Earthsci9_02kopnichevlo

Doklady Earth Sciences, Vol. 387A, No. 9, 2002, pp. 1063–1066. Translated from Doklady Akademii Nauk, Vol. 387, No. 4, 2002, pp. 528–532.
Original Russian Text Copyright 2002 by Kopnichev, Pavlis, Sokolova.
English Translation Copyright 2002 by MAIK “Nauka /Interperiodica” (Russia). GEOPHYSICS
Lithospheric Inhomogeneities and Strong Earthquake Sources
in the Central Tien Shan Region
Yu. F. Kopnichev1, G. Pavlis2, and I. N. Sokolova1
Presented by Academician V.N. Strakhov May 22, 2002 Mapping of the shear wave attenuation field in the relatively short time period. In total, we processed lithosphere of the Central Tien Shan region was carried about 500 earthquake records at the depth of 190– out on the basis of deep-focus Hindu Kush earthquakes 230 km and an epicentral distance (Δ) of 500–1100 km.
registered at more than 40 digital and analog stations.
The frequency filtration of vertical components was An intense attenuation stripe related to sources of two carried out during the analysis of digital records. We used strongest earthquakes (M ≥ 7.0) in the Tien Shan region a filter similar to the corresponding FSSS filter with the since the middle 1970s was distinguished in the west- following parameters: central frequency = 1.25 Hz, ern part of the area. It has been established that the width = 2/3 octave at the level equal to 0.7 of the max- structure of the lithospheric attenuation field substan- imum [5]. Analog seismograms were preliminarily tially changed over 10–20 years. The data obtained can scanned with the assistance of a wide-frame scanner be attributed to an intense rearrangement of the fluid and digitized at the frequency of 40 Hz.
field in the Earth’s crust and upper mantle, relatedamong other things to the preparation of strong earth- For analyzing the attenuation field characteristics, we used two parameters: logarithm of the ratio of max-imum amplitudes in S and P waves (S/P), as well as Recent data on the important role of mantle fluids in logarithm of the ratio of maximum in the S wave to the the preparation of strong crustal earthquakes [1–3] indi- coda level at t = 400 s, where t is counted from the cate that characteristics of the attenuation field of short- lapse time (S/c400). Maximal amplitudes were mea- period shear waves can be successfully used in the sured during the 10-s interval from the arrival time for future for long-term and middle-term predictions. Only longitudinal waves and the ±10-s interval from the the electrical conductivity of rocks is comparable with S wave arrival time on the travel-time curve for shear the attenuation of S waves in terms of sensibility to the liquid phase [4]. The aim of the present work is to studyspatiotemporal variations of the lithospheric attenua- In the considered range of epicentral distances for tion field in the Central Tien Shan and compare them hypocenters located at the depth of ~200 km, direct P with the seismicity of the region the last 20−25 years.
and S waves fall onto the M boundary at sufficientlylow angles and intersect the crust at i ~ 46°–49°. (For The study region (39°–45° N and 73°–81° E) com- the simplest two-layer model of a medium with the prises a substantial part of the Central Tien Shan region S wave velocity in the crust and upper mantle equal to along with the southern margin of the Kazakh Platform 3.5 and 4.6 km s–1 [6], respectively, the ray angle with in the north and the northwestern margin of the Tarim the vertical is equal to i ~ 71°–82°.) Taking into account S/P values, which can substantially differ even for close We used the records of deep-focus Hindu Kush stations, as well as earlier results of the attenuation field earthquakes obtained by 27 PASSCAL digital stations mapping for the northern Tien Shan region [7], one may and 11 KNET digital stations of the Kirgiz telemetric assume that the S/P parameter mainly characterizes the network in 1997–2000, as well as in 4 analog stations attenuation of shear waves at the distance of 30−60 km (CKM-III) in 1976–1999. The analysis of Hindu Kush southwest of the relevant stations in the Earth’s crust earthquake seismograms is convenient, since it allows (first of all, in its lower part at the depth of 30–55 km) us to obtain representative experimental material for a It was shown in [8, 9] that the S coda of Hindu Kush earthquake records was mainly formed by shear waves 1 Schmidt Joint Institute of Physics of the Earth, reflected from numerous subhorizontal boundaries in the upper mantle. In this case, as time elapses, S waves Bol’shaya Gruzinskaya ul. 10, Moscow, 123810 Russia on the coda intersect the lithosphere and asthenosphere at more and more acute angles. Hence, parameter Fig. 1. Map of the study region. Triangles denote seismic stations. Parameter: S/P (1) weak attenuation, (2) strong attenuation;
(3) source zone of the Suusamyr earthquake. Epicenters of large earthquakes in the strong attenuation stripe area: (4) M = 7.0, (5) M =
6.0; (6) Talas–Fergana Fault; (7) sources with anomalously high values of the helium isotope ratio.
S/c400 characterizes the correlation of S wave attenua- Figure 2 shows the relationship between average tion in the crust and upper mantle with distance from S/P values and average epicentral distance for the stud- the station (reflected S waves appearing on the coda at ied stations. It can be seen that parameter S/P for simi- t ~ 400 s intersect the M boundary at distances of ~10– lar Δ values can change by more than one order of mag- nitude (KOPG and TKM2 stations). Confidence inter-val for the average S/P value at the level of 0.9 variesfor different stations between 0.10 and 0.30.
The attenuation is generally weak for the Tarim massif margin, (except for the westernmost stations WQIA and KASH (Figs. 1, 2)). This statement is also valid for the following stations located at the northern margins of large depressions where following direct P and S waves intersect the lower crust: USP (Chu Depression), ANA (Issyk-Kul Depression), and KAI and NRN (Naryn Depression). Sufficiently high S/P values were obtained for stations KHA and KUU (southern margin of the Kazakh Platform), which is consistent with elevated velocities of P and S waves here relative to the Tien Shan region [6].
It follows from Fig. 1 that comparatively high S/P values are observed in the majority of the study region.
Against this background, one can see a strong attenua-tion stripe extending from KASH to TKM2 (minimum Fig. 2. Dependence of parameter S/P on epicenter distance.
corrected for Δ S/P values correspond to these stations).
Black and white circles denote strong and weak attenuation, In the southern part, this stripe extends along the Talas– respectively (Fig.1). The straight line shows a conditional Fergana Fault, which separates the western and central boundary separating the regions of S/P values correspond-ing to strong and weak attenuation.
Tien Shan regions, and in the KAZ station area turns to DOKLADY EARTH SCIENCES Vol. 387A No. 9 2002 LITHOSPHERIC INHOMOGENEITIES AND STRONG EARTHQUAKE SOURCES the north-northeast. A comparatively strong attenuation is recorded in the Zaili Fault (station TLG) area and thesouthern margin of the Issyk-Kul Depression (stationsULHL and KAR).
It should be noted that the sources of the two stron- gest earthquakes in the Tien Shan region during the last 25 years, namely, the Kashgar earthquake of August 23, 1985, (M = 7.0) and the Suusamyr earthquake of August 19, 1992, (M = 7.3), were confined to boundaries of theabove-mentioned attenuation stripe. This stripe also pro-voked the earthquake of January 9, 1997, (M = 6.0) in aregion where, according to instrumental and historicaldata, no events with M > 5.0 had been known before.
Figure 3 shows common envelopes of the S coda for several stations installed in and nearby the source zoneof the Suusamyr earthquake. The envelopes were con-structed beginning from the maximum in the S wave. Att < 400 s, the envelopes substantially differ in shape and reveal abrupt bends related to the increase and decreaseof slope. As the numerical simulation indicated [10],these bends resulted from the existence of zones char-acterized by a strong contrast of attenuation in the Earth’s crust and upper mantle (in the given case, nearthe recording stations).
The envelopes were plotted for two stations over different time intervals. It can be seen that the codadecay rate in 1992 was higher (relative to 1980) at sta- tion TORK located at a distance of ~20 km from thesource zone of the Suusamyr earthquake. At the sametime, it substantially decreased in 1991–1992 (relativeto 1976–1977) at station KRSU located approximately35 km southward (farther from the source zone).
Figure 4 demonstrates the dependence of the aver- age value of parameter S/c400 on distance. It should benoted that the dispersion of this parameter at the studiedstations is substantially (approximately two times) lower than that for parameter S/P. Relatively lowS/c400 values are observed for the majority of stations.
At the same time, very high values of this parametercorrespond to stations located in the source zone of the Fig. 3. Common envelopes of the S coda for stations located
in the source zone of the Suusamyr earthquake and its vicin-
Suusamyr earthquake (AML) and in its vicinity (TORK ity. (1) Records of stations KRSU and TORK in 1976–1977 and NICH). The envelopes are also relatively steep at and 1980, respectively; (2) the same in 1991–1992; stations located up to 60–70 km away from the source (3) envelope for station KAZ.
Let us note that the very high level of S wave with lithosphere were previously detected in source zones of respect to the coda and the impulsive character of this other strong earthquakes in the Tien Shan region [2, 3].
group at station AML is related to a weak attenuation inthe lower crust located in the south rather than the effect A detailed mapping of the attenuation field in this of focusing, since our data show that direct P and S region was carried out in [11] on the basis of numerous waves recorded by this station have very high fre- local earthquakes recorded by a remote highly sensitive quency spectra relative to the neighboring stations. At station. Judging from these data, the strong attenuation the same time, very high values of S/c400 for stations stripe identified by the authors of this work between the AML and TORK point to a sharp increase of attenuation KASH and TKM2 stations was absent during the in the lower crust and upper mantle as the Suusamyr 1970s. In addition, Figure 3 suggests that the structure earthquake source is approached. This effect can be of the attenuation field at KRSU and TORK underwent explained by the existence of a subvertical, strong attenuation zone penetrating from the lower crust into The comparatively fast change of the S wave atten- the upper mantle. Similar features of the fine structure of uation field can be related only to a rearrangement of DOKLADY EARTH SCIENCES Vol. 387A No. 9 2002 anomaly. Data of the Institute of Seismology, National Academy of Sciences, Kazakhstan suggest that a fairly vast area of seismic quietness can be distinguished south of station TKM2. Since 1996, earthquakes with source depths of ~20 km have been recorded here (it is known that increase in the share of relatively deep- focus events serves as an important prognostic indica-tor [2, 15]).
We are grateful to R.T. Beisenbaev for placing at our disposal analog records from station KUU.
1. Kopnichev, Yu.F., Dokl. Akad. Nauk, 1997, vol. 356, Fig. 4. Dependence of parameter S/c400 on the epicenter
2. Kopnichev, Yu.F. and Mikhailova, N.N., Dokl. Akad. distance. Black circles denote the stations located within the Nauk, 2000, vol. 373, no. 1, pp. 93–97.
source zone of the Suusamyr earthquake and its vicinity.
3. Kopnichev, Yu.F., Sokolova, I.N., and Shepelev, O.M., Dokl. Akad. Nauk, 2000, vol. 374, no. 1, pp. 99–102.
the fluid field in the Earth’s crust and upper mantle.
4. Berdichevskii, M.N., Borisova, V.P., Golubtsova, N.S., Judging from our previous data [1, 11, 12] and MTS et al., Fiz. Zemli, 1996, no. 4, pp. 99–107.
data [4], fluid-filled interconnected channels, which inter- 5. Zapol’skii, K.K., Eksperimental’naya seismologiya sect various tectonic structures existing in the lower (Experimental Seismology), Moscow: Nauka, 1971,pp. 20–36.
crust of the Tien Shan region. At the same time, the flu-ids can rise along the roots of large fault zones from the 6. Roecker, S., Sabitova, T.M., Vinnik, L.P., et al., J. Geo- phys. Res., 1993, vol. 98, no. 9, pp. 15 779–15 795.
upper mantle into the crust [2, 3, 13]. It should be notedthat very high (submantle) ratios of helium isotopes 7. Kopnichev, Yu.F., Dokl. Akad. Nauk, 2000, vol. 375, were recorded at the end of the 1980s in groundwater within the strong attenuation stripe (Fig.1). Previously, 8. Kaazik, P.B., Kopnichev, Yu.F., Nersesov, I.L., and such ratios were never encountered beyond the areas of Rakhmatullin, M.Kh., Fiz. Zemli, 1990, no. 4, pp. 38–49.
9. Kaazik, P.B., Kopnichev, Yu.F., and Rakhmatul- lin, M.Kh., Seismicheskie volnovye polya (Seismic Wave The sharp change of the attenuation field at stations Fields), Moscow: Nauka, 1992, pp. 16–26.
KRSU and TORK over 12–15 years points to a migra- 10. Kaazik, P.B. and Kopnichev, Yu.F., Vulkanol. Seismol., tion of fluids toward the Suusamyr earthquake source prior to this event. In addition, the attenuation anomaly 11. Kopnichev, Yu.F. and Nurmagambetov, A.N., Fiz. Zemli, in the AML station area indicates that the channels along which fluids ascended from the upper mantle 12. Kvetinskii, S.I., Kopnichev, Yu.F., Mikhailova, N.N., et al., were preserved here even within 7–8 years after the Dokl. Akad. Nauk, 1993, vol. 329, no. 1, pp. 25–28.
Suusamyr earthquake. This statement is consistent with 13. Kopnichev, Yu.F. and Sokolova, I.N., Fiz. Zemli, 2001, data on the rise of mantle fluids in source zones of strong earthquakes over several decades after theseevents [3].
14. Polyak, B.G., Kamenskii, I.L., Sultankhodzhaev, A.A., et al., Dokl. Akad. Nauk SSSR, 1990, vol. 312, no. 3, Sources of the two strongest earthquakes in the Tien Shan region during the last 25 years were confined to 15. Nersesov, I.L., Ponomarev, V.S., and Teitel’baum, Yu.M., the high-attenuation stripe. Hence, the next strong Dokl. Akad. Nauk SSSR, 1979, vol. 247, no. 5, pp. 1100– earthquake could also take place in the area of this DOKLADY EARTH SCIENCES Vol. 387A No. 9 2002

Source: http://www.kndc.kz/kndc/docs/publication/eng/topic2002_1.pdf