by Valentin ULOMOV, Dr. Sc. (Phys. & Math.), Otto Schmidt United Institute of Physics of the Earth, Russian Academy of Sciences (RAS)

The RAS United Institute of Physics of the Earth is through with an important project - the seismic mapping and zoning of Russia's territory. This work is of much significance for land use and antiseismic construction in areas hit by earthquakes. Our new seismic maps have merited awards at home and abroad. One map has been chosen as a component part of the global seismic map published in 1999 under UN auspices.

EARTHQUAKES INEVITABLE

Earth tremors are among the worst natural calamities in terms of destruction and environmental impact. Caused by the global evolution of our planet's lithosphere, a process that has been going on for hundreds of millions of years, earthquakes are inevitable. Still worse, they often occur unexpectedly, like a bolt from the blue.

Unfortunately, thus far we are unable to predict the exact time of imminent seismic disasters - let alone prevent them; but what we can do is to minimize the aftermath. Certain steps are quite possible toward this end. Informed opinion, for one. The population and the authorities of endangered areas ought to be aware of the threat and know how to combat the disaster, if worst comes to worst. Seismic zoning (SZ) is an essential part of this deal. Among other things, SZ is meant to contribute to rational land use and adequate antiseismic construction.

ILL-STARRED MAP

The world's first official SZ map dates back to the year 1937. Its maker was G. Gorshkov, a geologist employed at the Seismological Institute of the USSR Academy of Sciences (this research center was the forerun-

page 18

ner of the present RAS United Institute of Physics of the Earth named after Otto Schmidt). However, Gorshkov and the authors of the subsequent SZ maps (published, respectively, in 1957, 1968 and 1978) did not take account of specific characteristics of the seismic regime of corresponding territories, they did not duly assess a random component implicated in geodynamic processes. In fact, all those maps proved to be "deterministic", as we say. And what concerns the ideas of a probabilistic approach in evaluating the seismic threat, they gained recognition above all abroad, even though conceived in our country (namely, our Russian scientist S. Medvedev was the first to suggest such kind of approach in 1947; later on, in 1965, Yu. Riznichenko substantiated these ideas, and then other Russian researchers further developed this theory).

Seismic zoning is one of the most forbidding problems belonging as it does to the category of forecasts based on an incomplete body of information, slender and occasionally even faulty experience, and none too explicit methods. All that plus jerrybuilt houses and other structures led to immense material damage for the national economy and heavy casualties during seismic events.

The SZ map of the Soviet Union's territory (1978) was not up to the mark either: within a relatively short time there came a series of violent earthquakes, and their seismic intensity* was 2 to 3 points higher than indicated on the map. Those were the destructive Spitak earthquake in Armenia (1988); the Zaisan quake in Kazakhstan (1990); one at Racha in Georgia (1991); at Susamyr in Kirghizia (1992). Next came a very bad earthquake at Neftegorsk on Sakhalin that took a toll of two thousand lives and wiped out that community What a bitter balance!

In actual fact, the SZ-78 map was not a "general" one, for it was drawn up fragmentarily, in bits and pieces, for various regions and Soviet republics, according to different procedures and on the basis of disparate seismological material. At that time we did not even have a single catalogue of earthquakes for our country's territory, we had no description of appropriate procedures and of the initial data used by the compilers of that "hapless" map. Therefore, turning to the job of making a new SZ map in 1991, we had to start with a clean slate, so to speak.

Involved in this work was a large research collective from RAS centers, from RAS branches in Siberia, Urals and the Far East. The author of the present article superintended the project. The federal Ministry for Science offered assistance within the framework of the master program Global Changes of the Natural Environment and Climate (with RAS

* Seismic intensity - a measure (value) used to assess the ground movement with the passage of seismic (earthquake) waves; this value describes the degree of destruction of material objects, the character of changes caused on the terrestrial surface and the response of people, victims of the quake. It is expressed in 12 points of the macroseismic scale as well as in accelerations, rates, displacements and in other units; and it depends on the magnitude (on the Richter Scale) of the shock, the distance from the seismic focus, ground conditions and other factors. - Auth.

page 19

Vice-President Nikolai Laverov in charge).

Aware of the drawbacks of the SZ-78 map, we decided to incorporate our own research findings in a set of probabilistic maps of General Seismic Zoning (GSZ) designed for construction projects of different categories of responsibility and service life. That is to say, we would not draw up one SZ map, as it was done in this country before and is still being done in other countries. We proceeded from the seismogeodynamic (SGD) approach when seismicity is viewed as a result of deformation of the earth's crust and of the entire lithosphere. We sought to take due account of their structural features, strength characteristics and destruction processes at hierarchic levels of different scale. We could thus develop an integral GSZ methodology and compile uniform seismological and geological-geophysical electronic databases for the territory of Northern Eurasia covering Russia and other former Soviet republics as well as adjacent seismically unsafe regions. We evolved a unified model of occurrence of earthquake source zones (OES zones) with adequate seismological parameterization. Featured in our computations and constructions were extended, i.e. more realistic, earthquake sources (not just focal points as before). Besides, we drew upon the latest knowledge concerning the nonlinear nature of seismic processes and synergetic phenomena of self-organization of SGD structures.

WHAT WE NEED ARE MODELS

It became obvious long ago: seismic threat zoning, if predicated exclusively on what we know about the previous earthquakes and in the absence of adequate prognostic SGD models, is absolutely useless. These models should be based on structural and geodynamic regularities proper to the vast territory of Northern Eurasia as well. Such regularities are manifest in the hierarchical heterogeneity (non-uniformity) of present-day tectonic structures, from the lithosphere on to earth shell blocks of different rank; and they show up in the directionality of the geodynamic development of these structures. The interdependence between regional seismicity, and the structure and dynamics of the lithosphere is best expressed in the SGD interactions of lithosphere plates.

Most active are the convergent (coming together at a point) lithospheric structures on arcuate boundaries between lithosphere plates at the periphery of oceans; such structures are represented by subduction zones (when one plate underthrusts another) and by their relicts on the continents. These structures are orderly enough in dimensions and position. Globally, the mean statistical extension of each such region is equal to 3,000 +/- 500 km. Commensurate with this value are the predominant distances between the centers of their pairs nearest to each other. The dimensions of seismoactive areas and their spatial distribution have been found to be directly related to the magnitude of most probable earthquakes, an all-important factor for SZ.

page 20

Every region like that is characterized by a seismic regime proper to it. Therefore, to synthesize models of OES zones we selected those having the above-named parameters as an "initial" seismogenerating unit. The regularities discovered thereby were taken as a basis for dividing the territory of Northern Eurasia into seismogenetically homogeneous regions.

From a geological standpoint, this territory takes in four large platforms of different age - the East-European, West-Siberian, Turanian and Siberian, which are characterized by a relatively minor and diffuse seismicity; furthermore, Northern Eurasia also includes a number of orogenic (mountain-making) areas of exceedingly high seismic activity: Iran-Caucasia-Anatolia (Asia Minor); Central Asia; the Altai-Sayany-Baikal, Kurile-Kamchatka regions, among others. The Kuriles and Kamchatka, an area that has a subduction zone with earthquake sources at a depth of over 600 km, is the most mobile and seismoactive part of Northern Eurasia; it is a district of most violent tectonic tremors, a district where the bulk of SGD deformations in Northern Eurasia is released. Sources of intermediate depth of occurrence (from 30 to 300 km) are characteristic of two other typical zones of relict subduction - Vranica in the Eastern Carpathians and the Pamir- Hindu Kush district in Central Asia. Most of the earthquake sources are located in the upper part of the earth crust not deeper than 15 km; they pose the greatest threat because of their being rather close to the earth surface.

Major earthquakes occur in each of the above regions at least three times as often as it was believed before. The idealized picture of rectilinear plots of earthquake recurrence* used before resulted in the significant underrating of the seismic threat throughout the territory.

WONDERFUL STRUCTURE

Earthquakes originate in a discrete stratified in-blocks medium having a structure predetermined by the preceding geological epochs and ultimately, by the recent epoch. Tectonic tremor sources are not dispersed randomly but are associated with relatively narrow lineamental (rectilinear or almost rectilinear) zones of active faults. The dimensions of these faults and the distances between them depend, in turn, on the depth and strength characteristics of corresponding layers where crash movement took place. The thicker the layer, the deeper and more extended these faults are, and the more powerful the earthquake sources. The distances between the dislocation nodes of crossing faults and, accordingly, the dimensions of geoblocks formed by them have a clear tendency to grouping according to rank whereby their scale is doubled sequentially. This must be due to a regular twofold increase in the depth of the main boundaries of the divide in the earth crust and the upper mantle-boundaries reached by faults of corresponding ranks. For instance, the roof of the "granite" layer on continents lies at a mean depth of about 10 km; the boundary of the "basalt" layer is 20 to 25 km deep; the base of the earth crust - 40 to 50 km deep, of the lithosphere - 100 km, and of the asthenosphere - about 200 km deep. Further boundaries lie at a depth of about 400 and 700 km. Probably all geological horizons, the earth surface including, obey this fundamental law whereby the physical properties of matter undergo sudden changes with the doubling of its depth of occurrence.

Such kind of orderliness is a function of certain regularities not only in the systems of tectonic faults and geoblocks but also in the hierarchy of earthquake sources: the farther the sources are from one another, the stronger the quakes. Ranked according to magnitude intervals and the elastic (strain) energy released, earthquake sources obey definite laws in their distribution patterns both in real time and in space. The selfsame intriguing figure "two" turns up here again. The mean statistical dimensions (extension) of earthquake sources and the distances between the epicenters of their neighboring pairs are changed by a factor of two with every step of 0.5 magnitude. For instance, the extension of the earthquakes sources of M=6.5 approaches 25 km, that of the sources of M=7 is about 50 km; in a similar way, M=7.5 correlates with 100 km, M-8 - with 200 km, and so forth. The epicentral distances between pairs of sources are four times as long as the extension of respective sources, while the interrelationships of these values do not depend on the magnitude, that is these relationships are invariant with respect to the magnitude; this is the proof of self-similitude (fractality**) in the hierarchy of the dimensions of the interacting crustal blocks and of the sources of tectonic shocks.

Likewise orderly is the hierarchy of soliton-like ("solitary") deformation waves of seismic activation, the geons, according to the author's terminology; these are related to the dynamics of interacting geoblocks and to the directionality of synergic (cooperative) SGD processes. Propagating along faults of corresponding rank, geons produce (and destroy too) various "locks" ("catches") and thus precipitate earthquakes. Since these geo-dynamic processes develop at every hierarchical scale level, they are characterized by the same fractal dimensionality which is proper to the stratified in-blocks medium.

The interdependences in the orderliness of faults, geoblocks and earthquake sources, and also in the synergic development of SDG processes were taken as a basis of a model suggested by the author of the present article in 1987 - the fractal lattice model (FLM) of seismogenesis. It enabled a new approach to identification of OES zones relative to SZ of Northern Eurasia.

MODEL OF EARTHQUAKE SOURCES

We made use of the lineament-domain-focal (LDF) model which

* Such plots describe the seismic regime, i.e. the frequency of quakes of different magnitude. - Ed .

** Related to fractals, or objects characterized by self-similarity in a wide range of scales. - Auth.

page 21

then became a followup on our FLM model. Considered in the LDF model are four scale levels of earthquake sources: a large region of the above-named dimensions and integral characteristics of the seismic regime plus three basic structural elements of this regime- lineaments, which in a generalized form represent the axes of three-dimensional seismoactive fault or displacement structures (the groundwork of the LDF model); domains covering the geodynamically quasi-homogeneous volumes of the geological medium and characterized by diffuse seismicity; and potential earthquake sources indicative of the most dangerous sites of lineamental structures. Geons are the "engine": of the LDF model ofOES zones.

The lineaments, domains and potential sources are classified the same way as earthquakes, that is according to the maximum magnitude value with a step of 0.5 M. The minimum value of this characteristic along lineaments is equal to 6, for earthquake sources with a lower magnitude are identified not as reliably in the case of generalized seismic zoning, that is OES. But we can downgrade the lower threshold of magnitudes in detailed SZ. The upper threshold is determined by the actual SGD situation, while the lower one - by a minimal seismic threat which has to be taken into consideration in construction. In our studies we took the minimum M value equal to 4, and for minimum seismic intensity - 5 points on the macroseismic scale.

Yet earthquake sources are not rigorously aligned along lineament axes but deviate on both sides. Proceeding with our work, we obtained mean statistical values of such deviations. We found: the smaller the magnitude of a quake, the farther from the lineament axis its focus may be. Such kind of "dissipation" is determined by the dimensions of the regions of the dynamic impact of lineamental structures on the adjacent geological medium and by its fractal structure. Remarkably, the distribution of different-rank lineaments (correspondingly, of earthquakes of different magnitude) throughout Northern Eurasia by and large reflects a single structure of their hierarchical set. This confirms the validity of our concept about the structural-dynamic unity of the geophysical medium and the ongoing seismic processes there and, consequently, the relevance of the LDF model. The SGD characteristics obtained with its aid were used by us for identification of OES zones and for the modeling of a prognostic (virtual) seismicity essential for the computations of seismic "shakability" in Northern Eurasia.

VIRTUAL SEISMICITY

In spite of a vast amount of work done to systematize and unify earthquake catalogues for Northern Eurasia, the available body of data on seismic phenomena in this large territory is still all too scanty. Instrument observations - both in this country and abroad - were begun but 100 years ago, with quality data obtained much later, beginning with the mid-1960s. Historical evidence is not better off either. What we get from ancient civilizations are random bits of information on some of the major earthquakes, say, in the Near and Middle East, in Central Asia and in North China. Such evidence is altogether absent for sparsely populated regions of Siberia and the Far East. Likewise disjointed and unreliable are data on paleoseismodis-locations, i.e. geological traces left by the long-ago earthquakes. Hence the selfsame conclusion: we should develop SGD models that could take in any significant data on the geological structure, general geodynamics, paleoseismodislocations and the seismic regime of each particular region within Northern Eurasia.

Following the LDF model and our procedures, we calculated the rate of a flow of seismic events of different magnitudes for each region and drew up a model regional catalogue of earthquakes covering a period of dozens and even hundreds of thousands of years. Then we distributed all these events among the seismogenerating elements of Northern Eurasia - according to their dimensions and rank. At this point the LDF model of OES zones came alive, as it were, with each lineament, domain and potential sources "touching off tremors" of corresponding magnitude and frequency of recurrence. It thus became possible to make maps of virtual seismicity for any reasonable span of time (50, 100, 500 and more years) and study the effect of virtual quake sources on the earth surface.

At the concluding stage of our work we evaluated the seismic effect of each such focus; its location and frequency of occurrence were obtained in a playoff by means of a randomizer (random number generator), with due account being taken of the respective virtual catalogue, dimensions of the source, magnitude of the quake and the shock-wave attenuation law at a distance from the source. Computations were made for each nodal point of a square grid map covering the entire territory of Northern Eurasia at a step of 25х25 km. There were more than 100 thousand nodal points like that, and each was supplied with computer-fed information on the recurrence of different-intensity shocks; thereupon this information was used for mapping the extent of seismic hazard. Proceeding accordingly, we computed and drew up two types of maps. The first type, reflecting the periods of seismic intensity on the macroseismic scale, was evolved by registering the value of ground movements equal to one point. The other type of SZ maps incorporated seismic intensity values; we drew them up on the basis of the same data supplemented with concrete values assigned to the periods of shock recurrence (500, 1,000 and 5,000 years).

RECOMMENDED FOR CONSTRUCTION

As we have already said, seismic zoning (SZ) is possible if made on a probabilistic basis only. Such is the basis of our maps of general seismic zoning of the territory of Northern Eurasia compiled in 1997 (GSZ-97). They allow to assess the extent of seismic menace for objects of various service life periods and categories of

page 22

responsibility at three levels reflecting the rated intensity of jolts expected within a given area at an assigned probability during a definite time span. Thus, according to GSZ-97, the probability of a possible exceedance of earthquake intensity within 50 years shapes up as follows: 10 percent (map GSZ-97-A), 5 percent (GSZ-97-B) and 1 percent (GSZ-97-C), which corresponds to the mean periods of 500, 1,000 and 5,000 years for the recurrence of such tremors. The seismic effect prognosticated by these maps is coupled to average ground conditions in keeping with the present construction standards and rules (building code) operating in earthquake-prone districts.

Our set of GSZ-97 (А, В, С) maps was endorsed by Russia's Gosstroi (State Building Committee) for use in construction throughout this country. The GSZ-97-A map was recommended for the mass construction of residential, public and production buildings; the other two maps - GSZ-97-B and GSZ-97-C-for the erection of objects that should continue in service even during earthquakes and during work to eliminate their aftereffects (power and water supply, fire stations, communication facilities, transportation routes) and also for premises housing a large number of people (hospitals, schools, kindergartens, railway stations, air terminals, theaters, roofed-in markets, stadiums and like structures) and for buildings higher than 16 stories.

The GSZ-97 set is supplemented with maps indicating the recurrence periods for jolts of different intensity; this is likewise important for the practice of antiseismic construction because multiple seismic shocks may cause mechanical damages which, if accumulated, can sizeably reduce the strength of structures and, consequently, affect their resistance to subsequent quakes. Such high-danger objects as nuclear power stations are a special case. Accordingly, we drew up yet another map, GSZ-97-D, which corresponds to a quite rare pattern of earthquake recurrence - one shock per 10,000 years on the average.

The GSZ-97 maps have been published in a large print and in a wall format (scale-1:8,000,000). In 1998 this set won a first- class award at the International Exposition and Fair Innovations-98 and a medal from the

page 23

page 24

All-Russia Exhibition Center. One map, GSZ-97-A, showing peak accelerations of seismic oscillations throughout Northern Eurasia, in 1999 was entered into the Global Map of the Seismic Hazard.

THE THREAT IS GREATER THAN BELIEVED BEFORE

Thus for the first time ever we have been able to do the job of seismic zoning for all of Northern Eurasia, including its level lands and shelves of its border and inland seas. We have seen that a large part of the Russian Federation's territory is subject to stronger seismic tremors than it was believed earlier. There is a 1 percent probability of 6-point jolts occurring in half of the nation's territory within 50 years. The same probability value holds for violent and disastrous quakes of 8-9 points and higher in a third of our territory Given the 10 percent risk of a possible excess of seismic intensity within 50 years (i.e. using the GSZ-97-A), the extension of such areas, even though it shrinks by about 20 percent, still remains large enough. Judging by this map, nearly 30 percent of Russia's territory can be hit by 7-point seismic shocks. About 10 percent of the total area is under extremely dangerous zones of 8-9 and 9-10 points.

Among the most seismoactive regions are the Far East, southern Siberia and northern Caucasia. According to the GSZ-97-C map (the more so, the GSZ-97-D map), 6-7 point zones of European Russia pose a definite threat to high-risk construction projects as well; here, too, antiseismic preventive measures are imperative. Such zones cover the Middle Urals and adjacent districts, the Azov Sea and the \blga areas, the Kola Peninsula and contiguous territories. Besides, the locally induced seismic activation in the oil- mining districts of Tatarstan and in the ore-mining and processing enterprises of the Perm Region in the Urals can also cause some damage to the national economy. Next, the long low-frequency tremors of 4-5 points, which propagate over vast distances from the deep-seated foci of major earthquakes in the eastern Carpathians, are capable of damaging high-rise structures even at a very great distance from the epicenters, as far as the Moscow Region* (such structures are sensitive to seismic oscillations like that).

Reverberations of shallow, though strong earthquakes in western Turkmenia are felt in Russia as well. Seismically dangerous and thus ecologically vulnerable are the basins of the Black and Caspian Seas, the shelves of the Sea of Okhotsk, of the Chuckchee and Barents Seas, and also of the Laptev Sea; all these seas are oil and gas producers. There is a heightened risk of atomic power stations and other nuclear objects built in seismically active districts - even minor tremors can interfere with their normal operation.

Environmental modification as a result of man's economic activities and his action on the lithosphere of the earth is another seismic menace (extravagant mining of oil, gas and minerals, construction of major hydrotechnical projects, injection of industrial wastes and the like).

WHAT NEXT?

In their conceptual design, the maps of General Seismic Zoning of the territory of the Russian Federation (GSZ-97) differ from those conceived previously. Owing to their concrete probabilistic assessments, these new maps now enable bodies of government to make a qualitative evaluation of the seismic risk factor. Yet compared with the former seismic zoning maps, the GSZ-97 set of maps indicates somewhat larger areas of enhanced seismic danger; what is more, they show up areas formerly considered to be absolutely safe (aseismic). For a variety of reasons, using these maps in the building business calls for scientific counseling on the part of our research staff.

Such work should be carried out above all in the seismically endangered regions of this country (the Far East, southern Siberia and Northern Caucasia), and also in seismically dormant territories of the densely populated European part of Russia which, in fact, have been bypassed by seismologists in their studies (the Middle Urals, the Volga Region and other areas of the East-European platform). It is planned to effect scientific counseling within the framework of a goal-oriented federal program now at the gestation stage at Gosstroi (Seismosecurity of Russia); the following works are envisaged: larger-scale maps of GSZ-97 to be drawn up for members of the Russian Federation; supplying government bodies with lists of endangered population centers where seismic events are possible (three levels of hazard according to GSZ- 97 maps. A, B, C); large-scale maps of recurrence periods of different-intensity tremors to be compiled for the territories of the Russian Federation's members; entering new data into the catalogue of earthquakes in Russian Federation and in neighboring regions; backing up all projects with regional data on OES zones...

Let me stress in conclusion: we need an effective system of earthquake insurance. Such policy is bound to boost the efficiency of seismic-proof construction with the use of the GSZ-97 maps. In the absence of insurance policies and adequate supervision over the quality of construction, the government had (and will have in the future too) to hunt for funds to make good the damage inflicted by earthquakes. In other countries, for instance, such funds are drawn from respective insurance fees.

Now what concerns the further course of basic and applied research in the field of seismogeodynamics and seismic zoning: it should zero in on fundamentals of long-term prognostication of major earthquakes.

* See: A. Nikonov, "Earthquakes in Russia", Science in Russia, No. 3, 2000. - Ed.

Permanent link to this publication: