Fullscreen HTML5Med Res (??MB)HTML4

Can EARTHQUAKES by low frequency r The 2010 Christchurch earthquake. (Photo: http://rebuildingchristchurch.wordpress.com/2010/09/07/rebuilding-christchurch/) Do low frequency radio signal variations provide a clue to the onset of an earthquake? There are no ready answers but it’s tempting to investigate. This article gives some of the background and suggests how you can monitor low frequency radio signals using a simple preamp circuit feeding the sound-card input on your computer. A lthough seemingly a modern nightmare, earthquakes have always been a fact of life. Every year, several deadly 8-8.9, scores of 7-7.9, hundreds of 6-6.9 and thousands of 5-5.9 magnitude quakes strike around the globe. The Richter scales are logarithmic, so a magnitude 7 quake has a shaking amplitude 10 times greater than a 6. Further damage and deaths often result from the quake’s aftermath. Most fatalities in the offshore Indian Ocean Boxing Day 2004 earthquake (magnitude 9.2) were not caused by the earthquake itself but were Tsunamirelated. Deaths can run to hundreds of thousands but damage and casualties vary enormously, often relating 14  Silicon Chip to population densities, building techniques, terrain and soils – or sheer luck and timing. In 2010 the devastating January 12th Haitian and Canterbury’s (NZ) early morning 4th September quakes were of similar (7.1) magnitude but casualties of some 250,000 in Haiti contrasted with none in Christchurch. However on February 22nd 2011, a close and shallow 6.3 “aftershock” struck Christchurch at lunch time, killing hundreds and causing devastating damage. Precursors It’s only in recent times that it’s been realised earthquakes arise from the earth’s tectonic plates pushing and sliding against each other. Although their cause may be now known, and seismic monitoring well established, a recent TV Horizon documentary titled “ Why Can’t We Predict Earthquakes” (broadcast on SBS on January 24th 2011), lamented that earthquake prediction remains agonisingly elusive. Folklore links likely earthquakes to birds showing strange behaviour, well levels suddenly changing, flashes in the night sky, rainbow clouds, lunar-induced seismic tides, sunspots, ”earthquake weather”, gravity waves, radon gas seepage, the onset of headaches and even (predictably) conspiracy theories and electro-magnetic weapons. The danger that false alarms, arising siliconchip.com.au be predicted radio? By Stan Swan This seismogram, recorded by the McQueen’s Valley seismograph on Banks Peninsula (courtesy www.geonet.org.nz), displays the September 4 magnitude 7.1 Christchurch earthquake and some of its early aftershocks. The seismogram is coloured red if it is clipped, ie, the largest parts of the signal are not shown. If this was not done, then a large earthquake would obscure much of the seismogram from view – so if the signal is red, the real size is larger than shown. Lots of aftershocks can be seen on this image, with some of the biggest ones appearing out of the coda (dying away of shaking) of the main shock. from such incorrect quake alerts, may eventually desensitise people has to be considered as well. Studies have shown the pre-earthquake wandering of domestic animals, often dismissed as an urban myth, may indeed be a valid indicator. An recent Italian project showed striking toad breeding behavioural changes in the days before their (6.3) L’Aquila event of 6th April 2009. (At last – a possible use for cane toads?) What about radio signals? Radio propagation has long been related to solar activity and atmospherics, with the earth’s charged ionosphere an especially significant factor. In 1989 however submarine comsiliconchip.com.au munication monitoring first indicated seismic contributions as well, as significant signal surges were noted prior to California’s October 17th Loma Prieta earthquake that year. More recently, the French DEMETER (Detection of Electro-Magnetic Emissions Transmitted from Earthquake Regions) microsatellite (shown at right) detected ionospheric perturbations while passing over the September 29, 2009 Samoan and the January 12, 2010 Haitian earthquakes. DEMETER findings inApril 2011  15 dicate that shallow earthquakes, of magnitude 4.8 and larger occuring at night, show an associated decrease in natural EM radiation at around 3kHz. No changes were observed for deep quakes or for those that occur in the daytime. Although the disturbances were extremely small and only revealed statistically, this study demonstrated that seismic activity may also influence the ionosphere, with the effect perceptible at even the Low Earth Orbital (LEO) satellite altitude of some 700km before some earthquakes occur. P-wave earthquake alerts Fig.1 (above): the initial P-wave shock, akin to a longitudinal impulse along a slinky spring, travels at some 8 km/second but the more destructive sideways S-waves travel at only half this speed. Fig.2 (below): this sample seismogram reveals a ~38 second delay between them (relating to quite a distant earthquake) – perhaps just enough time to scramble to safety? Aside from EM alert possibilities, it’s actually already feasible to have very short advance warning of an earthquake by detecting the initial non-destructive P-wave (P = primary, push or pressure) compressions. These initial seismic waves travel more quickly through the Earth’s crust than the destructive “S” (secondary, shear or shake) transverse waves and subsequent rolling Rayleigh surface waves. P-waves, which are typically felt by humans as a bang or thump, travel at some 8km a second in dense earth (about twice the speed of S-waves), so advance warnings of perhaps seconds (for local rumbles) or up to about a minute (for deep or more distant quakes) may be possible. The effect is rather akin to seeing a lightning flash and hearing thunder some time later. A smoke alarm-sized P-wave seismic alert device already is widely marketed, especially in California. (See www.earthquakealert.com). As radio waves travel near instantly, it may be feasible to use such a device to “beat the P-wave” and radio ahead an alert by cell phone. Although time would be very precious indeed, even a few seconds warning may be enough to “Drop, Cover and Hold On”. Chile, which is on the Pacific’s Ring Of Fire, experienced a catastrophic 8.8 quake on 27th February 2010 and is considering such a country-wide alert system. The effectiveness of a prediction device relates to reliable P-wave detection, an acceptance that false alarms (arising from normal local ground vibrations) may occur and sufficient time to react. An alert of minutes would be much more valuable (even if less reliable), which is where very low frequency radio monitoring may assist. Natural radio and earthquakes The electromagnetic spectrum is full of transient natural RF signals, many often arising from the sun’s activity and the scientific jury remains out on just which, if any, are most related to earthquakes. The basis for the possible connection plausibly relates to stressed sub-surface rock layers generating voltages and signals in the manner of piezo crystals. Numerous accounts regarding the monitoring of different frequencies for broadband noise and static changes have been made, especially in such seismic regions as Italy – see www.nat- Spectran displays of useful VLF signals – at left is the 15.625 kHz horizontal scan of a colour TV while at right is the received signal from the Harold E Holt North West Cape (NWC) submarine communications station at Exmouth,WA. The latter is so powerful that it can usually be received throughout Australia even without a FET preamp! 16  Silicon Chip siliconchip.com.au Fig.3: the circuit of a wideband preamplifier suitable for use with a PC sound card. Because the frequencies of interest are so low, virtually any antenna you can put on this will theoretically be too short – so just use what you can! A good earth would also help greatly. About this feature: Fig.4: a suitable breadboard layout for the above preamp, drawn using "PEBBLE" (see SILICON CHIP, September 2009). hazards-earth-syst-sci.net/1/99/2001/ nhess-1-99-2001.pdf A review of the literature indicates that the most promising approaches for earthquake precursors may be: (1). Sub-Hertz and ELF magnetic transients in the .01 to 10Hz region, especially around 3Hz; (2). VLF electromagnetic transients around 10kHz; (3). VLF-LF broadband noise measurements in the 10 to 100kHz band; (4). LF-MF noise and propagation in the 150kHz to 2000kHz region; and (5). HF noise and propagation studies over 2MHz. For our purposes, monitoring the bands between 10kHz and 100kHz New Zealander Stan Swan is no stranger to SILICON CHIP readers, having written numerous articles over the years and is credited with introducing Australian and New Zealand readers to the PICAXE microcontroller. Although not a resident of Christchurch and therefore not directly affected, Stan contacted us not long after the September 2010 earthquake talking about some of the work he was doing in ultra-low-frequency radio earthquake “detection”. This article is the result, written and in fact in production as the news came through about the February 21 shake. And as we were about to go to press, on March 2 came the news that New Zealand had suffered yet another earthquake, fortunately (at 4.5) significantly lower in magnitude than Christchurch but this time located near the NZ capital city, Wellington – just across the harbour from Stan’s home in Eastbourne! Incidentally, in 1848 and 1855 Wellington suffered magnitude 7.1 and 8.0 earthquakes, the latter the largest ever recorded in New Zealand and causing considerable damage. are the most practical and initial investigation requires little more than a Windows PC with a working sound card – even an old clunker XP laptop with a 16-bit 48kHz sampling-rate soundcard will do nicely. But hang on – sound cards hearing radio? Yes – quite correct! Low Radio Frequencies (RF) produce a small signal that a sound card treats just like an equivalent audio signal from a microphone. And although sound ITU Abbrev. Designation Frequency Wavelength Typical EM signals 0 Sub-Hz 1 ELF 2 SLF 3 ULF 4 VLF 5 LF 6 MF 7 HF 8 VHF 9 UHF 10 SHF 11 EHF Sub Hertz Extremely Low Frequency Super Low Frequency Ultra Low Frequency Very Low Frequency Low Frequency Medium Frequency High Frequency Very High Frequency Ultra High frequency Super High Frequency Extremely High Frequency <3Hz 3Hz - 30Hz 30Hz - 300Hz 300Hz - 3kHz 3kHz - 30kHz 30kHz - 300kHz 300kHz - 3000kHz 3 MHz - 30MHz 30 MHz - 300MHz 300MHz - 3000MHz 3GHz - 30GHz 30GHz - 300GHz >100,000km 100,000km to 10,000km 10,000km to 1,000km 1,000km to 100km 100km to 10km 10km to 1km 1km to 100m 100m to 10m 10m to 1m 1m to 10cm 10cm to 1cm 1cm to 1mm Natural earth, ionosphere, space Deeply submerged submarines Sub. communication, mains grids Earth mode comms. – mine radio Near-surface sub. & cave radio. Long Wave radio, aircraft beacons Medium Wave AM broadcasting Short Wave radio, maritime, amateur FM radio, TV, aircraft & marine TV, cell phones, 2-way, WiFi, GPS Radar, satellite TV, microwave comms. Radio astronomy, microwave links The electromagnetic spectrum from "DC to Daylight" (well, almost). The bands/frequencies above 300kHz are pretty much understood but it's those below – and far below – which we are interested in here. siliconchip.com.au April 2011  17 The Richter Scale: not any more, it’s now the Moment Magnitude Scale The Richter magnitude scale, also known as the local magnitude (ML) scale, assigns a single number to quantify the amount of seismic energy released by an earthquake. It is a base-10 logarithmic scale obtained by calculating the logarithm of the combined horizontal amplitude (shaking amplitude) of the largest displacement from zero on a particular type of seismometer (Wood–Anderson torsion). So, for example, an earthquake that measures 5.0 on the Richter scale has a shaking amplitude 10 times larger than one that measures 4.0. The effective upper limit of measurement for local magnitude ML is just below 9 for local magnitudes and just below 10 for moment magnitude when applied to large earthquakes. The Richter scale has been superseded by the moment magnitude scale, which is calibrated to give generally similar values for medium-sized earthquakes (magnitudes between 3 and 7). Unlike the Richter scale, the moment magnitude scale reports a fundamental property of the earthquake derived from instrument data, rather than reporting instrument data which is not always comparable across earthquakes, and does not saturate in the high-magnitude range. Since the Moment Magnitude scale generally yields very similar results to the Richter scale, magnitudes of earthquakes reported in the mass media are usually reported without indicating which scale is being used. The energy release of an earthquake, which closely correlates to its destructive power, scales with the 3⁄2 power of the shaking amplitude. Thus, a difference in magnitude of 1.0 is equivalent to a factor of 31.6 ( = (101.0)(3 / 2)) in the energy released; a difference in magnitude of 2.0 is equivalent to a factor of 1000 ( = (102.0)(3 / 2)) in the energy released. Richter Description Earthquake effects magnitudes Less than 2.0 Micro 2.0–2.9 Minor 3.0–3.9 Minor 4.0–4.9 Light 5.0–5.9 Moderate 6.0–6.9 Strong 7.0–7.9 Major 8.0–8.9 Great 9.0–9.9 Great 10.0+ Epic Frequency of occurrence Micro earthquakes, not felt. About 8,000 per day Generally not felt, but recorded. About 1,000 per day Often felt, but rarely causes damage. 49,000 per year (est.) Noticeable shaking of indoor items, rattling noises. 6,200 per year (est.) Significant damage unlikely. Can cause major damage to poorly constructed buildings over 800 per year small regions. At most slight damage to well-designed buildings. Can be destructive in areas up to about 160km (100 miles) 120 per year across in populated areas. Can cause serious damage over larger areas. 18 per year Can cause serious damage in areas several hundred miles across. 1 per year Devastating in areas several thousand miles across. 1 per 20 years Never recorded in human history. Extremely rare (unknown) Courtesy Wikipedia – http://en.wikipedia.org/wiki/Richter_magnitude_scale cards typically handle signals up to around 24kHz this is quite adequate for this purpose. Just to clear up a confusion which often occurs: why can't you hear lowfrequency radio signals? Even though they may be in the audio frequency range, you can not hear low-frequency radio waves as they're an electrical rather than acoustic phenomena. Your ears cannot "detect" radio signals. Software Many specialised and complicated panoramic display sound card Windows programs are freely available, but it’s recommended you start with simple ones to get a feel for things. The tiny SAQrx (https://sites.google. com/site/sm6lkm/saqrx) should cope well, although Spectran (www.weaksignals.com) is better suited for more demanding work. 18  Silicon Chip Once installed verify operation by whistling into the computer’s mike to observe the resulting spectrogram. A valuable waterfall display option (plotting frequency versus time) is included in Spectran – it can be set to scroll sideways with a left mouse button click. This waterfall can be a visual goldmine when following transient signals, as they remain on screen long after they’ve ceased. It can be fascinating to “see” the spectrum of such everyday sounds as music, speech and bird calls! An averaging option further allows masking out of random noise to better show weak transmissions, and recordings can be saved to hard disk. More professional soundcard display offerings, especially the Spectrum Lab (http://dl4yhf.ssl7.com/spectra1. html), may suit once you are familiar with the panoramic technology. Apart from the PC, the only other hardware required to initially “hear” the nearby EM spectrum is a suitable 3.5mm phone plug (usually mono) and a short length of wire! Wavelengths at VLF are so long almost any handy length of insulated wire will do. Run the wire vertically if possible, and ensure it doesn’t snag or short to anything lively or your computer sound card may be damaged. Wind it in during any likely thunderstorms as well, at it could present a hazard to you and your computer. Performance In a typical built-up area mains noises (50Hz and harmonics) will promptly show themselves to indicate “receiver” operation but a more useful beacon can be the 15.625kHz horizontal scan oscillator of a PAL colour CRT TV set. This can usually be detected from many metres away. Assorted spusiliconchip.com.au rious signals may also be seen arising from normal PC operation. Removing the sound card input plug will verify the true nature of such “ghosts”. My urban location here in remote NZ meant other VLF transmissions were initially only weakly detected, although these were revealed better in quieter areas using a battery powered laptop and long wire antenna positioned well away from mains wiring. Readers living closer to powerful VLF submarine stations may find even a short hookup wire antenna will do! Other low frequency “noise” Aside from man-made noise, VLF reception may be further complicated by day/night variations and extensive tropical storms. Across the globe, lightning strikes almost continually (refer the World Wide Lightning Locator Network at http://webflash.ess.washington.edu) and it’s long been known that the violent electrical discharge may also even propagate upwards from storm clouds and influence the ionosphere. So-called “Schumann resonances” may then arise, caused by a lightning excited ~8Hz resonance in the waveguide cavity formed by the earth’s surface and the ionosphere. The bouncing EM pulses associated with such powerful lightning pulses and the resulting atmospherics (“spherics”) may propagate globally on low RF frequencies, to be heard as static crashes and even gliding frequency whistles and chirps. An auroral display may also produce such effects – listening to such atmospheric music can be part of the fun! Enhancement Given the very low frequency nature of the signals almost any simple preamplifier may be used to boost the input to the sound card. A complete receiver (such as the well known BBB-4 – http://www.auroralchorus.com/bbb4rx3.htm) could even be constructed for standalone listening, but this would not lack the panoramic display and recording features that PC monitoring allows. As broad-band low frequency boosting is needed, tuned circuits are not even utilised, although a suitable low pass filter may be needed to block any nearby powerful AM radio stations. After considering various low-noise opamps, a simple general-purpose siliconchip.com.au A typical published account Title: Geomagnetic precursors of intensive earthquakes in the 1-0.-2Hz frequency range of geomagnetic pulsations – Abstract Only Corporate Source: Joint Publications Research Service, Arlington, VA. During intensive geo-tectonic processes such as earthquakes, pulsations are observed in the geomagnetic field at a frequency of 0.02 to 1Hz with anomalously high amplitudes. These pulsations usually appear as beat phenomena lasting from several minutes to several hours. It has been found that the pulsations are excited only in magnetic components of the terrestrial electromagnetic field. The periods and amplitudes of the pulsations are nonlinearly related to the intensity of the earthquakes. Pulsations of this type are not observed when earthquakes do not occur. Additional analysis shows that frequently the pulsations precede intensive earthquakes by 10 to 200 minutes, then drop for about 1 hour, then appear once again during the actual earthquake. Oscillograms of such pulsations are presented. The periods and amplitudes of the geomagnetic pulsations preceding earthquakes are found to be linearly related to the magnitude of the earthquakes. A regression equation relating earthquake magnitude to pulsation characteristics is presented. Author: GOGATISHVILI, Y. M. CASI Accession No. 85N23178 Published: February 1985 (referenced at www.manuka.orcon.net.nz/eradio.htm) MPF102 N-channel J-FET was eventually used – see Fig.3. Layout and component values are not critical. This setup, powered by a 9V battery performed very well, revealing signals that were previously buried in the noise. The two back-to-back 3.3V zener diodes ensure any larger voltages on the antenna will be shorted to earth. The circuit draws around 4mA and can be easily assembled and enclosed in a small metal case to give shielding. Leads to the sound card should be shielded to reduce mains pick-up. Suggested monitoring approach Launching VLF monitoring satellites or erecting gigantic antenna farms, such as the military use, is naturally a tad daunting. Powerful VLF submarine communication transmitters conveniently already blanket the world so it’s suggested that initial monitoring merely follows the approach of simply checking their VLF signal strengths over an extended period. However, sudden variations may well arise due to solar storms (see www.swpc.noaa.gov) or the VLF site’s transmitting activity. Northwest Cape wasn’t set up just for your listening pleasure! Once the sound card-based equipment is organised at your location, use a stable set up and antenna so any on-screen changes will be noticeable. To help gain initial experience perhaps refer to your displays (and recordings?) when an earthquake has occurred somewhere, to see if unusual VLF activity was associated with transmitters near it. Details of the world’s latest earthquakes are soberingly shown at http:// earthquake.usgs.gov/earthquakes/ recenteqsww/ Conclusion Don’t expect instant answers in the VLF monitoring quest, as display checking may be akin to watching paint dry. Opinions may differ and findings are uncertain but seismic scientists worldwide earnestly scrutinise such displays in attempts to see if seismic and low-frequency radio signals act as possible earthquake precursors. It may well be a false quest, with no more pre rumble significance than the birds going quiet, your dog hiding under the bed – or the cane toads pausing their advances. But there just may be something in it – and your simple setup could help provide a valuable key or stimulate further investigations! References, extensions, published scientific studies and quoted web sites are conveniently linked via a resource site at www.manuka.orcon.net.nz/ eqradio.htm SC April 2011  19