Work Package 2 - Theoretical and experimental studies of plasma waves in the vicinity of earthquake epicentres


The main objectives of WP2 were:

  1. To study the electric and magnetic fields and their fluctuations associated with the precursors to earthquakes.
  2. To develop physical models that explain the associated seismo-electromagnetic turbulence identified in satellite data.


The spatio-temporal variations of EM signals on board the satellites are greatly influenced by the dynamical processes in the Earth’s magnetosphere. With the use of data collected from a single low orbiting satellite it is rather questionable to identify the cause of the observed anomalies. In the present project for supplementary cross-validation it is planned to use the data from high altitude magnetospheric satellites such as Cluster and THEMIS. The project participants have developed the effective methods of the satellite data processing that enable the identification of not only the propagation direction of observed waves, but also their plasma modes and nonlinear processes that affect their evolution and other characteristics of observed plasma turbulence (Balikhin et al., 2003; Walker et al., 2004; Hobara et al., 2007; Zhu et al., 2008). These techniques can be used for the cross-validation described above. These methods have been successfully tested for processes related to space weather problems and displayed their high effectiveness.



Task 2.1 – A survey of electromagnetic turbulence observed by DEMETER as it passes in the vicinity of earthquake epicentres

This task involved an investigation of the plasma wave environment of the ionosphere as measured by the DEMETER satellite as it passes over a seismically active region. Due to limitations on the availability of data it was decided to concentrate on the ULF electric field data set because it is available for the whole orbit where as the other data sets are typically only available for a limited fraction of each orbit (ie when the satellite operates in burst mode) or a limited set of parameters (eg spectra of one component of the electric field).

Task 2.1 investigated the plasma wave environment of the ionosphere in the vicinity of earthquake epicentres.

The results shown here were measured in the vicinity of the Sichuan earthquake epicentre and concentrate on the period of around 20 days preceding the earthquake. The first undertaking was to examine how the wave power varied within each half orbit when DEMETER passed within 2000km of either the earthquake epicentre or the point magnetically conjugate to it. The ULF waveform data were transformed into the frequency domain using an FFT and the resulting spectra were integrated to calculate the wave power in 10 frequency bands. The powers were then averaged over a period of 30 seconds in order to smooth the signal.

Figure 2.1 shows the ratio of the localised wave power to the average wave power for the orbit. Only data for night time orbits are plotted since the daytime orbits do not show any significant wave power in the ULF band. The X axis shows the number of the 30 second period within each orbit (there are typically around 70 such periods). The Y axis represents the satellite orbit, with 45 being the closest orbit to the time of the earthquake and zero referring to the orbit measured about 20 days prior to the earthquake and the colour indicates the ratio of the wave powers. The stars mark the times when DEMETER was closest to the earthquake epicentre. The intense signals corresponding to the time at the start (X~1) and end (X~70) of each orbit represent edge effect, and may be related to auroral activity.

Changes in wave power spectra

Figure 2.1: Changes in the wave power measured by DEMETER in the vicinity of the Sichuan earthquake.

From Figure 2.1 it can be seen that the wave power generally increases in the period just before the earthquake occurs in the vicinity of the epicentre. The other significant change in the wave power shown in this plot occurs over a similar time period but at a position that corresponds to DEMETER passing in the vicinity of the position magnetically conjugate to the epicentre. This increase could be due to waves propagating along the magnetic field lines from one hemisphere to the other. When comparing similar plots at different frequencies it is most noticeable that this onset in the wave power in most dominant at lower frequencies.

Example netpoint model

Figure 2.2: Variation of wave power with distance from the epicentre.

This increase in the wave power in the vicinity of the epicentre is also evident in Figure 2.2 which shows the mean (X-axis) wave power versus its variance during the period 2-3 days before the main earthquake. The symbols show the ratio of the wave power at distances 0-500km (blue), 500-1000km (red) and 1000-1500km (green) from the epicentre when compared to the wave power at distances 1500-2000km. These plots clearly show that as the spacecraft approaches the location of the epicentre the magnitude of the ULF fluctuations increases. This effect is observed in both the daytime orbits (circles) and night time (dots). However, for this effect to be apparent, DEMETER should be operating in burst mode rather than summary mode.

Figure 2.3: Variation of wave power at frequencies f<0.5Hz.

Finally, an examination of the spectra for these data periods reveals that bursts of wave noise occur at low frequencies, typically f<0.5Hz. Figure 2.3 shows the wave power measured below 0.5Hz as a function of time averaged for a 30 second period centred on the time when DEMETER is geographically closest to the epicentre (blue), closest to the conjugate point (red), or closest to the north (green) and south (magenta) conjugate points at the satellite altitude. This plot indicates that for a number of days before the onset of the earthquake there is an increase in the wave power below 0.5Hz.

Task 2.2 - Multi-satellite data analysis of Cluster, THEMIS, and DEMETER data to distinguish time periods when electromagnetic turbulence observed by the low orbit DEMETER satellite are the result of space weather related events in the magnetosphere.

In Task 2.2 other possible sources of waves were investigated to eliminate other potential sources of wave activity.

Task 2.1 showed that the DEMETER spacecraft measured an increase in the level of plasma waves when its orbit took it over a seismically active region. This increase may be due to sources other than seismic activity. Plasma waves in the inner magnetosphere may result from effects such as magnetic storms and substorms. Therefore it is important to eliminate these as sources for the increase in wave activity discussed in Task 2.1. The state of the magnetosphere maybe quantified by a set of geomagnetic indices, in particular the Dst index. Figure 2.4 shows the variation of the Dst index in the period before the Sichuan earthquake (May 12th, 2008). It clearly shows that there are no large decrease (of the order 100nT) which is indicative of the occurrence of geomagnetic activity caused by geomagnetic storms and substorms.

Example netpoint model

Figure 2.4: Variation of the geomagnetic Dst index in the run up to the Sichuan earthquake (May 12th, 2008).

This observation was also borne out in the electric field observations by the satellites of the Cluster and THEMIS missions as they passed through the inner magnetosphere. The average level of the electric field within the inner magnetosphere did not show a steady increase in the period leading up to the Sichuan earthquake and so it appears that the wave energy observed by DEMETER does not correspond to waves observed further out in the magnetosphere.

The origin of these waves appears to be terrestrial in nature. Two types of potential waves to carry energy from the EarthÕs surface to the ionosphere have been discussed in literature, namely the Internal Gravity Wave (IGW) and the Acoustic Wave. IGW are discussed elsewhere within the project. In this task, the occurrence of Acoustic Waves was investigated. Acoustic waves may be triggered by man-made sources (e.g. explosions, rocket launches etc.) or natural sources (e.g. seismic or volcanic activity, hurricanes, etc.). There have been a number of observations of increases in EM wave activity at ULF, ELF, and VLF frequencies associated with nuclear explosions. The acoustic waves generated propagate into the ionosphere where they cause disturbances in the conducting plasma causing current instabilities that lead to the generation of plasma waves. The EM waves propagate along the terrestrial magnetic field lines into the conjugate hemisphere where they have also been observed by orbiting satellites. Waves associated with acoustic sources tend to have a short lifetime, appearing as a noisy spot in the ionosphere in the vicinity of the epicentre.

Task 2.3 - Experimental results from Tasks 2.1 and 2.2 will be used to assess competing theories for electromagnetic earthquake precursors.

Tasks 2.1 and 2.2 investigated the low frequency plasma wave environment within the ionosphere. It was demonstrated that during the precursory period leading up to the Sichuan earthquake there was an increase in the low frequency plasma wave activity as the DEMETER passed over the seismically active area. During this period, the magnetosphere was quiet with the occurrence of no geomagnetic storms and substorms. Cluster and THEMIS observations showed that the level of low frequency wave activity in the night-time inner magnetosphere showed no general increases during this period. Therefore it was concluded that the increase in wave activity was due to terrestrial sources including seismic activity. Based on these results and others obtained within the SEMEP project this task reports a scenario that connects the production of Internal Gravity Waves due to seismic activity, their propagation into the upper atmosphere where their effects are observed in the propagation of VLF transmitter signals and the increase in ULF wave activity at the altitude of DEMETER.

Task 2.3 put together observations from this work package and others to suggest a scenario to explain the observation of ULF waves in the vicinity of a seismically active zone. The source of these waves is the continual deformation of the crustal layers of the Earth within the seismically active zone. This deformation results in a number of processes that are capable of launching either Internal Gravity Waves (IGW) or Acoustic Waves such as

We hypothesize that, prior to an earthquake the dynamics in the atmosphere develops according to the following scheme. The large scale thermal inhomogeneity regions above the seismically active regions appear within a few days before strong earthquakes. Thermosphere with decreasing temperature with altitude is unstable stratified if the vertical temperature gradient exceeds that adiabatic lapse rate. Investigating the vortex structures in the Earth’s atmosphere with finite vertical shear of zonal wind, we obtain a more realistic estimate for the velocity of vortex structures. We show that the vortex velocity in the coordinate system moving with the zonal wind can be substantially smaller than the velocity of sound. Owing to collective vortex motions in the atmosphere and related motion in the ionosphere, the density of plasma and atomic oxygen increase. This agrees with DEMETER observations of ionospheric density variations recorded before the 2010 M=8.8 earthquake in Chile. We believe that our research can help to advance in understanding of the interpretation of structures observed before the earthquake in a few days.

At the interface of the lower ionosphere/atmosphere these perturbations affect the propagation characteristics of terrestrial transmitter signals if their path passes over a seismically active region. Higher in the ionosphere, low orbiting satellites such as DEMETER observe irregularities in particle density and an increased plasma wave activity, the frequency of which is similar to one observed from the spectral analysis of terrestrial transmitter signals. Once in the ionosphere, the IGW can drive a myriad of instabilities resulting in their dissipation as energy is transferred to smaller scales, causing wave activity at higher ELF, VLF, and HF frequencies as have often been observed.



D2.1 A statistical study of plasma waves observed by the DEMETER satellite.
D2.2 A model for the low-frequency seismo-electromagnetic turbulence.
D2.3 Elaboration of a self consistent physical scenario for the earthquake.