This comment by Dr. Friedemann Freund was published in the 18 May 1999 issue of EOS, Vol. 80, No. 20.

Comment: Earthquake Prediction Is Worthy of Study

Imagine a densely populated region in the contiguous United States haunted over the past 25 years by nine big earthquakes of magnitudes 5.5 to 7.8 killing hundreds of thousands of people. Imagine further that in a singularly glorious instance a daring prediction effort, based on some scientifically poorly understood natural phenomena, led to the evacuation of a major city just 13 hours before an M = 7.8 earthquake hit. None of the inhabitants of the evacuated city died, while in the surrounding, non-evacuated communities 240,000 were killed and about 600,000 seriously injured. Imagine at last that, tragically, the prediction of the next earthquake of a similar magnitude failed, as well as the following one, at great loss of life.

If this were an American scenario, the scientific community and the public at large would buzz with the glory of that one successful, life-saving earthquake prediction effort and with praise for American ingenuity. The fact that the next predictions failed would likely have energized the public, the political bodies, the scientists, and the funding agencies alike to go after a recalcitrant Earth, to poke into her deep secrets with all means at the scientists' disposal, and to retrieve even the faintest signals that our restless planet may send out prior to unleashing her deadly punches.

However, this is not an American scenario. The San Andreas fault is bad enough but not that bad, and by virtue of its being a transform fault, it is different from the situations in other parts of the world where enormous earthquakes are by-products of major subduction zones. Maybe because it's not an American scenario could we read "A Misuse of Public Funds: U.N. Support for Geomagnetic Forecasting of Earthquakes and Meteorological Disasters" (Eos, September 29, 1998), a Forum piece by Wallace Campbell, renowned seismologist, which opens with the sentence: "The legitimate scientific community needs to be alerted to the expenditure of considerable public funds for pseudoscientific projects that build false hopes of protection from geophysical hazards."

The maligned document is a report from Beijing, China, titled Manual on the Forecasting of Natural Disasters: Geomagnetic Methods. The authors are five Chinese scientists, Xiaoping Zeng, Yunfang Lin, Chunrong Xu, Ming Zhao, and Yuechen Zhao, while Jeanne-Marie Col and Arthur N. Holcombe are United Nations officials who served as facilitators with Jean J. Chu, a Chinese American, also at the United Nations, to bring a little-known sector of Chinese endeavor to the attention of the English-reading West.

"Fuzzy thinking . . . pseudoscientific nonsense," claims Campbell, who accuses the United Nations of having squandered public funds. If this accusation holds, it could lead to disciplinary action against those who are alleged to have committed such malfeasance. A couple of e-mails and phone calls to the United Nations helped me to shed light into the alleged financial morass. According to the information provided, the only funds that the United Nations used to organize the workshop in Beijing of which this manual is an product were donations obtained specifically for this purpose, outside the United Nations budget. While the Forum is meant to foster scientific discourse and to provide room for personal opinions, it is unfortunate that Campbell neglected to check the facts. Lurking behind his angry words is an enmity that many seismologists seem to harbor toward those who dare think about earthquake prediction. For them, earthquakes are to be objectively and dispassionately dissected. They proudly point out how much seismology has contributed to our knowledge about the spasms of the Earth. Trying to predict such spasms is a dangerous game. Any ever-so-slightly optimistic assessment of mankind's chances to forecast such natural disasters may raise false hopes. Once the public gets fired up, the science can get easily burned.

Two questions arise. Does such harsh judgment apply to the authors of the manual from Beijing? Is it helpful to discredit an entire community of scientists on the basis of the alleged wrongdoing of a few?

Addressing the second question first, I would like to point out that, even if there are some whose good intentions may be questionable, there are always others whose motives are altruistic. They just want to help—especially when tens or hundreds of thousands of lives may be at stake. They go into a controversial field such as earthquake prediction out of love for the unknown. They accept the challenge that nature throws at us. They know their odds are not good, but they do not give up. Uncertainty of success does not deter them. The efforts of these individuals do not deserve to be labeled illegitimate.

It is here where I disagree with Campbell. As Ziman [1998] puts it in a recent review of Gordon Moran's book, Silencing Scientists and Scholars in Other Fields, "Knowledge is created as much by heated argument as by ice-cold experimentation. The norms and practices of scientific communication pit researchers verbally against one another, but strictly limit their rhetorical weapons." Campbell uses innuendoes to discredit the interdisciplinary search for the subtle signals by which the Earth may divulge an impending disaster.

I am in a somewhat favorable position to comment on the clash of the minds. I have no prior involvement in earthquake prediction research, but for a few years I have followed the relevant literature. I attended a National Science Foundation-sponsored workshop on earthquake prediction in Southern California, convened close to where the M = 7.3 Landers earthquake would hit on June 28, 1992, just 11 days later. All of the participants dispersed without the faintest premonition of what would occur—all except two, who had said that, on the basis of their data, a fairly large earthquake in this region might be imminent. I followed the proceedings of a workshop in Tokyo in 1993 on "Electromagnetic Phenomena Related to Earthquake Prediction" and attended a workshop of the same kind in 1997. I listened to speakers from different countries, including America, who brave the ire of some big name seismologists. I came away with the feeling that those lone fighters are a dedicated bunch, sincere and mostly capable, sometimes scared by the antagonism thrown at them, often defiant, and always thankful to whoever lends them their ear.

The Earth's utterings are faint and often confusing and few claim to know how to read them. Yet radio communication researchers have long observed that the ionosphere can be disturbed for weeks over regions where an earthquake will eventually hit. It is impressive to read the careful analysis of radiowave transmission data that confirm, after the fact, that the deadly M = 7.1 Kobe earthquake of January 17, 1995, was accompanied by an ionospheric oscillation that began a few days before and decayed over a few days after the quake [Molchanov et al., 1998]. It is impressive to view the ground data from Japan and China, which suggest faint geoelectric and geomagnetic anomalies associated with impending earthquake activity. It is puzzling to listen to reports of apparent ground temperature anomalies extracted from satellite data over remote areas in China but unconfirmed by ground expeditions. Many geoelectric and geomagnetic observations come out of the countries of the former Soviet Union [Gokhberg et al., 1987]. There are also millivolt ground potential differences on which the VAN group (so called after the initials of the three inventors Varotsos, Alexopoulos, and Nomicos) in Greece [Varotsos et al., 1993] has built a system of hotly debated earthquake prediction [Masood, 1995].

According to Li Hui and Mervis [1996], China's earthquake prediction program is "arguably the most effective in the world despite an elusive and very modest definition of success." The foreword to the manual likewise acknowledges that "prediction of earthquakes by geomagnetic methods is still in the exploratory stage of data acquisition. . . . Although we have had [in China] more than ten fairly successful predictions, none can be said to be accurate in the strict sense." Should we not be thankful to the United Nations for having sponsored a document that allows interested scientists in the West to get a glimpse of Chinese thinking of which we still know so little? "The Chinese are more comfortable than Western scientists with conflicting and contradictory evidence," says L. Comfort, quoted by Li Hui and Mervis [1996], ". . . they just broaden the scope of their inquiry and collect more data."

Pseudoscience? I think this expression used by Campbell is derogatory. The central issue is our lack of understanding of the physical processes leading to electric and electromagnetic (EM) signals emitted from the ground. They require strong currents. The manual speaks with bewilderment to the issues at hand, and so does the community of earthquake researchers. As an outsider, I have the privilege to approach seemingly intractable questions from a different and possibly useful fresh perspective. For some years, I have been interested in defects in minerals from a solid state physics viewpoint. We identified defects that had been entirely overlooked in the past. We generically call them "peroxy" because they arise from lattice oxygen oxidized from its normal O2- state to the O- state. Two O- form a peroxy link, O- – O- or Si/O-O-\Si. Peroxy links derive from small amounts of H2O incorporated as Si-OH into the matrix of nominally anhydrous minerals when crystallizing in an H2O-laden environment [Freund et al.,1993]. Upon cooling below 500C, pairs of Si-OH seem to undergo a redox conversion to molecular H2 plus peroxy [Freund, 1987]. Since most rock-forming minerals are nominally anhydrous and since all geological environments are H2O-laden, all igneous rocks in the Earth's crust will most likely contain peroxy.

To understand the next step requires a brief excursion into semiconductor physics. An O- in an O2- matrix represents a defect electron or "positive hole" [King and Freund, 1984]. Two positive holes that tie a peroxy bond represent a positive hole pair (PHP), electrically dormant. When a PHP awakens, it generates a highly mobile charge [Freund et al., 1993]. While we had known for some time that PHP split upon heating, we noted with surprise that PHP also split when shocked—not necessarily a powerful shock but a simple acoustic wave such as emitted by brittle fracture. We therefore conducted low velocity impact experiments [Borucki and Freund, 1999] using cylindrical rock cores 20 mm in diameter and up to 700 mm in length. We first worked with gabbro, a quartz-free rock. As depicted in the inset in Figure 1, we equipped gabbro cores with a ring capacitor around the front end, a photodiode to detect light emission, and a plate capacitor at the end.

Figure 1 shows the signals recorded upon a 90 m/s impact. The moment the 3.2-mm steel ball hits the rock is marked by a weak light blip. About 100 microsec. after impact, a positive charge arrives at the front ring capacitor. About 150 microsec. after impact, the positive charge arrives at the back plate capacitor. As the front ring capacitor reaches 450 mV, a burst of light is emitted from the front end, often visible with the naked eye. Positive capacitor voltages indicate positive charge carriers, consistent with positive holes. The delayed light burst is due to a corona discharge, caused by the high electric field, up to 106 V/cm, that is predicted to build up at the rock surface [King and Freund, 1984].

When the impact is directed onto the side of a rectangular block, a surprisingly strong current can be sent through, indicating that the propagating charge cloud causes the rock to become a conductor, albeit only for a short time. When the energy of impact is increased, the shocks become more powerful. Using granite and 6.3-mm aluminum projectiles traveling at 1-2 km/s, the propagation of the S and P waves can be "seen" through the piezoelectric response of quartz crystals.

Figure 2 shows results obtained with a setup sketched in the inserts. Figure 2a depicts from top to bottom the EM emission and the positive voltage on the rock surface in the plane of impact. Figure 2b depicts from top to bottom the positive potentials recorded by capacitive sensors and the voltage on the rock surface near the base plane, modulated by current injections. Prominent is the charge that engulfs the entire rock, building up over about 1 ms after impact and fading over the 2-3 ms.

The intriguing first insight gained from such experiments is that acoustic waves generate charges in a common igneous rock. These charges are positive and emit EM signals as they propagate. After each shock, the charge carriers remain "alive" for a few milliseconds, but every new shock reactivates them anew.

It is too early to draw conclusions from these experiments as far as earthquake prediction is concerned, but the discovery of highly mobile positive holes in common igneous rocks opens a window from which we can take a fresh look at some of the enigmatic electric and electromagnetic signals coming out of the Earth around earthquake time.

Pseudoscientific nonsense? Wishing to understand the signals that come from beneath our feet remains an important question of our time, inside and outside the United States.


Friedemann Freund, SETI Institute, at the NASA Ames Research Center, MS 239-4, Moffett Field, CA 94035-1000, USA. email:

For an updated discussion of the author's research as related to earthquake prediction, see:
Freund, F., Time-Resolved Study of Charge Generation and Propagation in Igneous Rocks
J. Geophys. Res. 105: 11,001-11,019 (2000).


Gabbro core, instrumented as depicted in the inset and impacted with a 3.2 mm steel ball at 90 m/sec. Time of impact marked by vertical arrow.
Channel 2: Front ring capacitor voltage;
Channel 3: Back plane capacitor voltage;
Channel 4: Front light emission



Granite block, 25 x 25 x 20 cm, grounded at the bottom and instrumented as depicted in the insets with three magnetic pick\endash up coils, two electrodes at the plane of impact and near the bottom, and three capacitors, impacted with a 6.3 mm aluminum ball at 1.46 km/sec (arrow)


Freund, F., and J. G. Borucki, Charge carrier generation and charge cloud propagation following 100 m/sec impacts on igneous rocks, in Atmospheric and Ionospheric Electromagnetic Phenomena Associated with Earthquakes, edited by M. Hayakawa, pp. 839-857, Terra Sci., Tokyo, 1999.

Freund, F., Hydrogen and carbon in solid solution in oxides and silicates. Phys. Chem. Miner., 15, 1-18, 1987.

Freund, F., M. M. Freund, and F. Batllo, Critical review of electrical conductivity measurements and charge distribution analysis of MgO, J.

Geophys. Res., 98, 22,209-22,229, 1993. Gokhberg, M. B., I. L. Gulfeld, and V. I. Liperovsky, Electromagnetic precursors in the earthquake system: search problem, Vestnik Acad. Nauk USSR, N3, 45-53, 1987.

King, B. V., and F. Freund, Surface charges and subsurface space charge distribution in MgO containing dissolved traces of water, Phys. Rev. B Condens. Matter, 29, 5814-5824, 1984.

Li Hui, and J. Mervis, China's campaign to predict earthquakes, Science, 273, 1484-1486, 1996.

Masood, E., Greek earthquake stirs controversy over claims for prediction, Nature, 375, 617, 1995.

Molchanov, O. A., M. Hayakawa, T. Oudoh, and E. Kawai, Precursory effects in the subionospheric VLF signals for the Kobe earthquake, Phys. Earth Planet. Inter., 105, 239-248, 1998. (This reference is missed in Eos 18 May 1999 issue.)

Varotsos, P., K. Alexopoulos, and M. Lazaridou, Latest aspects of earthquake prediction in Greece based on seismic electric signals, II, Tectonophysics, 224, 1-37, 1993.

Ziman, J., Silencing scientists and scholars in other fields: Power, paradigm controls, peer review and scholarly communication, Nature, 395, 856-857, 1998.