Long-term earthquake prediction in the circum-pacific convergent belt

Investigation of the time-dependent seismicity in 274 seismogenic regions of the entire continental fracture system indicates that strong shallow earthquakes in each region exhibit short as well as intermediate term time clustering (duration extending to several years) which follow a power-law time distribution. Mainshocks, however (interevent times of the order of decades), show a quasiperiodic behaviour and follow the ‘regional time and magnitude predictable seismicity model’. This model is expressed by the following formulas $$\begin{gathered} \log T_t = 0.19 M_{\min } + 0.33 M_p - 0.39 \log m_0 + q \hfill \\ M_f = 0.73 M_{\min } - 0.28 M_p + 0.40 \log m_0 + m \hfill \\ \end{gathered} $$ which relate the interevent time,T _t (in years), and the surface wave magnitude,M _f , of the following mainshock: with the magnitude,M _min, of the smallest mainshock considered, the magnitude,M _p , of the preceded mainshock and the moment rate,m ₀ (in dyn.cm.yr^?1), in a seismogenic region. The values of the parametersq andm vary from area to area. The basic properties of this model are described and problems related to its physical significance are discussed. The first of these relations, in combination with the hypothesis that the ratioT/T _t , whereT is the observed interevent time, follows a lognormal distribution, has been used to calculate the probability for the occurrence of the next very large mainshock (M _s ≥7.0) during the decade 1993–2002 in each of the 141 seismogenic regions in which the circum-Pacific convergent belt has been separated. The second of these relations has been used to estimate the magnitude of the expected mainshock in each of the regions.     

Long-term earthquake prediction in the Aegean area based on a time and magnitude predictable model

The Aegean and surrounding area (34°N–43°N, 18°E–30°E) is separated into 76 shallow and intermediate depth seismogenic sources. For 74 of these sources intervent times for strong mainshocks have been determined by the use of instrumental and historical data. These times have been used to determine the following empirical relations: $$\begin{gathered} \log T_t = 0.24M_{\min } + 0.25M_p - 0.36\log \dot M_0 + 7.36 \hfill \\ M_f = 1.04M_{\min } - 0.31M_p + 0.28\log \dot M_0 - 4.85 \hfill \\ \end{gathered} $$ whereT ₁ is the interevent time, measured in years,M _min the surface wave magnitude of the smallest mainshock considered,M _p the magnitude of the preceding mainshock,M _f the magnitude of the following mainshock, \(\dot M_0 \) the moment rate in each source per year. A multiple correlation coefficient equal to 0.74 and a standard deviation equal to 0.18 for the first of these relations were calculated. The corresponding quantities for the second of these relations are 0.91 and 0.22. On the basis of the first of these relations and taking into consideration the time of occurence and the magnitude of the last mainshock, the probabilities for the occurrence of mainshocks in each seismogenic source of this region during the decade 1993–2002 are determined. The second of these relations has been used to estimate the magnitude of the expected mainshock.     

Long-term earthquake prediction along the western coast of South and Central America based on a time predictable model

The repeat times,T, of strong shallow mainshocks in fourteen seismogenic sources along the western coast of South and Central America have been determined and used in an attempt at long-term forecasting. The following relation was determined: $$\log T = 0.22M_{\min } + 0.21M_p + a$$ between the repeat time,T, and the magnitudes,M _min, of the minimum mainshock considered andM _p , of the preceding mainshock. No dependence of the magnitude,M _f , of the following mainshock on the preceding intervent time,T, was found. These results support the idea that the time-predictable model is valid for this region. This is an interesting property for earthquake prediction since it provides the ability to predict the time of occurrence of the next strong earthquake. A strong negative dependence ofM _f onM _p was found, indicating that a large mainshock is followed by a smaller magnitude one, andvice versa. The probability for the occurrence of the expected strong mainshocks (M _s ≥7.5) in each of the fourteen seismogenic sources during the next 10 years (1992–2002) is estimated, adopting a lognormal distribution for earthquake interevent times. High probabilities (P ₁₀>0.80) have been calculated for the seismogenic sources of Oaxaca, Chiapas and Southern Peru.     

A time- and magnitude-predictable model for generation of shallow earthquakes in the Aegean area

Repeat times of strong shallow mainshocks have been determined by the use of instrumental and historical data for 68 seismogenic sources in the Aegean and surrounding area (34°N–43°N, 18°E–30°E). For 49 of these sources at least two interevent times (three mainshocks) are available for each source. By using the repeat times for these 49 sources the following relation has been determined: $$\log T_t = 0.36M_{\min } + 0.35M_p + a$$ whereT _t is the repeat time, measured in years,M _p the surface wave magnitude of the preceding mainshock,M _min the magnitude of the smallest earthquake considered and “a” parameter which varies from source to source. A multilinear correlation coefficient equal to 0.89 was determined for this relation. By using the same repeat times for the 49 seismogenic sources, the following relation has been determined between the magnitude,M _f , of the following mainshock andM _min andM _p . $$M_f = 0.95M_{\min } - 0.49M_p + m$$ wherem is a constant which varies from source to source. A multilinear correlation coefficient equal to 0.80 was found for this relation. The model expressed by these two relations is represented by a scheme of a time variation of stress under constant tectonic loading. In this scheme, the maximum stress values during the different seismic cycles fluctuate around a value, τ₁, in a relatively narrow stress interval, expressing the high correlation coefficient of the relation between LogT andM _p . On the contrary, the minimum stress values fluctuate around a value, τ₂, in a much broader stress interval. However, each of these minimum stress values becomes lower or higher than τ₂ if the previous one is higher or lower than τ₂, respectively, expressing the negative correlation betweenM _f andM _p .     

Predicting earthquakes by analyzing accelerating precursory seismic activity

During 11 sequences of earthquakes that in retrospect can be classed as foreshocks, the accelerating rate at which seismic moment is released follows, at least in part, a simple equation. This equation (1) is ,where is the cumulative sum until time,t, of the square roots of seismic moments of individual foreshocks computed from reported magnitudes;C andn are constants; andt _fis a limiting time at which the rate of seismic moment accumulation becomes infinite. The possible time of a major foreshock or main shock,t _f,is found by the best fit of equation (1), or its integral, to step-like plots of versus time using successive estimates oft _fin linearized regressions until the maximum coefficient of determination,r ²,is obtained. Analyzed examples include sequences preceding earthquakes at Cremasta, Greece, 2/5/66; Haicheng, China 2/4/75; Oaxaca, Mexico, 11/29/78; Petatlan, Mexico, 3/14/79; and Central Chile, 3/3/85. In 29 estimates of main-shock time, made as the sequences developed, the errors in 20 were less than one-half and in 9 less than one tenth the time remaining between the time of the last data used and the main shock. Some precursory sequences, or parts of them, yield no solution. Two sequences appear to include in their first parts the aftershocks of a previous event; plots using the integral of equation (1) show that the sequences are easily separable into aftershock and foreshock segments. Synthetic seismic sequences of shocks at equal time intervals were constructed to follow equation (1), using four values ofn. In each series the resulting distributions of magnitudes closely follow the linear Gutenberg-Richter relation logN=a–bM, and the productn timesb for each series is the same constant. In various forms and for decades, equation (1) has been used successfully to predict failure times of stressed metals and ceramics, landslides in soil and rock slopes, and volcanic eruptions. Results of more recent experiments and theoretical studies on crack propagation, fault mechanics, and acoustic emission can be closely reproduced by equation (1). Rate-process theory and continuum damage mechanics offer leads toward understanding the physical processes.     

Application of the time-predictable model in Peninsular India for future seismic hazard assessment

It has been the belief among Earth scientists that the Peninsular Shield is aseismic, as the region attained stability long ago. However, the earthquake at Koyna (10 December 1967), Bhadrachalam (13 April 1969), Broach (23 March 1970), Hyderabad (30 June 1983), Khillari (30 September 1993), Jabalpur (22 May 1997), Gujarat (26 January 2001), and additional ones of smaller magnitudes, altered this concept. This area has experienced many widely distributed shallow earthquakes, some of them having large magnitudes. It is now widely accepted that seismic activity still continues with moderate events. Therefore, a need has arisen to take into consideration recent seismological data to assess the future seismic status of Peninsular India. Earthquake generation model has been studied to develop the statistical relations with surface wave magnitude (M _S ≥ 4.5). Five seismogenic sources showing clustering of earthquakes and including at least three main shocks of magnitude 4.5 ≤ M _S ≤ 6.5 giving two repeat times, have been identified. It is mainly based on the so-called “regional time-predictable model”. For the considered region it is observed that the time interval between two consecutive main shocks depends on the preceding main shock magnitude (M _p ) not on the following main shocks magnitude M _f suggesting the validity of time predictable model in the region.     

Long-term earthquake prediction in the Marmara region based on the regional time- and magnitude-predictable model

In order to estimate the recurrence intervals for large earthquakes that occurred in the Marmara region, this region, limited with the coordinates of 39°–42°N, 25°–32°E, has been separated into seven seismogenic sources on the basis of certain seismological criteria, and regional time- and magnitude-predictable model has been applied for these sources. Considering the interevent time between successive mainshocks, the following two predictive relations were computed: log T _t = 0.26 M _min + 0.06 M _p –0.56 log M ₀ + 13.79 and M _f = 0.63 M _min ? 0.07 M _p + 0.43 log M ₀ ? 7.56. Multiple correlation coefficient and standard deviation have been computed as 0.53 and 0.35 for the first relation and 0.66 and 0.39 for the second relation, respectively. On the basis of these relations and using the occurrence time and magnitude of the last mainshocks in each seismogenic source, the probabilities of occurrence P(Δt) of the next mainshocks during the next five decades and the magnitude of the expected mainshocks were determined.     

The time of occurrence and the magnitude of the largest aftershock over India

Summary The time of occurrence and the magnitude of the largest aftershock in relation to the main shock have been studied for India and its neighbourhood based on the USCGS data during the years 1963–1971. It is found that the largest aftershock occurs within 2 hours after the main shock in about 50% of the cases and frequency of occurrencen(t) of the largest aftershocks decreases hyper-bolically with the intervalt after the main event and could be represented by a law of the formn(t)=At ^–h whereA andh are constants. The probability of occurrence of the largest aftershock within 2 hours of the main shock is found to be higher over island are regions of the world. The difference (M ₀–M ₁) of the magnitude of the largest aftershockM ₁ to that of the main shockM ₀ as a measure of aftershock activity does not show any marked regional variation over India and its neighbourhood, as was reported by Mogifor Japan. Examination of the values ofM ₁/M ₀ and the constantb in Gutenberg-Richter's frequency magnitude relationship reveals a range of variation in both; high values ofM ₁/M ₀ have been found to be associated with high values ofb in many tectonic earthquakes and thus not, restricted to reservoir associated seismic activity.     

Analysis of foreshock sequence of the 1975 Haicheng earthquake of M S 7.3

The “earthquake nucleation” is discussed in this paper. The acceleration is a property of the nucleation phase and is a necessary condition of earthquake instability too. If the acceleration property of this nucleating process is described by the equation dΘ/dt=C/(t _f−t) ⁿ , the process can be summarized briefly that the rate of cumulative seismic release is proportional to the inverse power of the remaining time to failure. Based on this principle, the foreshock sequence of the 1975 Haicheng earthquake withM _S7.3, was analysed backward. It is stated clearly that the time-to-failure and magnitude of the mainshock can be predicted successfully if the coefficientr ² attains to the maximum. In the estimation of mainshock time, the error can generally be less than, or far less than, one-half the remaining time between the time of the last used data point and the mainshock. Contribution No. 95A0024, Institute of Geophysics, SSB, China. This study is the "Eighth-Five" contract project ofSSB.     

Intrinsic and Scattering Seismic Attenuation in Southwestern Anatolia

The intrinsic dissipation and scattering attenuation in southwestern (SW) Anatolia, which is a tectonically active region, is studied using the coda waves. First the coda quality factor (Q_c) assuming single scattering is estimated from the slope of the coda-wave amplitude decay. Then the Multiple Lapse Time Window (MLTW) analysis is performed with a uniform earth model. Three non-overlapping temporal data windows are used to calculate the scattered seismic energy densities against the source-receiver distances, which, in turn, are used to calculate separate estimates of the intrinsic and scattering factors. In order to explore the frequency dependency, the observed seismograms are band pass-filtered at the center frequencies of 0.75, 1.5, 3.0, 6.0 and 12.0. The scattering attenuation (Q_s⁻¹) is found lower than the intrinsic attenuation (Q_i⁻¹) at all frequencies except at 0.75 Hz where the opposite is observed. Overall the intrinsic attenuation dominates over the scattering attenuation in the SW Anatolia region. The integrated energy curves obtained for the first energy window (i.e., 0–15 s) are somewhat irregular with distance while the second (i.e., 15–30 s) and third (i.e., 30–45 s) data windows exhibit more regular change with distance at most frequencies. The seismic albedo B₀ is determined as 0.61 at 0.75 Hz and 0.34 at 12.0 Hz while the total attenuation factor denoted by L_e⁻¹ changes in the range 0.034–0.017. For the source-station range 20–180 km considered the scattering attenuation is found strongly frequency dependent given by the power law Q_s⁻¹ = 0.010*f^−1.508. The same relations for Q_i⁻¹, Q_t⁻¹ (total), Q_c⁻¹ and (expected) hold as Q_i⁻¹ = 0.0090*f^−1.17, Q_t⁻¹ = 0.019*f^−1.31, Q_c⁻¹ = 0.008*f^−0.84 and respectively. Compared to the other attenuation factors Q_c⁻¹ and are less dependent on the frequency.     

2020年云南巧家MS5.0地震微震检测及发震构造初步探讨

利用匹配定位方法对2020年5月18日云南巧家Ms5.0地震震后24h震源附近台站.记录的连续波形进行遗漏地震扫描和定位,共识别出327个地震事件,约为台网目录的2.4倍,最小完整震级由最初的ML1.9降至ML1.1.随后,依据最新目录计算了震后震源区的b值,并结合余震展布形态,初步分析此次地震发震构造.研究结果显示,...     

Re-visiting large historical earthquakes in the Colombian Eastern Cordillera

A re-assessment of the historic seismicity of the central sector of the Colombian Eastern Cordillera (EC) is made by revision of bibliographic sources, by calibration with modern instrumental earthquakes, and by interpretations in terms of current knowledge of the tectonics and seismicity of the region. Throughout the process we have derived an equation to estimate M_w for shallow crustal earthquakes in Colombia using the length of isoseismal VIII, L_VIII:

Self-organized Fractal Seismicity of Reservoir Triggered Earthquakes in the Koyna-Warna Seismic Zone,Western India

— Analysis of the Koyna-Warna earthquake catalog (1968–1996) shows that on an average there is a positive correlation between the b value (decrease) and fractal dimensions (decrease in both D₂^s and D₂^t) of earthquake epicenters 0.5 and 2.5 years prior to 1973 (M5.2) and 1980 (M5.5) events, respectively, except a negative correlation for about five years (1988–1993) prior to the 1993/1994 sequence (M5.4). This positive correlation indicates a weaker clustering, or that the epicenters tend to fill the two-dimensional plane. While the origin of the negative correlation seems to be that during periods of large events (low b value), there is strong clustering around the main shock epicenter (high fractal dimension). Interestingly, during the last year (1995–1996) of the studied period both the b value and correlation dimensions rose significantly, suggesting that stress release occurs through increased levels of low magnitude and increasingly scattered seismicity, suggesting an increased risk of larger magnitude events. Incidentally, during 2000 three earthquakes of magnitude M 5.0, one earthquake of M 4.0, 45 earthquakes of magnitude M 3.0–3.9, and several thousand earthquakes of M < 3 have occurred in the region. Thus it can be inferred that at local scales the relationship yields both positive and negative correlation that appears to be controlled by different modes of failure within the active fault complex.Acknowledgement. The authors are grateful to Dr. B.K. Rastogi of NGRI for providing the catalog of Koyna earthquakes and for useful scientific discussions. The comments of Dr. I. G. Main have improved the quality of paper for which we extend to him our sincere thanks. One of the authors (AOM) thanks the Third World Academy of Science and the Council of Scientific and Industrial Research, India for the Postdoctoral Fellowship award under which this work was carried out.     

Surface fault traces,fault plane solution and spatial distribution of the aftershocks of the September 13, 1986 earthquake of Kalamata (Southern Greece)

A shallow earthquake ofM _S=6.2 occurred in the southern part of the Peloponnesus, 12 km north of the port of the city of Kalamata, which caused considerable damage. The fault plane solution of the main shock, geological data and field observations, as well as the distribution of foci of aftershocks, indicate that the seismic fault is a listric normal one trending NNE-SSW and dipping to WNW. The surface ruptures caused by the earthquake coincide with the trace of a neotectonic fault, which is located 2–3 km east of the city of Kalamata and which is related to the formation of Messiniakos gulf during the Pliocene-Quaternary tectonics. Field observations indicate that the earthquake is due to the reactivation of the same fault.A three-days aftershock study in the area, with portable seismographs, recorded many aftershocks of which 39 withM _S1.7 were very well located. The distribution of aftershocks forms two clusters, one near the epicenter of the main shock in the northern part of the seismogenic volume, and the other near the epicenter of the largest aftershock (M _S=5.4) in the southern part of this volume. The central part of the area lacks aftershocks, which probably indicates that this is the part of the fault which slipped smoothly during the earthquake.     

Ground-motion Attenuation Relation from Strong-motion Records of the 2001 Mw 7.7 Bhuj Earthquake Sequence (2001–2006), Gujarat, India

Predictive relations are developed for peak ground acceleration (PGA) from the engineering seismoscope (SRR) records of the 2001 M_w 7.7 Bhuj earthquake and 239 strong-motion records of 32 significant aftershocks of 3.1 ≤ M_w ≤ 5.6 at epicentral distances of 1 ≤ R ≤ 288 km. We have taken advantage of the recent increase in strong-motion data at close distances to derive new attenuation relation for peak horizontal acceleration in the Kachchh seismic zone, Gujarat. This new analysis uses the Joyner-Boore’s method for a magnitude-independent shape, based on geometrical spreading and anelastic attenuation, for the attenuation curve. The resulting attenuation equation is,

Determining the time and space domains for estimating b value in seismic zoning

To take the seismic zone that includes the great shock with M _s≥8.2 as the statistical unit of estimating b value can often lead to more large variance, because the seismogenic zone of the great shock with M _s≥8.2 are larger than that delineated in general seismic zone. Two-level statistical units are considered in this paper. The seismic province is the first level unit that is suitable for group of earthquakes including the great shock of M _s≥8.5. A seismic province can be divided into several seismic zones. They can be taken as the second level unit for group of quakes in which the super magnitude of the greatest shock do not exceed 8. Because of the nonstationarity in time of seismic activity, the unbalancedness of data and differential of seismic temporal series feature in different areas need to be considered when we select the time period for estimating b value. According to local conditions, the time period is selected at one’s discretion in order to reflect seismicity level of this statistical unit in future 100 years. Contribution No. 98A02039, Institute of Geophysics, China Seismological Bureau, China.     

A study on the focal parameters of Shidian M =5.9 earthquake sequences in 2001

On the basis of about 300 earthquake wave forms observed in the Shidian M _S=5.9 sequences on April 12, 2001 recorded in Kunming Digital Seismic Network, the spectra of shear wave have been used to estimate the focal parameters of these earthquake sequences. The results show that within the magnitude range of 1.5–5.3, the seismic moments are 10¹⁰–10¹⁶ N·m, the corner frequencies are 0.2–0.8 Hz, radii of the focal rupture are 200–2 500 m and the stress drops are 0.1×10⁵–20×10⁵Pa. Through the statistical analyses of variation of corner frequency f _c and stress drop Δσ with time, it is discovered that the average corner frequency of the foreshock sequences is obviously lower than that of the aftershock sequences. Contrarily, the average stress drops Δσ of the foreshock sequences are clearly higher than that of the aftershocks. It is considered that these variation characteristics of average corner frequency and stress drops before and after the main shock have index significance to the precursory information before a strong earthquake. The higher stress drops for the foreshock sequences show that the higher shear stresses have been stored in the area of main shock. After the main shock, most of the stresses have been released, so the aftershock sequences show a rupture process of lower stresses. Foundation item: Scientific and Technological Key Project of Yunnan Province (2001NG46)     

Aftershocks of February 19, 1968 earthquake in Northern Aegean Sea and related problems

Summary The aftershock activity associated with the North Aegean Sea earthquake of February 19, 1968 (M _s =7.1) has been studied with emphasis on the time and magnitude distributions. Regarding the focal mechanism it was noted that the sense of first motion remains rather unchanged throughout the sequence.The secondary series are not distinct in space or time or mechanism from the primary sequence. Although the main shock was followed by almost 3000 aftershocks (M _L 2.1) which form the basis for this study, principal strain has been released by the main shock which is responsible for the 82 per cent of total strain release in the sequence.     

Monitoring the West Bohemian earthquake swarm in 2008/2009 by a temporary small-aperture seismic array

The most recent intense earthquake swarm in West Bohemia lasted from 6 October 2008 to January 2009. Starting 12 days after the onset, the University of Potsdam monitored the swarm by a temporary small-aperture seismic array at 10 km epicentral distance. The purpose of the installation was a complete monitoring of the swarm including micro-earthquakes (M _L < 0). We identify earthquakes using a conventional short-term average/long-term average trigger combined with sliding-window frequency-wavenumber and polarisation analyses. The resulting earthquake catalogue consists of 14,530 earthquakes between 19 October 2008 and 18 March 2009 with magnitudes in the range of − 1.2 ≤ M _L ≤ 2.7. The small-aperture seismic array substantially lowers the detection threshold to about M _c = − 0.4, when compared to the regional networks operating in West Bohemia (M _c > 0.0). In the course of this work, the main temporal features (frequency–magnitude distribution, propagation of back azimuth and horizontal slowness, occurrence rate of aftershock sequences and interevent-time distribution) of the recent 2008/2009 earthquake swarm are presented and discussed. Temporal changes of the coefficient of variation (based on interevent times) suggest that the swarm earthquake activity of the 2008/2009 swarm terminates by 12 January 2009. During the main phase in our studied swarm period after 19 October, the b value of the Gutenberg–Richter relation decreases from 1.2 to 0.8. This trend is also reflected in the power-law behavior of the seismic moment release. The corresponding total seismic moment release of 1.02×10¹⁷ Nm is equivalent to M _L,max = 5.4.     

Regional Variations of the <Emphasis Type="Italic">ω</Emphasis>-upper Bound Magnitude of GIII Distribution in the Iranian Plateau

The Iranian Plateau does not appear to be a single crustal block, but an assemblage of zones comprising the Alborz—Azerbaijan, Zagros, Kopeh—Dagh, Makran, and Central and East Iran. The Gumbel’s III asymptotic distribution method (GIII) and maximum magnitude expected by Kijko—Sellevoll method is applied in order to check the potentiality of the each seismogenic zone in the Iranian Plateau for the future occurrence of maximum magnitude (M_max). For this purpose, a homogeneous and complete seismicity database of the instrumental period during 1900–2012 is used in 29 seismogenic zones of the examined region. The spatial mapping of hazard parameters (upper bound magnitude (ω), most probable earthquake magnitude in next 100 years (M₁₀₀) and maximum magnitude expected by maximum magnitude estimated by Kijko—Sellevoll method (max M_{K ? S}^max) reveals that Central and East Iran, Alborz and Azerbaijan, Kopeh—Dagh and SE Zagros are a dangerous place for the next occurrence of a large earthquake.