Earthquake's Regional Forecasting Using Reliable Geomagnetic Precursor

INRNE, BAS, Sofia, Bulgaria

Everyday Monitoring

Geomagnetic quakes as Earthquake Precursor

Every day Sofia (one component) and Skopje, Kiev (vector variometer mode) geomagnetic and earthquake monitoring. The analysis of Kiev data for estimation of geomagnetic quake (the value of Sig jump) need more statistic. Explanation for the next Sofia, Skopje, Kiev, Lvov,Pag and Dusheti, Georgia (DSH_Ge) grafics: The earthquake precursor (geomagnetic quake) is SigD 10% jump. The approximate time window for incoming event is defined from the next Earth Tide extrem (Dennis Milbert source ccde). In minimum the time period is approximately ±1 day and in maximum ±2 days.
EQ data from:
Reliability time control from:
SChtM>2e4, Mag>3, Distance<800 km. The index of SChtM is the time difference between local Tide extremum and occred Eq time. The index of Magnitude is the distance in hundred km between Gm device and eq's epicenter.

p.s.Nov 09, 2011 In our (me and Lazo Pekevski) visit in Georgia in the framework of EU FP7 IRSES 2011 Project "Complex research of earthquake'e forecasting possibilities,seismicity and Climate chang correlations"
the Dusheti, Georgia flux gate magnetometer data were included in our every day monitoring.

p.s.Aug 01, 2013 In our (me and Lazo Pekevski) visit in Grocka, Serbia in July in collaboration with Milena Cukavac and Dr. Spomenko Mihailovich the Grocka flux gate magnetometer data were included in our every day monitoring.

Last visit

Amazing and shiny stats
Blog Counters

December 21, 2014

Important note: Please see the explanations for the developing the "Geomagnetic approach" in the time of EU IRSES BlackSeaHazNet (2011-2014) project: BlackSeaHazNet Scientific Report - EU FP7 IRSES project 2011-2014

Last udate:

08 Apr 2008 SofiaBg          

06 Jan 2014 Skopje           Skopje, Macedonia

20 Jul 2017 Kiev           Kiev, Ukraine

06 Jan 2014 Lvov           Lvov, Ukraine

12 May 2018 Pag          Panagurichte, Bg

21 Sep 2012 DSHGe      Dusheti, Georgia

12 May 2018 GckSrb     Grocka, Belgrade, Serbia

12 May 2018 SuaRo     Surlari, Romania

06 Feb 2014 LAquila     LAquila, Italy

21 Mar 2015 ArmeniaGeomagnetic     Stepanavan, Armenia

06 Feb 2014 ArmeniaBoreHole     Noemberyan borehole, Armenia

24 Aug 2017 KAK          Kakioka, Japan

Real time EMSC-CSEM reliability test,  


It is useful to stress that the author’s interests to the earthquake’s prediction problem arise as a result of complex research of the Black Sea ecosystem about 15 years ago (Mavrodiev, 1998). During the time of gathering the historical data for the ecosystem was observed, that the Crime earthquake in 1928 occurred, as an evidence for electromagnetic and earthquake correlations. Such hypothesis has been proposed by the academician Popov in the early 20- 30-ties of the last century (private communication).

According the INTERMAGNET requirements for measuring the geomagnetic field (see Fig.1) on Earth surface (, the accuracy is ±10 nT for 95% of reported data and ±5 nT for definitive data, with one sample per 5 seconds, in the case of Vector magnetometer (F(XYZ) or F(HDZ)) and 1 nT, with 3 samples per second, for Scalar Magnetometer (F).

The geomagnetic vector projection H is measured with relative accuracy less or equal to1 nT by a fluxgate, feedback based device of rather original and simple, but powerful construction. (know- how of JINR, Dubna, Boris Vasiliev, 1998, private communication). It is used with 2.4 samples per second. Due to technical reasons the sensor was oriented under the Horizon in a manner that the measured value of H is around 20000 nT (Figures 1 and 2).

The minute averaged value     and it’s error     are     and     ,
where   , are the measured Nm=144 times per values of the field and their experimental error. The standard deviation     and it’s error     for every minute are:
Analyzing the correlations between the behavior of the geomagnetic field, Earth tidal gravitational potential and the occurred earthquakes (1999 to 2001) it turns out that the daily averaged value of   ( and   ), which we denote by Sig (dSig), is playing the role of earthquake precursor.

Figure 3 illustrates the behavior of geomagnetic field component and its variation for a period without earthquakes precursor in the region. Figures 4 and 5 illustrate the behavior of geomagnetic field and its variation, which is unusual. Is this case there is a geomagnetic quake, which is precursor for incoming event (earthquake or earthquakes). One has to be sure that there are not a cosmos or Sun wind reasons for the geomagnetic quake ( see the site: )


For example. in Figure 4 the predicted time was 3 +/- 1 June, 2002 and the prediction was confirmed with earthquake, occurred at 03/06/2002 02:04, Lat41.95N, Lon23.10E, Dep8, Mag2.6, Ml, 50 km from Sofia, SChtM = 598 [Mag/r2].
The preliminary Fourier analysis of data gives the fact that the bigger geomagnetic variations are cased from the arriving for hours time period new frequencies with periods from 10-th of seconds till 10-th of minutes and with very specific amplitude behavior. Such spectrum, which arrives for hour period of time, is invisible for minute samples measuring.
The probability time window of the incoming event (or events) is defined by the next date of the Earth tidal potential extremum with tolerance approximately for the tidal minimum +/-1 day and for the maximum +/-2 days- Figure 6.

The uncertainty problem of distinguishing the predicted event (or group of events – aftershocks) from the events which occurred in the region in the predicted time window is solved with new earthquake function SChtM:

  [Thousand km]

The sense of function SChtM is a density distribution of earthquake's Magnitude. In the point of measurement SChtM is logarithmically proportional to the energy influence of the earthquakes. It is important to stress out that the first consideration of the Magnitude and distance dependences was obtained on the basis of nonlinear inverse problem methods. Obviously, the nearer and biggest earthquake (relatively biggest value of SChtM) will bear more electropotential variations, which will generate more power geomagnetic quake. At this stage of the researching the measure of daily geomagnetic state can serve the averaged for 24 hours (1440 minutes) values of standard deviations:

(1)       ,  

The simple and usually working criterion for evidence of geomagnetic quake is when Sig increases for two consecutive days i, i+1 and the differences between the values of Sig are bigger than the mean arithmetic sum of their errors dSig:

(2)   Abs(Sig i+1 - Sig i) > (Dsig i+ DSig i+1)/2.

If criteria (2) are fulfilled and there are not a Cosmos and Sun generated variations of geomagnetic field, it could be concluded that the geomagnetic quake has happened. Such quake is unique precursor for incoming earthquake and in the next minimum or maximum of the local Tidal gravitational potential somewhere in the region this predicted earthquake will occur. For some of the cases, the criteria (2) have to be calculated for decades of hours or ten’s of minutes. The signal for earthquakes with different epicenters we observe in the case when the specific behavior of field and its standard deviation occurs more than ones per day at different hours.
The analysis of the precursor function Sig on the basis of special digital 5 points derivatives can serve in the future for creating the algorithm for automated alert system.It is obvious that the more detailed time window can be achieved by analyzing the daily variations of tidal potential, calculated every hour.
As an example, the parameters of predicted events (Figure 5) are presented in Table 1.

Table 1

DMYhm Latitude Longitude Dept [km] Magnitude Distance [hundred km] ChtM
03/07/2003 15:49 42.96   25.29 10 2.6 1.10 219
03/07/2003 20:51 41.96   23.28 12 2.9 0.48 684
05/07/2003 21:58 40.35   26.14 2 4.0 2.97 70
06/07/2003 19:10 40.46   26.01 10 5.7 3.16 89
06/07/2003 19:39 40.62   25.25 2 4.2 2.35 111
06/07/2003 20:02 40.72   25.98 20 3.2 2.54 71
06/07/2003 20:10 40.46   26.08 10 5.0 3.20 76
06/07/2003 20:48 40.28   26.08 2 4.1 3.01 70
06/07/2003 21:58 40.38   26.10 10 3.9 2.92 70
06/07/2003 22:05 40.34   26.00 2 3.9 2.92 71
06/07/2003 22:42 40.95   26.00 10 4.6 2.80 89
07/07/2003 00:24 40.25   25.98 2 3.7 3.00 64
07/07/2003 00:48 40.44   25.87 18 3.4 2.77 66
07/07/2003 07:15 41.67   24.88 10 3.1 1.21 230
07/07/2003 16:17 40.37   25.91 10 3.3 2.85 61
08/07/2003 02:48 41.82   22.93 12 2.8 0.65 465
08/07/2003 12:00 42.84   23.32 10 2.5 0.50 565

At this stage of the study all earthquakes have the same SChtM for different definitions of the magnitude. After the developing of mathematical models of empirical and theoretical dependences between incoming earthquake processes, magnetic quake and parameters of earthquake on the basis of inverse nonlinear problem we will obtain a set of SChtM functions in correspondence with the different definition of magnitude. The volumes, its depth, chemical and geological structures of the region have to be included in the dependences as well.
From January 1, 2004 is using function SChtM, which is the density of earthquake energy in the measurement point:
SChtM=exp( Magnitude ) / (Req+Distance)2, Req = 0.40+Dept [hundred km]
The other variables are the same.
In the vernal and autumnal equinox the approach (2) needs more statistics on the bases of geomagnetic data from at least 3 monitoring points for discovering the more hidden dependences between geomagnetic quakes, more flatter tidal variations and incoming event or events.

SChtM/June 2004, Sofia