Rocks That Crackle and Sparkle and Glow: Strange Pre-Earthquake Phenomena
| Scientific Paper | |
|---|---|
| Title | Rocks That Crackle and Sparkle and Glow: Strange Pre-Earthquake Phenomena |
| Read in full | Link to paper |
| Author(s) | Friedmann T Freund |
| Keywords | rocks, electric charge, magnetic field, electromagnetic, earthquakes |
| Published | 2003 |
| Journal | Journal of Scientific Exploration |
| Volume | 17 |
| Number | 1 |
| No. of pages | 35 |
| Pages | 37-71 |
Read the full paper here
Abstract
Seismic waves are the most dramatic and most intensely studied manifestations of earthquakes. However, we also know of non-seismic phenomena, which precede large earthquakes. Some of them have been reported for centuries, even millennia. The list is long and diverse: bulging of the Earth's surface, changing well water levels, ground-hugging fog, low frequency electromagnetic emission, earthquake lights from ridges and mountain tops, magnetic field anomalies up to 0.5% of the Earth's dipole field, temperature anomalies by several degrees over wide areas as seen in satellite images, changes in the plasma density of the ionosphere, and strange animal behavior. Because it seems nearly impossible to imagine that such diverse phenomena could have a common physical cause, there is great confusion and even greater controversy. This explains why reports on nonseismic pre-earthquake phenomena are regarded with suspicion in the scientific community. This may change with the recent discovery that igneous and metamorphic rocks, which make up a major portion of the Earth's crust, contain electric charge carriers, which have been overlooked in the past. These charge carriers are defect electrons in the valence band, i.e., positive holes. Under normal conditions they are dormant, but when they "wake up", the rocks begin to sparkle and glow. This paper describes the physical and chemical nature of these positive holes, how they are introduced into minerals and rocks, and how they become activated. Evidence will be presented that, once the positive holes are generated, currents propagate through the rocks leading to electromagnetic emission, to positive surface potentials, to corona discharges, to positive ion emission, and to mid-infrared radiation. These phenomena are expressions of the same fundamental process: the "awakening" of dormant positive hole charge carriers that turn rocks momentarily into p-type semiconductors.
Overview
Friedmann T Freund — writing from San Jose State University and NASA Ames — sets out to rescue an entire category of observation that mainstream seismology has largely written off. Reports of earthquake lights, ground-hugging fog, magnetic anomalies, satellite-detected thermal anomalies, ionospheric disturbances and strange animal behaviour before large quakes have accumulated for centuries from every tectonically active region on Earth. Freund identifies three reasons why geophysicists remain sceptical: the phenomena are too varied to plausibly share a cause; some earthquakes produce them while comparable ones do not; and no physical process is known that could produce even one of them, let alone all. His methodological starting point is explicitly contrarian: "if phenomena have been reported so many times by a wide range of observers across centuries and from different parts of the world, they must hold some truth," and the failure to explain them may mean the missing piece is in our knowledge rather than in the observations.
The proposed missing piece is a charge carrier. Freund argues that igneous and high-grade metamorphic rocks contain peroxy defects — he calls an intact O–O bond or O3Si/OO\SiO3 link a "positive hole pair" (PHP) — which are electrically dormant but which, when broken by stress, release mobile defect electrons in the oxygen 2p valence band. These "p-holes" momentarily turn an insulating rock into a p-type semiconductor. From this single mechanism he derives ground currents, low-frequency electromagnetic emission, magnetic anomalies, positive surface potentials, corona discharges, positive ion emission and non-thermal infrared. The departure from the mainstream account is twofold: it claims a charge carrier that standard rock-conductivity measurements have been systematically destroying for decades, and it treats the anomalous-precursor literature as data rather than folklore.
The argument
The conductivity conundrum
Several of the reported precursors require large, time-varying currents in the ground as a sine qua non. Only two mechanisms have well-understood physics: piezoelectricity and streaming potentials. Freund dismisses both. Of the rock-forming minerals only quartz is piezoelectric, and in a batholithic granite the crystals are randomly oriented, so local voltages cancel and no long-range field results. Streaming potentials in porous rock are capped by the ionic conductivity of the pore brine, giving "as low as a few millivolts over distances as large as a kilometer." Meanwhile the laboratory literature holds that oxide and silicate minerals are essentially non-conducting until ionic conduction switches on around 700–800 °C, or until partial melting. Hence the dilemma: field observations demand currents that no accepted mechanism can supply, and "when a set of observations cannot be explained within the framework of existing knowledge, the tendency is not to believe the observation."
An artefact of experimental design
Freund's key move is to challenge the measurements themselves. Conductivity experiments on insulating crystals are plagued by surface currents that swamp the bulk signal; the standard remedy is a guard electrode plus prolonged high-temperature pretreatment to burn off what is assumed to be "dirt." He asks whether that "dirt" conductivity was the real signal all along, and whether decades of pretreatment have been destroying "an important piece of information that is germane to the samples under study."
The theoretical case comes from King and Freund (1984): when p-holes are activated in a sample of dielectric constant ε surrounded by a medium of lower dielectric constant — vacuum or air — they redistribute into a thin surface charge layer. Since any laboratory measurement requires removing the rock from the surrounding medium of matched dielectric constant in which it sat inside the Earth, the surface charge layer is not contamination but an unavoidable consequence of the measurement geometry.
Where the peroxy defects come from
The chemistry traces to a 1970s anomaly. Finely divided MgO made by decomposing ultra-pure Mg(OH)2 released not water but molecular hydrogen, in amounts of order 1–2% of the oxygen present — far too much to attribute to a sample with under 5 ppm cation impurity. The only available electron donor was oxygen itself: 2 OH− → O22− + H2, with two oxygens going from the 2− to the 1− state and forming a peroxy bond. Later work found the same conversion in MgO single crystals grown from water-bearing melts, and then in olivine, feldspar, garnet, dunite, andesite and diorite. Since any oxide or silicate crystallizing in an H2O-laden environment incorporates some OH−, Freund concludes that PHPs are probably present in most igneous and high-grade metamorphic rocks. He notes that reviewers rejected this on the grounds that peroxy species are highly oxidized and cannot exist in reducing environments — an objection he calls fallacious because the conversion occurs around 400–500 °C during cooling, where the mineral is locked into a metastable rather than an equilibrium state.
Waking the p-holes: impact and load
Fracturing MgO in front of a fast mass spectrometer produced atomic oxygen within microseconds, suggesting activation by the acoustic shock itself. That led to impact experiments on rocks, instrumented with capacitive sensors, induction coils, contact electrodes and photodiodes. Low-velocity impacts (~100 m/s steel spheres) produced a positive charge arriving at a ring sensor about 100 μs after impact and at a plate sensor ~50 μs later, giving p-hole propagation speeds of 100–300 m/s — in agreement with the theoretical maximum set by phonon-frequency electron hopping (~1012 Hz over ~3 Å jumps). As the surface potential approached the predicted 400 mV, a photodiode recorded a major flash from the rim, accompanied by an EM spike: a corona discharge, as predicted from the ~400,000 V/cm field implied by a 10–100 nm charge layer. In a diorite block under 25 V bias, the impact was followed ~150 μs later by the onset of current — the insulating rock had become conductive. Medium-velocity impacts (1.5 km/s into granite) activated p-holes throughout the whole block, with the charge cloud arriving after the P and S waves had died away.
Static loading gives the same result. A granite block loaded only in its central cylinder developed a positive surface potential rising with load to +1.7 V at failure, measured 7.5–15 cm from the stressed volume — so the carriers flow out of the source region into unstressed rock. An ion collector biased at −47 V drew 40–80 pA, indicating positive ions emitted from the surface. Freund maps the p-hole generation rate onto the dislocation-generation curve: it begins at the end of the elastic range, peaks during plastic flow, and falls off as dislocations coalesce into microcracks before failure.
Non-thermal infrared
Satellite thermal anomalies of 2–3 °C over tens to hundreds of kilometres, appearing 5–10 days before an event and vanishing 1–2 days after, cannot be real heating: the energy to warm that rock volume by 1 °C would exceed the total release of an M ≈ 7 earthquake, and thermal conduction could not heat and cool it in days. Using a QWIP camera at 8.3 μm on granite in a 1500-ton press, Freund recorded infrared excess beginning above 20 bar, rising to the equivalent of +0.5 K at 4σ, and emitted from the unloaded half of a block 5 cm from the deformed region — far beyond diffusive heat transport in the minutes available. He attributes it to p-hole recombination at the surface.
Broad-brush consequences
Outflowing p-hole clouds are self-limiting, since the source volume charges negatively and the surroundings positively; a backflow of protons from pore-water electrolysis or electrons from ground follows, and two coupled counter-currents of unequal mobility can oscillate. Freund reads the 1999 Taiwan magnetometer record this way: station LY, 20 km from the Chi-Chi rupture, showed 100–200 nT pulses — nearly 0.5% of the dipole field — beginning about two months before the M 7.7 event and continuing to the M 7.1 Chai-Yi aftershock, with asymmetric pulses of several hours' duration, while Pacific coast stations used for subtraction showed nothing comparable. The same currents radiate in the ULF band, offering a link to the 0.007–0.013 Hz noise increase recorded in Japan over 4000 km away. Corona discharges at convex topography account for earthquake lights, including the moving "sheet of light" reported 19 km from the 1988–89 Saguenay epicentre, and for the high-frequency static noted by radio operators before the 1960 Chile and 1964 Alaska events; ball lightning is offered as a possible detached plasma volume. Emitted positive ions supply condensation nuclei for pre-quake fog, and — since positive air ions are associated with headache, nausea and irritability, as with Foehn, Chinook and Santa Ana winds — a physiological route to the animal-behaviour reports that Milne's textbook dismissed as impossible to accept. Finally, a raised positive ground potential coupling to the ionospheric F layer as one plate of a planetary capacitor is offered for the ionospheric perturbations.
Assessment
The paper's real strength is that it is an experimental paper, not a speculative one. Freund is not merely proposing that a unifying mechanism might exist; he identifies a specific solid-state defect, gives its chemical origin, predicts its carrier velocity from phonon frequencies, predicts the surface potential and field strength from dielectric theory, and then measures all of them. The predictions and the measurements agree: 100–300 m/s carrier speed, ~400 mV surface potential, corona discharge at the rim, current onset in a nominally insulating diorite, and infrared emission from rock that is demonstrably not hot. The observation that infrared and surface charge appear centimetres away from the deformed volume, faster than heat could diffuse, is a genuinely difficult result to explain any other way, and the critique of conductivity methodology — that decades of high-temperature pretreatment may have annealed away the very carriers being sought — is a serious methodological point rather than special pleading.
The weaknesses lie almost entirely in the extrapolation. The laboratory work spans centimetres, seconds and megapascals; the phenomena to be explained span hundreds of kilometres, weeks and the whole crust, and the scaling is asserted rather than demonstrated. Freund concedes as much — "we have not yet conducted experiments to measure such currents directly," "details of this scenario are of course still highly uncertain," the solid-state plasma outbreak is "still speculative" — but the discussion section nonetheless proceeds to claim a great deal of territory on that basis. The self-limiting-current argument cuts against the model more sharply than the paper allows: if outflow generates a countervailing field that stops it, generating the sustained multi-week current oscillations inferred from the Taiwan magnetometer data requires a backflow mechanism that is named (protons, ground electrons) but not quantified. Nothing in the paper estimates whether the p-hole concentration in a realistic crustal volume can deliver 100–200 nT at 20 km, which is the single number that would make or break the geophysical claim.
Several links in the chain are correlational or anecdotal in a way the rest of the paper is not. The Taiwan magnetic data are unpublished, courtesy of a colleague, and are a single sequence; establishing that such pulses precede earthquakes rather than merely occurring near them requires the false-alarm rate, which is not given — and this is precisely the statistical objection that sank earlier precursor claims (the Geller and Turcotte critiques Freund cites in passing but does not answer). The animal-behaviour section rests on an untested chain: p-holes to surface charge to ion emission to airborne positive ions to physiological distress to observable behaviour, each step plausible and none measured together. The ball lightning material adds nothing testable. And the ionospheric coupling invokes field magnitudes over 100 km that the paper itself flags as surprising without resolving.
The closing pages shift from physics to sociology — peer review, mainstream editors, Schopenhauer's three stages of truth — and this is the paper at its least persuasive. The physical evidence assembled here does not need that framing and is weakened by it. Taken narrowly, on its laboratory results, the case that stressed rocks release mobile positive-hole carriers is strong and was worth pressing; taken broadly, as a unified account of every reported precursor, it remains a programme rather than a result. Freund is largely honest about which is which, and the two decades of subsequent work on stress-activated charge carriers suggest the narrow claim has held up better than the sceptics of 2003 expected.