Figure 1: Earthquake epicenter and Dobrovolsky area.

The mainshock consisted of two subevents and was characterized by supershear rupture. The aftershock activity was law and confined to five clusters, which cover a zone too long for a Mw 6.9 earthquake (about120 km of fault rupture length derived from the aftershock distribution) while according to Wells and Coppersmith the rupture length is between 35 and 60 km.






  1. Accelerated Moment Release (AMR) & Revised-AMR

In terms of AMR/R-AMR methods for the analysis and characterization of the regional seismicity from Catalogue, we showed that a slight improvement came from the application of R-AMR instead of AMR, but acceleration was not as evident as in the previous case of Chile 2014. Figure 2 shows results from R-AMR in which an acceleration in seismicity is visible.




Figure 2: Results from R-AMR analysis (Greece).





Two different approaches were developed to search EQ precursor using geomagnetic Swarm data: MASS method and Wavelet method. Both analyses are based on Level 1B MAGxLR Swarm product.


  1. MASS algorithm (Magnetic Swarm anomaly detection by Spline analysis)

The Greece EQ has a lower magnitude (M6.9,lower than M7.0) than the Chile ones (M8.2). This is reflected by the smaller Dobrovolsky area.

For North Aegean Sea 2014 earthquake the algorithm MASS (Magnetic Swarm anomaly detection by Spline analysis) was applied as an automatic research of magnetic anomalies with 3 different thresholds (kt= 2, 2.2 and 2.5). The tracks are marked as “anomaly” only if the centre of the moving window is in DbA and if geomagnetic conditions are quiet (|Dst|≤20 nT and ap≤10 nT).

Figure 3 shows an example of anomalous track.


Figure 3: Example of anomalous track detected with MASS algorithm (Sat_A, 13 May 2014, Greece).



Figure 4: Cumulative number of anomaly detected by MASS one month before and one month after M6.9 24 May 2014 Northern Aegean Sea EQ. Threshold is kt = 2.5 and the anomalies are selected only with geomagnetic quiet condition(|Dst| ≤ 20 nT and ap ≤ 10 nT).

Figure 4 shows the cumulative number of anomaly tracks detected by MASS. The number of anomalies found after the EQ is bigger than those detected before EQ (nEQ>1). In addition, the number of anomalies identified by Bravo is greater than the average number of anomalies revealed by Alpha and Charlie. Both these parameters together with R2 ≥0.97 in Y component, provide an indication that most of the revealed anomalies may have not a relation with the LAIC, rather to an external source to the Earth (for example coupling with the magnetosphere).




  1. Wavelet Analysis

The Wavelet spectral analysis has evidenced the existence (in general) of anomalous families, each characterized by some features. High frequency large impulsive anomalies in the first case emerge from background at night-time even in quiet magnetic conditions (Figure 5). It is not easy to recognize the lithospheric origin. In fact, Figure 28 (right) shows a couple of such pulses: one inside DbA just over the epicentre and the other outside. Trains of these pulses appear in Figure 28 (left) when the satellite flies over a region external to the DbA.





Figure 5: Example of output figure for Wavelet analysis, Aegean Sea EQ (Greece).

Analyzing this event, it can be concluded that discrimination of emerging signals from the background by applying this method alone is very challenging and that it may be necessary the adoption of some other more complex scheme of strategy.



  1. SWARM – IONOSPHERE from Satellite

The Swarm data related to the study of ionospheric background were used to tune the search for earthquake-related anomalies. Satellite-based data for the ionospheric characterization of the EQ events are mainly those referred to the LP (Langmuir Probe) instrument aboard the SWARM satellites, which determine local properties of the plasma, such as temperature and density by measuring the collected current due to electrons and ions. The electron density Ne is the relevant parameter used for the ionospheric characterization of the EQ events. Two different methods were developed to analyze Swarm ionospheric data: NeSTAD and NeLOG.


  1. Method I: NeSTAD

The NeSTAD analysis has been applied to the Swarm constellation data as described in D3203-Updating Chapter 3. In particular, LP and IBI data available in the period from 24 April to 23 June have been used to derive the track anomaly parameters.

Then, to tag the interesting track anomalies for this particular event, the following criteria have been applied:

  • R>Rthr=0.7 and standard deviation of the filtered track below σthr=0.1 or standard deviation of the filtered track above σthr=0.1 independently of R value.
  • Only night time tracks have been selected (18 to 06 LT), because during night time and at mid latitudes the ionosphere is expected to be less turbulent.
  • Quiet ionospheric conditions (absolute value of Dst in the considered day not exceeding 20 nT).

An example of tagged anomaly is provided in Figure 6.


Figure 6: Identified anomaly with the NeSTAD algorithm and tagged as interesting for the Aegean Sea M6.9 EQ event. Tagged anomaly refers to Swarm Alpha satellite on 14 march 2014.


Figure 7: - Cumulative number of anomalies identified through the NeSTAD algorithm for Swarm satellite Alpha.


Figure 7 shows the cumulative number of the tagged anomalies through the NeSTAD algorithm for Alpha satellite. The black dashed line indicates the day in which the Aegean Sea M6.9 EQ event occurred. The red boxes indicate the days in which disturbed geomagnetic conditions have been recorded and that are not included in the tagging process. The two blue dashed lines indicate the time interval during which the given Swarm satellite covered the DbA of the event at night time. It is interesting to note that only one anomaly has been tagged before the EQ event (Alpha: 29 April 2014), while the most of anomalies occurred after the EQ event occurrence.


  1. Method II: NeLOG

The automatic search NeLOG of ionosphere electron density anomaly from Swarm data is applied to this case study. A sample is classified as anomaly if the residual value exceeds a threshold kt times the RMS of the residual after the spline fit. A track is selected as an “anomaly” if it has more than 10 anomaly samples in DbA and if geomagnetic indices are |Dst|<20 nT and ap<10 nT. Tracks are selected within a mean local time between 22 and 6.

The NeLOG algorithm detects some anomalies 10 days before EQ. The anticipation time is compatible with those given in the literature for the electron density anomalies that may precede earthquakes. In particular the anomalies are characterized by an increase or decrease of Ne.

Figure 8 shows an example of anomaly track detected by NeLOG method ten days before the North Aegean Sea EQ (Sat_A, M6.9, 24 May 2014).



Figure 8: Anomaly detected by NeLOG in satellite Alpha 10 days before the M6.9 24 May 2014 Northern Aegean Sea EQ.


  1. IONOSPHERE Ground-based

i) Ionosondes

Ionosphere ground-based analysis was performed for the Aegean Sea EQ using ionosonde data from Athens Ionospheric Observatory (Latitude: 38.0° N, Longitude: 23.5° E -Penteli site, Athens, Greece). The ionospheric anomaly found before the 24/05/2014 earthquake is shown below. A single anomaly has been identified, occurring on 11 February 2014 at 07 UT. Figure 9 shows that during the first part of the day, when the ionospheric anomaly occurred, there were quiet conditions making the anomaly likely to be not related to geomagnetic or auroral activity.


Figure 9: The ionospheric anomaly of the 11.02.14 using observed Δh’Es, δfbEs, and δfoF2 variations and 3-hour ap. It can be considered a significant ionospheric anomaly prior to the 24/05/2014 earthquake.


ii) GNSS

Ground based GNSS data have been analysed by two different approaches: the “multi-station GNSS analysis” and the “single-station GNSS analysis”.

The multi-station approach makes use of 16 IGS stations to highlight the geographical distribution of possible anomalous structures related to pre-EQ effects on the ionosphere. Considering the data acquired in the period from 24 April to 24 May 2014, the geographical distribution of the anomalies seems to show a minimum in the area surrounding the epicenter (see Figure 10).


Figure 10: Geographic distribution of TEC anomalies detected from 24 April to 24 May 2014 over an area including the DbA.


Single station approach has been applied to one of the closest GNSS station to the epicenter, located in Dionysos (dyng, 38.04 N,23.56 E), considering data acquired from 17 May to 24 May 2014 (one week before the EQ). The spectrogram of the oscillation component extracted from vTEC signal (see Figure 119) highlights the presence of a large power corresponding to 0.03 mHz (about 9 hours period) during both 18 May and 22 May (6 days and 2 days before the EQ respectively).


Figure 11: Continuous wavelet spectrum of the oscillation component of vTec over dyng station (Aegean Sea EQ).



R-AMR and AMR techniques were applied to the study of Greece EQ but seems they did not work as well as in 2014 Chile EQ. Possible explanation for this could be going back to the following reasons:

- the region (or the fault system) is not equivalent to a critical system

- the EQ epicenter is surrounded by sea

NeLOG algorithm detects anomalies within 10 days from the EQ date and in the same period also MASS detects magnetic anomalies. In particular, the magnetic anomalies detected by MASS are distributed along the tracks while electron density anomalies are more concentrated. The anticipation time is compatible with values found in literature for electron density anomalies preceding earthquakes, so these events are very good candidate for LAIC effect.

The tagging procedure on NeSTAD output allows to identify only one anomaly for satellite before the EQ, but several anomalies are also immediately after the EQ.