in-situ observation of night-time F region currents with the CHAMP satellite. Geophys. Res. Lett. 29. doi:10.1029/2001GL013845. Lühr, H., Maus, S., Rother, M., 2004. The noon-time equatorial electrojet, its spatial features as determined by the CHAMP satellite. J. Geophys. Res. 109, A01306. doi:10.1029/2002JA009656. Maus, S., Rother, M., Holme, R., Lühr, H., Olsen, N., Haak, V., 2002. First scalar magnetic anomaly map from CHAMP satellite data indicates weak lithospheric field. Geophys. Res. Lett. 29. doi:10.1029/2001GL013685. Maus, S., Rother, M., Hemant, K., Stolle, C., Lühr, H., Kuvshinov, A., Olsen, N., 2006. Earth's lithospheric magnetic field determined to spherical harmonic degree 90 from CHAMP satellite measurements. Geophys. J. Int. doi:10.1111/j.1365-246X.2005.02833.x. Mavromatou, C., Hadjicontis, V., 2001. Laboratory investigation of the electric signals preceding the fracture of crystalline insulators.
G. Balasis, M. Mandea / Tectonophysics 431 (2007) 173–195 In: Teisseyre, R., Majewski, E. (Eds.), Earthquake Thermodynamics and Phase Transformations in the Earth's Interior. Academic Press, pp. 501–517. Ogawa, T., Oike, K., Miura, T., 1985. Electromagnetic radiations from rocks. J. Geophys. Res. 90, 6245–6249. Olsen, N., Haagmans, R., Sabaka, T.J., Kuvshinov, A., Maus, S., Purucker, M.E., Rother, M., Lesur, V., Mandea, M., 2006. The swarm end-to-end mission simulator study: a demonstration of separating the various contributions to Earth's magnetic field using synthetic data. Earth Planets Space. 58 (4), 359–370. Parrot, M., 2002. The micro-satellite DEMETER. J. Geodyn. 33, 535–541. Parrot, M., Berthelier, J.J., Lebreton, J.P., Sauvaud, J.A., Santolik, O., Blecki, J., 2006. Examples of unusual ionospheric observations made by the DEMETER satellite over seismic regions. Phys. Chem. Earth. 31, 486–495. Pham, V.N., Boyer, D., Chouliaras, G., Le Mouel, J.L., Rossignol, J.C., Stavrakakis, G.N., 1998. Characteristics of electromagnetic noise in the Ioannina region (Greece); a possible origin for so called Seismic Electric Signal (SES). Geophys. Res. Lett. 25, 2229–2232. doi:10.1029/98GL01593. Pulinets, S.A., Boyarchuk, K.A., 2004. Ionospheric Precursors of Earthquakes. Springer, Berlin, Germany, p. 315. Purucker, M., Ishihara, T., 2005. Magnetic images of the Sumatra region crust. EOS Trans. 86 (10), 101–102. Röder, H., Braun, T., Schumann, W., Boschi, E., Buettner, R., Zimanowski, B., 2005. Great Sumatra earthquake registers on electrostatic sensor. EOS Trans. 86 (45), 445. Stolle, C., Lühr, H., Rother, M., Balasis, G., 2006. Magnetic signatures of equatorial spread F, as observed by the CHAMP satellite. J. Geophys. Res. 111, A02304. doi:10.1029/2005JA011184. Torrence, C., Compo, G.P., 1998. A practical guide to wavelet analysis. Bull. Am. Meteorol. Soc. 79, 61–78.
195
Trigunait, A., Parrot, M., Pulinets, S., Li, F., 2004. Variations of the ionospheric electron density during the Bhuj seismic event. Ann. Geophys. 22, 4123–4131. Tyler, R., Maus, S., Lühr, H., 2003. Satellite observations of magnetic fields due to ocean tidal flow. Science 299, 239–241. Tzanis, A., Vallianatos, F., Cruszow, S., 2000. Identification and discrimination of transient electrical earthquake precursors: fact, fiction and some possibilities. Phys. Earth Planet. Inter. 121, 223–248. Uyeda, S., Nagao, T., Orihara, Y., Yamaguchi, T., Takahashi, I., 2000. Geoelectric potential changes: possible precursors to earthquakes in Japan. PNAS 97, 4561–4566. Uyeda, S., Hayakawa, M., Nagao, T., Molchanov, O., Hattori, K., Orihara, Y., Gotoh, K., Akinaga, Y., Tanaka, H., 2002. Electric and magnetic phenomena observed before the volcano-seismic activity in 2000 in the Izu Island Region, Japan. PNAS 99, 7352–7355. doi:10.1073/pnas.072208499. Vallianatos, F., Nomikos, K., 1998. Seismogenic Radioemissions as Earthquake Precursors in Greece. Phys. Chem. Earth 23 (9), 953–957. Varotsos, P., Lazaridou, M., 1991. Latest aspects of earthquake prediction in Greece based on Seismic Electric Signals. Tectonophysics 188, 321–347. Varotsos, P., Alexopoulos, K., Lazaridou, M., 1993. Latest aspects of earthquake prediction in Greece based on Seismic Electric Signals, II. Tectonophysics 224, 1–37. Varotsos, P., Lazaridou, M., Eftaxias, K., Antonopoulos, G., Makris, J., Kopanas, J., 1996. Short term earthquake prediction in Greece by Seismic Electric Signals. In: Sir Lighthill, J. (Ed.), The Critical Review of VAN: Earthquake Prediction from Seismic Electric Signals. World Scientific Publishing Co., Singapore, pp. 29–76. Varotsos, P., 2005. The Physics of Seismic Electric Signals. Terra Sci. Pub, Tokyo, p. 338.