by Gennady SOBOLEV, RAS Corresponding Member, Joint Institute of Physics of the Earth named after Otto Schmidt
Among elements earthquakes are, perhaps, the most dreadful and dangerous. Predicting them is a challenge for science. To meet it, different methods are used, including laboratory modeling which yields encouraging results.
Lithosphere-the planet's rock shell-sometimes gives gigantic cracks. The moving edges of such rupture produce tough waves. Traveling at a speed of a few kilometers per second, they quickly reach the surface with often catastrophic consequences.
This is the mechanism of an earthquake. According to geological and seismological data, an earthquake is a result of a sudden outburst of dynamic tensions caused by processes going on in the bosom of the earth (matter differentiation, convective rock mass flows, etc.).
Occasionally, the length of a rupture appearing on the surface from the bowels reaches hundreds of kilometers, while vertical and horizontal displacements of its edges-a few meters. For example, after the Spitak earthquake of December 7, 1988, a cliff up to 2 m high appeared in 8 km from the completely destroyed town. However, just a decade later only an expert could note traces of the past shift: the earth is fast to heal wounds. And people build new houses in places once swept by "a subsoil hurricane".
Seismic shocks are extremely swift-passing, although produce disastrous effects: towns and settlements get annihilated as a matter of seconds, hundreds and thousands of people are killed (and sometimes as many as a million!). Projecting the statistics of earthquake victims in the 20th century into the near future we can reach a dismaying conclusion: in the current decade earthquakes may affect at least 300 - 350 thousand people, and the damage will amount to $100 - 150 billion.
In Russia the seismically hazardous regions are, primarily, Kamchatka, the Kurile Islands, Sakhalin, the Baikal Basin, the Baikal-Amur Railroad area and Northern Caucasia * . Within the same decade thousands of people in our country will be exposed to the danger, while the economic loss is likely to approach R300 billion. These mind-boggling figures may be substantially reduced if we are able to cope with the problem of a short-term (as a matter of days or hours) earthquake forecast and maximally accurate pinpointing their epicenters.
It is a common knowledge that strong shocks are preceded by a series of weak ones, as well as anomalies of electric and magnetic fields, discharges of radon, helium and of other gases. A slump or, conversely, an upsurge of the groundwater level is registered, domestic and wild animals get nervous or even go completely amok.
However, such foreshocks are not reproduced in all cases or in any locality, their combination and duration are rather volatile, and the physical nature still contains many mysteries. That is why it is so important to comprehend the laws shaping up the focus of underground shocks. Observing the dynamics of natural earthquakes is clearly insufficient: each time they happen in new geological conditions, at a different, and often considerable, depth, where the conclusion about rock properties is largely the realm of guess based on circumstantial evidence. Full-fledged preliminary study is not always possible either: after all, in most cases a calamity catches us unawares.
No doubt, a lot of questions can be answered by a strict physical experiment. But it presupposes recurrence of experiment under controlled testing conditions. And that is attainable in the laboratory only, i.e., studying tough waves originating in rock samples at ruptures, similar in parameters to seismic vibrations at earthquakes. To be certain, an experimental model is deliberately simplified as opposed to the extremely complex lithospheric mass, therefore the laws discovered in the simulated circumstances require superimposition upon the observation data of real events.
The key problem of such laboratory studies is related to the impossibility to more or less adequately com-
* See: V. Ulomov, "Seismic Menace in Russia", Science in Russia, No. 6, 2001. - Ed.
Graph of consequences of world earthquakes. Dots show the average values of damage and death rate for decades.
Change of acoustic phenomena during a micro-earthquake modeling in a granite sample. Signals emitted by the focus are shown in red, those from the surrounding areas-in blue. The diagram in the upper part demonstrates distribution of micro-cracks in the sample, which concentrate in the focus 14,200 seconds after the start of the experiment.
pensate for the time effect. In the subsoil, conditions for abrupt tectonic breakdowns build up in the course of hundreds or thousands of years, and an experiment should be carried out within a few hours.
Quite recently, with the use of the equipment developed by the US Geological Survey (Menlo-Park, California), a unique modeling of an underground shock focus was carried out. Participating in the analysis and processing of the obtained data were also specialists of the RAS Joint Institute of Physics of the Earth named after Otto Schmidt and the RAS Institute of Physics and Technology named after A. Ioffe.
The experiment boils down to the following. Samples were placed in special chambers under a high pressure. Then they were exposed to additional vertical stress as a result of which the tested granite monolith developed ruptures producing micro- earthquakes of sorts (similar to seismic vibrations). Thus, it was visually demonstrated how a destructive tectonic process is developing.
14,200 seconds after the test had started the tested granite developed a micro- earthquake focus in the form of a cluster of cracks. Simultaneously, the acoustic phenomena stimulated by the process substantially changed. Signals coming from zones around the focus at a certain point in time subdued, while the number
Diagram of signals from ruptures in the granite sample. Singular signals - green dots (1), double signals - blue dots (2), clusters - red dots (3). The generating micro-cracks accumulate in time in the micro-earthquake focus. As a result, the dot dispersion sharply decreases along the Y-axis showing the vertical distance from the displacement plane.
Foreshocks distribution chart. Above-a lull period (away from the epicenter marked with an asterisk is an activation period). Below-clusters of weak seismic shocks are denoted with crosses. P.K. - Petropavlovsk Kamchatsky, U.K. - Ust- Kamchatsk
of pulses emitted directly by the focus was upsurging manifesting the activation phase. There is every ground to interpret these phenomena as foreshocks.
Also symptomatic may be the "generation" of tone clusters-a set of characteristic audio oscillations emitted by ruptures co-located in time and space. They appear as the focus starts to develop. The emitting micro-cracks are concentrated in the center of the future rupture. Hence, the foreshock simultaneously indicates the time and the place of the simulated micro-earthquake. At that point the stress of the sample is quickly falling, as the rock monolith is losing its carrying ability and starts to disintegrate.
Unfortunately, a seismologist is unable to observe similar processes going on in the depth of lithosphere to compare with laboratory models. Obviously, natural tectonic activity is caused by a combination of different factors unaccounted for in simulated conditions. Nevertheless, in some cases it becomes possible to identify the above foreshocks.
Algorithms developed in our institute help isolate periods of seismic calm, activation and clusters against the background of all kinds of noises. For that purpose signals are analyzed emitted by numerous weak shocks that pose no threat.
Interesting and encouraging results were yielded by the study of materials related to major Kamchatka earthquakes (March 2, 1992, November 13, 1993, December 5, 1997) after the RTL method which has received international acclaim. RTL stands for region-time-length. The method registers zero marks corresponding to the perennial seismic background level. A reduction by a few units of the average-square error a reliably registers a lull. Subsequent return to the zero level indicates an approaching earthquake.
In the reviewed region seismic lull came 3 years before a powerful shock, while an activation phase lasted for about 1.5 years. Before the Kamchatka earthquake of December 5, 1997, information about a strong subsoil shock was transmitted to the Russian Federation Ministry of Emergencies. The forecast was made 16 months before the event.
The spatial distribution of fore-shocks is very remarkable too. A lull appears away from the focus, while activation starts in its immediate proximity. It is also there that clusters appeared a year before sharp vibrations. In other words, the laws elucidated through laboratory experiments got confirmation.
Similar materials were obtained by the studies of other powerful earthquakes in Russia, Japan, Turkey, Italy and USA, however in all of these cases the prognostic anomalies were identified only retrospectively.
Evolution of RTL parameter, based on data of Kamchatka earthquakes on March 2, 1992(1), November 13, 1993 (2,) and December 5, 1997(3). Zero marks show the perennial seismic background level. A reduction by a few units of the average-square error о reliably registers a lull. The graphs show that seismic lull settled 3 years before a powerful shock; the activation phase lasted for about 1.5 years. White arrow shows the date when information about a strong subsoil shock was transmitted to the Russian Federation Ministry of Emergencies. Asterisks correspond to earthquake epicenters.
The existing "code of ethics" requires of scientists to make available to colleagues in the concerned countries projections of tectonic disasters that may occur in their territories in a routine manner, rather than make official announcements. This recommendation adopted by the international seismological community is absolutely justified. It purports to avert a panic which may be caused by such news, which in itself is fraught in serious material damage, irrespective of whether the prediction proves true or false. The more so since the forecast accuracy still leaves much to be desired. False alarms that happen quite frequently may sometimes be qualified as economic subversion.
So, what makes a forecast reliable? First of all, it is a profound understanding of the nature of foreshocks; prompt feed in of data about anomalies in different geophysical fields; drawing on the amassed experience of accurate earthquake prediction. It is also important to carry on laboratory experiments and observations on specially equipped sites.
Unfortunately, since the beginning of the 1990s some prognostic grounds in this country have been closed, and the number of stations registering variations of seismic parameters has considerably decreased. Now they are being gradually restored.
Drawing analogy with weather forecast, we might say that geophysicists have learnt to approximate time and place of an "underground thunderstorm", but are still very far from accurate and prompt positioning of individual "lightning" discharges in time and space.
A medium-term forecast is presently most useful, ensuring a high-probability pinpointing of strong subsoil shocks within an interval of a few months or years. It is not aimed at provisions for evacuation, rather it seeks to prepare people in advance for the eminent calamity, to facilitate planning and conduct of the required construction works, consolidation of less rigid buildings and structures, relocation of dwellers from dilapidated houses and so forth.
In the future it may be quite possible to create weak artificial earthquakes by the impact of explosions, electromagnetic fields and other means. The objective of such activities is to reduce dynamic stresses in the projected focus and thus mitigate the risk of sudden powerful seismic shocks. Laboratory experiments in progress now are called to determine the required strength, frequency and duration of such impacts. The research supplies new hopes in man's confrontation with one of the most frightening elements.
Illustrations supplied by the author
Permanent link to this publication:
LUnited States LWorld Y G