lightning scatter: a faint and rare mode of propagation

HAL Id: hal-00638547https://hal.archives-ouvertes.fr/hal-00638547

Submitted on 5 Nov 2011

HAL is a multi-disciplinary open accessarchive for the deposit and dissemination of sci-entific research documents, whether they are pub-lished or not. The documents may come fromteaching and research institutions in France orabroad, or from public or private research centers.

L’archive ouverte pluridisciplinaire HAL, estdestinée au dépôt et à la diffusion de documentsscientifiques de niveau recherche, publiés ou non,émanant des établissements d’enseignement et derecherche français ou étrangers, des laboratoirespublics ou privés.

Lightning scatter: a faint and rare mode of propagationJean-Louis Rault

To cite this version:Jean-Louis Rault. Lightning scatter: a faint and rare mode of propagation. VHF Communications,2005, 2, pp.111-120. �hal-00638547�

https://hal.archives-ouvertes.fr/hal-00638547

https://hal.archives-ouvertes.fr

Reflections on layers of the ionosphere,reflections on ionised meteorite trails,echoes on airplanes, EME, reflectionson auroral ionised clouds … Variousmodes of radio propagation have beenexplored and used for a long time byradio amateurs.

Could lightning also be capable ofscattering radio waves?

This article deals successively with atheoretical and a practical approach tolightning scatter.

1.0

Theoretical point of view

1.1 Is a flash of lightning able to

reflect radio waves?

Any ionised medium is liable to reflect aradio wave. A thunderbolt is a violentelectric discharge that heats and ionisesthe ambient air. Temperatures can reach20 or 30,000°K and the electron densitycan raise up to 1017

to 1018 electrons percm3 [2]. Knowing that the electrondensity necessary to get full radio wavesreflection is as follows [4]:

m and e being respectively the mass andthe electric charge of an electron and f

N

being the frequency of the reflectedwave, one can see that a flash of light-ning is theoretically able to reflect theentire radio spectrum.

Replacing m and e by their numericalvalues and using electrons per cm3 for N

e

, the highest reflected frequency is:

with fN

given in MHz and Ne

in electronsper cm3.

So one can see that 3 x 109 electrons percm3 is a density that is high enough toallow a reflection of VHF or UHF radiowaves.

A lightning flash ionised channel can beseveral kilometres long [6], with a diam-eter being a few centimetres in size. 80 %of the flashes are of cloud-cloud type(CC, see fig 1), 20% being of the cloud-ground type (CG, see fig 2).

The extremely high temperature gradientaffecting the air layer close to electricdischarge could possibly participate in

Jean-L. Rault F6AGR

Lightning scatter: a faint and

rare mode of propagation

2

2

e

fmN N

e

⋅⋅=

π(1)

41024,1 ⋅

=e

N

Nf (2)

VHF COMMUNICATIONS 2/2005

111

the radio waves reflection.

1.2 What could be the effective

duration of reflection from a flash of

lightning?

A CC or a CG lightning flash is com-posed of several phases.

At the beginning, low intensity precur-sors (where electrical current reaches afew hundreds amperes) appear in ahighly charged part of a cloud. When aconductive channel is connected betweentwo parts of a cloud with opposite polari-ties, or between a cloud and the ground, areturn stroke appears which carries ahuge quantity of electricity (several ten

thousands amperes).

A complete lightning flash includes sev-eral return strokes and can last severalhundreds of milliseconds [2], [3].

1.3 What is the probability of

occurrence of lightning flashes?

Most of the 3000 thunderstorms thatappear each day around the world occurin the equatorial area.

In Europe, the occurrence is around sometens of electrical activity days per year[1]. See Fig 3 for a map of Franceshowing an example of the yearly statis-tics. Knowing that a single thunderstorm

Fig 1: Cloud toCloud (CC) flash.

Fig 2: Cloud toground (CG)thunderbolt.


112

generates hundreds or thousands of light-ning flashes, one can see that the prob-ability to get some echoes is not negligi-ble.

1.4 What is the maximum distance

one can expect for lightning echoes ?

The maximum echo range for a transmit-ter and a receiver located both at groundlevel (if we consider they have the samealtitude) depends on the mirror altitude.

The range is as follows:

R being the Earth radius and h the heightof the reflecting part of the lightningflash.

For example, a height of 5000m gives arange of 500km, assuming that the mirroris located half way of the transmitter andreceiver, and a little bit more if theatmosphere refraction is taken into ac-count.

2.0

Practical experiment

A reliable test procedure has to be estab-lished, to be sure to catch, record andanalyse any lightning scatter.

Just listening to distant beacons during astormy day is too subjective and notconvincing enough to prove that light-ning scatter really exists.

The following key points were taken intoaccount when establishing the test pro-gramme:

• choice of a radio beacon transmittinga stable and well known signal

• absence of interference around thebeacon frequency

• distance between beacon and receiverlarge enough to avoid any receptionwhen there is no ordinary tropo-spheric propagation

• probability of frequent thunderstormson the beacon/receiver path

Fig 3: Annualstatistics aboutthunderstorms inFrance.

hR

RRD

+

⋅⋅= arccos2max


113

• automatic record of receiver audiooutput to allow further batch analysis

• simultaneous automatic record of ra-dio noise generated by the lightningflashes in order to allow further cor-relations analysis

• monitoring of thunderstorm predic-tions and real time activity thanks toweather agencies Internet providers

With all these prerequisites in mind, acampaign of systematic audio recordswas performed during summer 2004. Theradio amateur beacon, F5XAG, was cho-sen because it fulfils most of the requiredcriteria.

At each end of the 648km path underinvestigation (see fig 4), the equipmentwas as follows:

Fig 4: F5XAG toF6AGR path.

Fig 5:Schematicdiagram ofexperimentalequipment.


114

Beacon:

Name: F5XAGLocation: IN93WCAltitude: 550mFrequency: 432.413MHzERP: 40WBeam: NNE

UHF receiver (432MHz):

Name: F6AGRLocation: JN18DQAnt. altitude 66m

Antenna 2 x 10 element yagi RHCP/LHCP

Receiver ICOM IC-821H + masthead preamplifier

LF receiver (137kHz):

Name: F6AGRLocation: JN18DQAnt. altitude 56mAntenna 23 turns 1.2m2 square

loopReceiver ICOM IC-738

Antennas configurations are shown in

Fig 6:144MHz and432MHz aerials.

Fig 7: 137kHz receivingloop antenna.


115

figs 6 and 7.

The 137kHz amateur band was chosen asthe lightning flash monitoring frequencybecause it gives a good compromise onthe detection range of European thunder-storms. Watching flashes on VLF wouldhave given too many pulses generated byvery distant thunderstorms. On the otherhand, the energy radiated by a thunder-bolt in the VHF/UHF band is quite small,

and so the range on these bands islimited.

2.1 Receiver/recorder configuration

The audio outputs of the 432MHz and137kHz receivers are connected to a PCfitted with a stereo sound card (see fig 5).The computer is also used to analyse therecords and to track any interesting ech-oes.

Fig 8: F5XAG long burst of signal.

Fig 9: F5XAG lightning scatter.


116

The UHF receiver is equipped with anOCXO, which is mandatory to tune theVFO to the correct frequency withoutreceiving any permanent signal (providedthat the beacon transmits on the correctfrequency).

Both audio channels are recorded inparallel, with 16bit resolution. The sam-ple frequency being adjusted to get thebest compromise between the audio passband and the volume of stored data. Forexample, one hour of stereo recordingwith an audio pass band of 4kHz and a16bit resolution is more than 115Mb.

Recordings are performed in real time onthe hard drive of the computer and thenstored on 4.7Gb data DVDs. Data com-pression such as those performed by MP3algorithms are not usable, because theydistort the signal too much, so only WAVrecordings offer the necessary recordfidelity.

To identify any faint and short echoes, a

solution is to use a spectral analysis tool.Although an FFT algorithm is not thebest tool to track short pulses, FFTsoftware is very easy to find and todownload from Internet.

A graphical display showing frequencyon the Y axis, time on the X axis andsome colours to give an amplitude indi-cation is very easy and pleasant to exam-ine visually. A quick look is much moreeffective for identifying a short and faintecho than spending a very long time tolisten to white noise.

Two complementary software tools wereused for the experiment:

• CoolEdit 2000 from Syntrillium

• Spectrum Lab, developed by Wolf-gang Büscher DL4YHF

CoolEdit 2000 is very valuable for jug-gling with long audio records, offeringuseful functions such as Rewind, For-ward, simple FFT controls, time and

Fig 10: Real timeobservation oflightning strikes(21 July 2004). ed -looks better incolour, see VHFCommunicationsweb site.


117

frequency zooms, all of which are verysimple to use.

Spectrum Lab is a powerful spectralanalysis tools kit allowing many param-eters adjustments. However it requiressome knowledge of signal and dataprocessing to be fruitfully controlled.

3.0

Results of the experiment

Several 24 hours-a-day audio recordshave been performed during calm andthundery days of summer 2004.

The first result is that on this N/S path,

bursts of signal coming from the F5XAGbeacon were clearly received night andday. Each burst was some tens of secondslong, separated by minutes or tens ofminutes of silence. An example of such aburst is shown in fig 8. Some bursts showa typical Doppler effect indicating thatthe signal is may be reflected from highaltitude airplanes, but some otherspresent some frequency splitting anddrifts which are not easily explainable.

Several occurrences of lightning scatteron 432MHz were clearly identified in thesummer 2004 records. Fig 9 shows anexample of such an echo received on Julythe 21st around 20:30 local time.

The lower trace (432 MHz channel)shows a beacon echo around 1200Hz thatgoes on for about 500 milliseconds. Thesignal to noise ratio is about 10dB. Nonoise at all generated by the lightning

Fig 11: F5XAR lightning scatter.


118

flash itself was detected. The upper traceshows the corresponding 137kHz activ-ity, which consists in broadband spikesgenerated by each electrical discharge.Fig 10 shows the real time status ofthunderbolt ground hits at 20:30.

During summer 2004, a sea thunderstormin Brittany allowed another lightningscatter hunt on 144 MHz. The groundpath was 82 km long (see fig 12). Theconditions were as follows:

Fig 11 shows an example of the echoeswhich were clearly identified. The hori-zontal line indicates that the beacon wasreceived most of the time and the verticalline shows the wide band noise receivedfrom the flash itself. On the example, theecho was composed of two successivebursts, with a total duration of less than300mS. The ratio echo/permanent carrierwas better than 20dB.

Further to the encouraging results ob-tained on 144 and 432MHz, old recordsperformed previously in 2002 on 21MHzfor a meteor scatter study were re-ana-lysed carefully in order to track anypossible lightning scatter on that band.The station used as transmitter was apowerful short waves French broadcaststation (Radio-France International) thatis very useful to track meteor scatter

activity. The 21MHz path was 250kmlong.

The results were amazing and lightningechoes were identified at a rate of about 6per minute (see fig 13). The refractedcarrier was received permanently and theecho level was 6 to 10dB over thepermanent carrier. The length of eachecho was a few hundreds milliseconds.

4.0

Conclusion

This lightning scatter experiment showsthat radio scattering from thunderboltsreally exist.

But many questions remain unan-swered… What is the best location andorientation of a lightning flash referred toa transmitting and a receiving station?Are some frequencies better than others?What could be the maximum length of anecho?

Elves and sprites, triggered by powerfulpositive lightning flashes have been dis-covered recently, thanks to sensitivevideo cameras. Could these large lumi-

Fig 12: F5XAR toF6AGR/P path.


119

nous discharges happening in the lowerpart of the ionosphere also contribute tothe scattering of radio waves?

That's another interesting story!

5.0

Bibliography

[1] Les Orages, Frank Roux, 354 pages,Documents Payot, ISBN 2-228-88328-X

[2] The Lightning Flash, Vernon Cooray,574 pages, IEE Power series, ISBN 0-85296-780-2

[3] Lightning, Physics and effects,

Vladimir A. Rakov et Martin A. Uman,Cambridge University Press

[4] Télécommunications et Infrastruc-tures. Gérard Barué, THALES Ellipses,ISBN 2-7298-1323-3

[5] La Propagation des Ondes, SergeCanivenc F8SH, Tome 1, 256 pages,SORACOM Editions, ISBN 2-904-032-23-1

[6] Lightning flash lengths deduced fromVHF radiation for a Colorado Thunder-storm, Etude de Eric Defer & James E.Dye, National Center for AtmosphericResearch, Boulder, 2002.

Ed - Colour pictures from this article canbe found on the VHF Communicationsweb site - www.vhfcomm.co.uk

Fig 13: Radio France lightning scatter.


120