Osservatorio Permanente Emissioni Radiosismiche

OPERA 2015 – Osservatorio Permanente Emissioni Radiosismiche

Osservatorio Permanente Emissioni Radiosismiche
(Radio-seismic Emission Permanent Observatory) by Renato Romero

PART I – Premises and start of the monitoring, until the 2018 end
(see the section below for the complete list of collaborators who work on this project)
Skip to part II – Data analysis and considerations or directly to monitored events

In this research project we will look at a particular type of hypothetical seismic precursors: radio emissions at very low frequencies. According to many publications, some of which also appeared in authoritative scientific bulletins, these signals would be present for all earthquakes of magnitude greater than 6, at low frequencies, with intensity strong enough to allow worldwide detection.

Low frequency radio precursor signals would begin several hours before the event, and given their intensity they would be receivable on the whole planet: they would cease immediately after the earthquake. Many of these studies are based upon data acquired with non-commercial, bespoke systems, having technical specifications comparable to those of more expensive professional equipment such as induction coils or flux gates; others are based on professional systems.

The question arises: if these signals are strong and they are always present at every high-intensity earthquake, why are they currently not used to make predictions? The OPERA project, with a one year listening campaign, will attempt to answer this question.


There usually is a seismic event of some significance to direct the attention of the media about ability to predict earthquakes, leaving transpire that, thanks to the work of researchers, it will soon be possible to have the forecasts as today we already have the weather. At every opportunity it seems that the aim of the forecast is now at your fingertips. But the ones who have few years on the shoulders will not be in trouble to remember a certain recurrence of the news.

Back to the mind by analogy the future proposed in the late ’60s, in the wake of the enthusiasm immediately after the conquest of the Moon: in the early ’70s was English TV series such “SPACE 1999” to hypothesize within 30 years the exploitation of the moon as a nuclear waste repository , with a human colony on a permanent basis on the satellite.

SPACE 1999 logo
Weather forecast
BBC earthquake reports

Sooner we should get down on Mars, then the solar system and to follow, the colonization of the entire universe. Today, however, after almost half a century, except for six signs of life on the International Space Station, we still are here, on Earth. When it comes to earthquake prediction are therefore faced with another illusion as the space conquest? With this comment we do not want to immediately exclude the opportunity to predict earthquakes.

But we also must objectively observe that this topic lends itself to a lot of speculations. Opera will try to base its conclusions solely based on records, acquired from monitoring stations, integrated in some cases from data coming by geophysical observatories in northern Europe.


The OPERA project is a planned activity, which consists on one year VLF systematic monitoring (2015), with a writing from time to time a technical report of any episode that exceeds a certain threshold scale, calculated by comparing the earthquake intensity and the distance from it for each single monitoring unit. OPERA 2015 stands for Osservatorio Permanente Emissioni Radiosismiche (Permanent Observatory of Radioseismic Emissions).

The first OPERA Project, was in 1997 and it was addressed to the frequency of occurrence, the hours more propitious, and the seasons in which it was possible to listen radio nature signals, like Whistlers. The results were published on INSPIRE Journal, but the project was realized with absolutely primitive resources, including a cassette recorder, a boiler timer, the RS-4 INSPIRE electric field receiver and an aerial loop to detect magnetic field. Despite the extreme craftsmanship of devices, and the measurements in a single site only, the obtained results at that time are still comparable to those we obtain without difficulty by our network nowadays.

therefore now we have a lot more: OPERA 2015 is an ambitious project aimed at producing an official document certifying the results of our research, using data gathered during a campaign Listens for one year, using the existing monitoring network.

It will include the monitoring of all seismic events on a global scale with a magnitude greater than 7.5, and, gradually to decline, down to the earthquakes of magnitude 4 occurred in proximity of at least one of the five stations, plus all the geomagnetic pulsations occurred, relating to solar storms.

At the end of the campaign a detailed analysis of the signals collected from workstations (a report for each event with the signals collected from the participating stations) will form a rich documentation about the extremely low electromagnetic frequencies and its relevance or not to seismic events, including volcanic phenomena given that a station is right on the slopes of Etna volcano.

With this project we don’t want invalidate anyone: taking advantage of the high seismicity of the Italian peninsula we only wish to make an organic search based on a considerable number of observations, performed by monitoring stations spread on our national territory.

In the above chart the world seen from Italy in an azimuthal chart (map edited with Azimuth Map V 3.2 by Tony Field VE6YP). The OPERA project will refrain as much as possible from any kind of conjecture: final conclusions will only be the result of these measures.


The VLF Openlab Observatory is an Italian network of VLF monitoring stations. They share the spectrogram settings, like brightness, contrast, scroll time, FFT resolution… so that results can be easily compared: it is very easy this way to understand if the same signal has been received from one or more stations simultaneously.

This structure consists of six stations and begins from the free collaborations within the activities of OpenLab www.vlf.it, the portal since 1999 dedicated to the techniques of receiving and to the study of natural radio signals of very low frequency. In addition to publishing articles vlf.it has always had attention to VLF monitoring: permanent recordings begin in 2002 from the station of Cumiana (01 in the chart above), but the network as we know it today, begins in 2007 with the activation of the station at Pontese dew (02), in Gran Paradiso Park, activated with the support of the CSP (Regione Piemonte Research Center). The main objective of the network was, since from the beginning, the monitoring of signals at very low frequency, with a focus on the presence of possible radio seismic precursors.

Italy map with OpenLab monitoring stations siting

Detail of station 1-2-6: red lines indicate tectonic lines (full chart here)

Along the way they added two locations on the slopes of Etna volcano in Sicily and allocated inside the National Park structure (03) an other one in Emilia Romagna Region (04), a station on the National Institute of Radio Astronomy of Bologna (05) in autumn 2012, and last the monitoring station in the city of Turin (06), on the hills near the city .


The VLF OPENLAB OBSERVATORY monitoring structure is today composed by:

01 – IK1QFK CUMIANA (TO), NW Italy (Azimuth chart)

Maintained by Renato Romero (Network coordinator) and Luca Feletti
Equipped with: two induction coil, big Marconi antenna, geophone, ogg vorbis streaming, GPS

02 – CSP VLF MONITORING STATION From Pontese dew (TO), NW Italy (Az. chart)

Maintained by CSP Team: Roberto Borri, Luca Seoli and Claudio Re
Equipped with: multiturns aeral loop and differential electric field receiver (ADA)
03 – ETNA RADIO OBSERVATORY From NICOLOSI (CT), Etna Park, Sicily, S Italy (Az. chart)

Maintained by Rosario Catania – Alessandro Longo – Salvo Caffo
Equipped with: Induction coil, orthogonal VLF loop, geophone.

04 – ROMAGNA OBSERV. From Fiumicino(FC) / Sogliano al Rubicone(FC), N Italy. (Az. chart)

Maintained by Federico Scremin
Equipped with: Induction Coil, marconi T antenna

05 – NORTHERN CROSS RADIOTELESCOPE From Medicina (IT – Bologna), N Italy (Az. chart)

Maintained by Jader Monari, Massimo Silvestri, and Renzo Cabassi
Equipped with: induction coil, static electric field receiver.


maintained By Claudio Re and Fabrizio Francione
Equipped with: Marconi Antenna, differential electric field receiver (ADA), fluxgate
The stations are followed by operators locally and centrally coordinated through TeamViewer.


During the year of activity, OPERA will monitor seismic events that exceed an attention threshold: they will be chosen with a scalar criterion later explained. Even low frequency earth’s geomagnetic activity will be monitored: in the OPERA tabs will also be included the receptions of geomagnetic pulsations, effect of solar flares and solar wind investing our planet.
Here is the summary of signal categories and workstation involved from time to time:

– SRS (spectral resonances structures)
– PC1 (type 1 geomagnetic pulsations)
– PAROXYSM (lava flows)
–> when received from two or more stations
–> when received by one or more stations + Kiruna
–> all the stations
–> station that exceed a scalar criterion (documented below)
–> only if the monitoring station is positioned just above the earthquake
–> Etna volcano monitoring station only
–> all the monitoring station

The events database will be available on-line to the voice “MONITORED EVENTS” below and updated as that they occur.
As earthquake data source will be used the following official websites: – Istituto Nazionale di Geofisica e Vulcanologia

Also for reference of the geomagnetic activity we prefer to use data from institutional observers:

Their data are referenced, its equipment ise maintained by professionals, and the data reported have metro-logical value. They are known worldwide for the quality of their reception. Moreover, the presence of a strong signal worldwide coverage, if seen from amateur stations, less sensitive, and positioned in places with high electromagnetic pollution (as in an home apartment), will can not escape to an high sensitivity observatory placed in areas carefully chosen among the most quiet ones of the planet.


We started from an initial hypothesis: if radio seismic precursors exist they will be easier to be received if we are near the earthquake and if the earthquake is of high intensity. This statement may seem trivial but allows us to establish a choice criterion. If we attribute to a given magnitude an arbitrary amount of RF energy, calculating the attenuation of an electromagnetic signal, we get a gradient that indicates the possibility of receiving a precursor, determined by the combination of distance and amount of generated RF energy.

Since the possibility of propagation of a signal in the soil increases with the wavelength, it is most likely that if in depth are generated radio waves, only those with lowest frequency will be able to arrive up to the surface. And low frequencies means long wavelength and long Fresnel distance: in most cases our monitoring station will be in reactive area. The attenuation computing must consider the isotropic attenuation from free space (that in levels means 1 / 1, linear), the magnetic filed attenuation in reactive area according to Biot Savart Law (in level goes as 1 / r 3), and electric field attenuation in reactive area (in level as 1 / r2). Also according to the known data available on the propagation of these signals we have achieved a table that summarizes with good approximation the sum of these three behaviours: it is calculated considering that the distance covered increases with the cube of the power generated, assuming that our centre band is 10 Hz.

Starting from Wikipedia Richter scale reference (http://it.wikipedia.org/wiki/Scala_Richter) we calculated:

This assessment is arbitrary, because it does not consider some factors, we know: not necessarily all earthquakes can be source of RF signals, not all earthquakes occur at the same depth, not all soils are equal… but it allows us to establish a choice criterion, to avoid to monitor thousands of seismic events, but choose only those, where the probability to detect hypothetical precursors is higher. Anyway for events that are not highlighted here, we can still, if necessary, recover all recordings (spectrograms and source wave files) from our database, also in a second time. But here’s how we got to calculate the selection table above.

Since the ’50s, many mathematical models for the study of the propagation of electromagnetic waves in cavity have been developed from the ELF (<30 Hz) up to VLF (3-30 kHz) range. In these models, the wave guide consists of two concentric spherical surfaces, with a reflection coefficient almost unitary, one of which is representative of the Earth’s surface and the other of the lower limit of the ionosphere. Within this guide, the field sources are modelled by vertical electric dipole (VED), vertical magnetic dipole (VMD) and by horizontal electric dipole (HED). In order to support mathematically the method of Radio-seismic Indicator (RI), previously introduced, we can consider the presence of a VED source within a wave guide such as that described. The effective wave guide height is about 60-90 km. In the range of SLF / EFL, this height is less than the free-space wavelength of considered electromagnetic phenomena; so only a zero-order mode can propagate.

With respect to the following figure:

with a VED source, the electric and magnetic fields (for TM waves) can be expressed through the Maxwell’s equations, using a potential. Taking into account the particular geometry of the problem and the type of active sources, it can be demonstrated that the components of these fields are so expressible:


The potential U, through which the field components are expressed, satisfies the Helmholtz equation:

Using appropriate boundary conditions, derived from the expression of normalized surface impedances that define the wave guide, the Helmholtz equation can be solved, obtaining the expression of the potential U as a function of r and θ.
Through mathematical manipulations here omitted, but entirely explained in [1], this expression can be used to obtain, a very complex complete formulas for the electromagnetic field of a VED in the Earth-ionosphere wave guide. In SLF, the above expressions can be suitably approximated (as stated in [1]) and transformed into a easier to use form. In ELF, in which we are interested in, it is not possible to use the above simplification as the wavelength, in this case, is comparable with the Earth’s circumference. Then, some algorithms were developed (some of them are described in [1]) in order to overcome the above mathematical difficulties and obtain usable formulas.


Earth’s radius: 6.370 km
current moment of the dipole: 1 A m
ground conductivity: 10^-4 S/m
ionosphere conductivity: 10^-5 S/m
ionospheric equivalent reflection height: 70 km

using the above algorithms, the magnitude of the electromagnetic field components can be obtained and showed in the following figure for f = 1 Hz, 3 Hz, 5 Hz, 10 Hz, 20 Hz and 30 Hz.

The method of Radio-seismic Indicator (RI), introduced to discriminate those earthquakes potentially interesting with respect to the electromagnetic activity, among their totality, is a tool, easy to use, whose validity is limited to the above purpose, but which, in any case, does not counteract the rigorous mathematical theory.
In fact, we can evaluate, for example, the magnetic field magnitude as a function of distance with the RI method and compare the result with those obtained with the algorithm described in [1].
The result of the comparison is shown in the following figure and we can note a good behaviour of RI method.

For further details, we suggest the consultation of the listed references.
A text in electro magnetics is also useful to find the theory about the Maxwell’s equations and their solution in particular cases with respect to the sources and geometry of the problem. Although relative to a particular scope and devoted to frequencies higher than those considered here, we want remember the study conducted by Bannister and Fraser-Smith [3], whose results constitute a reference element for the argument of our interest.
In [1] is also shown a comparison between the results obtained with the algorithms developed for the solution of the field equations and those obtained by Bannister.

An observer can evaluate the Radioseismic Indicator (RI) using the following formula:

where M is the earthquake magnitude (in Richter scale unit) and D is the distance (in km) between the event and the observer. RI is conventionally expressed in dBe. We have decided to use the letter ”” to identify an””arthquake reference situation (at which RI is equal to 0 dBe): an earthquake whose magnitude is equal to 4, in Richter scale unit, and whose epicenter is located at 100 km from the observer.

Here below a graphic representation of Radio-seismic indicator, with a simply on line calculator:

Radio-seismic calculator
(works with Firefox)

Insert magnitude:
Insert distance (Km):

A simply version of radio seismic calculator in Excel is available here:

The complete Excel sheet we use to evaluate seismic list compatibility

[1] Weiyan Pan, Kai Li “Propagation of SLF/ELF Electromagnetic Waves”; Springer, 2018
[2] Peng HY, Tao W, Pan WY, Guo LX “Numerical integral method for ELF fields excited by vertical electric dipole in asymmetric Earth-ionosphere cavity”; Chin J Radio Sci, 2012
[3] Bannister PR, Fraser-Smith AC “Reception of ELF signals at antipodal distances”
[4] Bannister PR “ELF propagation update”; IEEE, 1984
[5] Bannister PR, Wolkoff EA, Katan JR, Williams FJ “Far-field, extremely-low-frequency propagation measurement”; Radio Sci
[6] Watt A. D. “VLF Radio Engineering”; Pergamon Press, 1967

This part has been developed by Luca Feletti and Renato Romero

The activities of the research project can also be followed with mobile devices using our Opera2015 APP.
The APP can be downloaded from the following links:






Or directly using the QR code.

Made by Rosario Catania, Etna Radio Observatory.


It may seem a trivial statement, but the basic requirement to recognize a seismic anomaly is to know exactly what is the ordinary situation. Usually it is the natural background constituted for example by Schumann resonances or geomagnetic pulsations during solar storms, but it can also be “ordinary” signals due to anthropogenic disturbances: they are many and they shouldn’t be interpreted as precursors. Disturbances also arise from microphonic effect: our sensors are sensitive to the wind, rain and storms: it is essential to recognize at a glance their signals, their mark on spectrograms. Hereinafter a small gallery with signals related to anomalies collected from various monitoring stations, and certainly not of seismic origin.




TRAIN 5 km away from COILS



How to start up a VLF observatory
How to put a VLF observatory on-line
Create your own streaming VLF Audio program

Some preliminary notes at the beginning of the monitoring

Here some preliminary notes at the beginning of the monitoring

We use FFT analysis because it is the best way to graphically display the natural background noise, made by Schumann Resonances from 5 to 50 Hz, and Geomagnetic Pulsations over a hiss background below 5 Hz. But what are we really looking for? We are seeking anomalies extraneous to the noise background: burst, glitch, hiss, whistle, tone… or anomalous background noise level variations. According to the spirit of Openlab, which governed all work so fare published on the site www.vlf.it, also the results of this research campaign will be shared with everyone. In addition to the spectrograms for each event we will also be publishing wave files sources: in this way you will be able to reprocess the original signal, directly from your home.

At the beginning of the monitoring activities we wondered if it was necessary to calibrate receiving systems, because all workstations are using sound card and we know that sound cards have not a precise reference frequency. We did test it but, due to high thermal excursion, stations although calibrated didn’t keep the alignment for a very long time. After various tests we have therefore decided to not calibrate monitoring station frequency reference. We have preferred to prioritize “the time reference” of each location. Any workstation runs a time client software, periodically connected to a time server. Hours reported on workstation files are therefore accurate within +/- 20 ms. For any audio file, available for analysis, we are sure of the starting time accuracy: if you require a precise time reference for the entire recording, open an audio editor, and normalize the file length to 4 or 8 hours exactly (14400 seconds or 28800) . This way the times will correspond with an accuracy of +/- 20ms through the entire recording.

Online Casino Promotions

Signals until now received by our network (the first monitoring station is active since 2002), lead us to suppose it will be not possible to arrive to an earthquake forecasting. But we will be very happy if one day we can say: “we were wrong!”

OPERA 2018

Osservatorio Permanente Emissioni Radiosismiche
(Radio-seismic Emission Permanent Observatory)
By Renato Romero

PART II – Data analysis and considerations

The events selection treated in OPERA project was made on data provided by INGV, the Italian Institute of Geophysics and Volcanology. The National Seismic Network (RSN) has traced and reported the data of 15’532 earthquakes, occurred between January 1 and December 31, in Italy and world in 2015. In the same year, in Italy and in geographically neighboring areas, occurred on average just over 40 earthquakes a day, an earthquake almost every half hour.

Table activity of 1st January 2015 with 28 events. Of all these only one, in yellow highlighted, exceeds the magnitude threshold of 3.2.

Compared to previous years the number of localized earthquakes is significantly decreased: in fact, both in 2013 as in 2014, there were more than twenty thousand events recorded in Italy, mainly due to some seismic sequences, with numerous events, which lasted in the months. Like every year, the vast majority of recorded earthquakes had a magnitude of less than 2.0: more than 13’000 events.

Map of Italy, with 2018 earthquakes and our 6 monitoring stations

Detail of the area around Sicily

Of 15532 episodes documented by INGV, only those with magnitude greater than 3.2 were selected for our study , which have been around 279. On these 279 was applied the RadioSeismic Index Calculation cdescribed above, selecting 45 events to be analyzed, those with the highest probability of finding associated radio emission.

Apuestas Deportivas Argentina

January 2018 table activity for magnitude more than 3.2: light blue highlighted the events with high RI index.

The selection threshold was placed with an index of -10 dBe RI, with a few positive exceptions where the earthquake had been detected in a very intense way by the station geophones, or perceived directly by the people. And with some negative exception in cases where the stations interested were out of service for maintenance. This led to an analysis which covered 41 events, with a greater possibility of existence of associated radio signals, most single and someone multiple, including also some paroxysms.

Total selected events table.

SEISMIC MONITORED EVENTS (which exceeded the scalar threshold)

Free web counters

This website and its pages do not incorporate profiling cookies, but only technical ones, used to allow the correct counter working (Shinystat only)