Increased correlation range of seismicity before large events manifested by earthquake chains

“Earthquake chains” are clusters of moderate-size earthquakes which extend over large distances and are formed by statistically rare pairs of events that are close in space and time (“neighbors”). Earthquake chains are supposed to be precursors of large earthquakes with lead times of a few months. Here we substantiate this hypothesis by mass testing it using a random earthquake catalog. Also, we study stability under variation of parameters and some properties of the chains. We found two invariant parameters: they characterize the spatial and energy scales of earthquake correlation. Both parameters of the chains show good correlation with the magnitudes of the earthquakes they precede. Earthquake chains are known as the first stage of the earthquake prediction algorithm reverse tracing of precursors (RTP) now tested in forward prediction. A discussion of the complete RTP algorithm is outside the scope of this paper, but the results presented here are important to substantiate the RTP approach.     

Seismic reversal pattern for the 1999 Chi-Chi, Taiwan, MW 7.6 earthquake

We in this study have calculated the standard normal deviate Z-value to investigate the variations in seismicity patterns in the Taiwan region before and after the Chi-Chi earthquake. We have found that the areas with relatively high seismicity in the eastern Taiwan became abnormally quiet before the Chi-Chi earthquake while the area in the central Taiwan with relatively low seismicity showed unusually active. Such a spatially changing pattern in seismicity strikingly demonstrates the phenomenon of “seismic reversal,” and we here thus present a complete, representative cycle of “seismic reversal” embedding in the changes of seismicity patterns before and after the Chi-Chi earthquake.     

Estimation of Coulomb stress changes and induced earthquake occurrences prediction in the region of Three Gorges Reservoir,China

In the region of Three Gorges Reservoir (TGR) in China, there has been occurrence of several frequent earthquakes of moderate intensity since reservoir impounding occurred in 2003. These earthquakes are generally believed to be induced by reservoir impoundment and water-level variations. Usually, the geo-stress will change, when natural earthquakes occur. Following this principle, this paper adopted the rate and state theory to simulate and estimate Coulomb stress changes in the TGR region and obtained the pattern of Coulomb stress changes with time and the event sequence as well as the distribution of Coulomb stress changes in space. First, the TGR regional catalogue was analyzed and processed, leading to quantification of the magnitude of completeness and all of the parameters that are used in the stress–seismicity inversion process, including the reference seismicity rates, characteristic relaxation time, fault constitutive parameters, and stress rates. Second, the temporal evolution of the stress changes in different time windows was computed and analyzed, and it was found that there is an association between the Coulomb stress changes and rates of increase in the cumulative number of earthquakes. In addition, the earthquake occurred in November 2008 (M _S = 4.1) was analyzed and attempted to simulate the distribution of stress changes in space through the stress–seismicity inversion model. The results proved that the modeled area coincides with the historical area of earthquakes that occurred after 2008. Finally, a prediction was made about the earthquake productivity rates after 2015, which showed a declining earthquake rate over time that ultimately returned to the background seismicity. This result is essentially in agreement with Omori’s law. To conclude, it is rational to use the stress-inversion method to analyze the relationship between induced earthquake seismicity and local stress changes as well as to simulate the area of earthquake occurrence and productivity rates of reservoir-induced earthquakes.     

鄂尔多斯块体周缘地区近期地震活动性与汶川地震应力触发作用的关系

采用β统计对汶川地震前后鄂尔多斯块体周缘地区的地震活动率进行了空间扫描分析,并采用JiChen的震源破裂模型计算了汶川地震产生的库仑破裂应力变化,以研究鄂尔多斯块体周缘地区近期地震活动性与汶川地震应力触发作用的关系。结果发现,鄂尔多斯块体西南缘弧形断裂束的南东段与南缘渭河盆地的地震活动率在汶川地震后提高显著,其他区域的地震活动率没有明显提高,库仑破裂应力计算得到两个区域的应力变化范围分别为0.005～0.02 MPa和0.001～0.01 MPa,表明汶川地震有可能触发了这两个区域的地震活动。鄂尔多斯块体东缘的山西断陷带处于库仑破裂应力计算的应力增加区,应力变化范围为0～0.012 MPa,2009年3月以来发生的4次ML4.5～5.2级强有感至微破坏地震有可能被汶川地震所延迟触发。b值、地震能量释放率与空间相关距离SCL等地震活动性参数随时间变化扫描结果显示,该区域可能处于不断趋近高应力累积的状态,其未来大震有可能提前发生。西缘地区为应力减小区,其目前的地震活动处于正常水平状态。     

Spatio-temporal evolution of Coulomb stress in the Pacific slab inverted from the seismicity rate change and its tectonic interpretation in Hokkaido, Northern Japan

We estimated spatio-temporal evolution of Coulomb stress within the subducted Pacific slab in Hokkaido from the analysis of seismicity rate change. For this purpose we used earthquake catalog from the Institute of Seismology and Volcanology (ISV), Hokkaido University for the period 1993/4/1–2006/12/31 after relocating to compensate location errors due to the heterogeneous P- and S-wave structure beneath Hokkaido. We found that spatial pattern of Coulomb stress change inverted from the seismicity rate change is comparable with static change in Coulomb stress estimated from dislocation models. Our results and analyses reveal important insights on spatio-temporal pattern of deformation of the subducted Pacific slab in terms of Coulomb stress change. We found that the 2003 Tokachi Oki earthquake (Mw = 8.0) pervasively perturbed Coulomb stress in a regional scale with a significant impact to trigger the 2004 Kushiro Oki earthquake. The 2004 Kushiro Oki earthquake (Mw = 7.0) is another significant stressing event that changed the pattern of Coulomb stress in the area. We found that stressing events with magnitude smaller than 7.0 has minimal impact on Coulomb stress change in the Pacific slab. Similarly, comparatively deep focused large earthquakes could not change Coulomb stress significantly. Further the pattern of Coulomb stress change after the 2003 Tokachi Oki earthquake correlates the pattern of afterslip distribution in Hokkaido.     

Seismicity patterns before the M=5.8 2002, Palermo (Italy) earthquake: seismic quiescence and accelerating seismicity

Seismic quiescence and accelerating seismic energy release are considered as possible spatio-temporal patterns of the preparation process of the 6 September 2002 Palermo, Italy, earthquake (M 5.8). The detailed properties of the quiescence are analyzed applying the RTL algorithm. The RTL algorithm is based on the analysis of the RTL prognostic parameter, which is designed in such a way that it has a negative value if, in comparison with long-term background, there is a deficiency of events in the time–space vicinity of the tested point. The RTL parameter increases if activation of seismicity takes place. The RTL algorithm identified that a seismic quiescence started from the beginning of November 2001 and reached its minimum at the end of May 2002. The Palermo 2002 earthquake occurred 2 months after the RTL parameter restored its long-term background level. The application of a log-periodic time-to-failure model gives a “predicted” (in retrospect) magnitude M=6.2 main shock on 5 May 2002.     

Global strain rates in western to central Himalayas and their implications in seismic hazard assessment

The Himalayas has experienced varying rates of earthquake occurrence in the past in its seismo-tectonically distinguished segments which may be attributed to different physical processes of accumulation of stress and its release, and due diligence is required for its inclusion for working out the seismic hazard. The present paper intends to revisit the various earthquake occurrence models applied to Himalayas and examines it in the light of recent damaging earthquakes in Himalayan belt. Due to discordant seismicity of Himalayas, three types of regions have been considered to estimate larger return period events. The regions selected are (1) the North-West Himalayan Fold and Thrust Belt which is seismically very active, (2) the Garhwal Himalaya which has never experienced large earthquake although sufficient stress exists and (3) the Nepal region which is very seismically active region due to unlocked rupture and frequently experienced large earthquake events. The seismicity parameters have been revisited using two earthquake recurrence models namely constant seismicity and constant moment release. For constant moment release model, the strain rates have been derived from global strain rate model and are converted into seismic moment of earthquake events considering the geometry of the finite source and the rates being consumed fully by the contemporary seismicity. Probability of earthquake occurrence with time has been estimated for each region using both models and compared assuming Poissonian distribution. The results show that seismicity for North-West region is observed to be relatively less when estimated using constant seismicity model which implies that either the occupied accumulated stress is not being unconfined in the form of earthquakes or the compiled earthquake catalogue is insufficient. Similar trend has been observed for seismic gap area but with lesser difference reported from both methods. However, for the Nepal region, the estimated seismicity by the two methods has been found to be relatively less when estimated using constant moment release model which implies that in the Nepal region, accumulated strain is releasing in the form of large earthquake occurrence event. The partial release in second event of May 2015 of similar size shows that the physical process is trying to release the energy with large earthquake event. If it would have been in other regions like that of seismic gap region, the fault may not have released the energy and may be inviting even bigger event in future. It is, therefore, necessary to look into the seismicity from strain rates also for its due interpretation in terms of predicting the seismic hazard in various segments of Himalayas.     

A worldwide test of three long-term premonitory seismicity patterns—a review

Worldwide analysis of the clustering of earthquakes has lead to the hypothesis that the occurrence of abnormally large clusters indicates an increase in probability of a strong earthquake in the next 3–4 years within the same region. Three long-term premonitory seismicity patterns, which correspond to different non-contradictory definitions of abnormally large clusters, were tested retrospectively in 15 regions. The results of the tests suggest that about 80% of the strongest earthquakes can be predicted by monitoring these patterns.Most of results concern pattern B (“burst of aftershocks”) i.e. an earthquake of medium magnitude with an abnormally large number of aftershocks during the first few days. Two other patterns, S and Σ often complement pattern B and can replace it in some regions where the catalogs show very few aftershocks.The practical application of these patterns is strongly limited by the fact that neither the location of the coming earthquake within the region nor its time of occurrence within 3–4 years is indicated. However, these patterns present the possibility of increasing the reliability of medium and short-term precursors; also, they allow activation of some important early preparatory measures.The results impose the following empirical constraint on the theory of the generation of a strong earthquake: it is preceded by abnormal clustering of weaker earthquakes in the space-time-energy domain; corresponding clusters are few but may occur in a wide region around the location of the coming strong earthquake; the distances are of the same order as for the other reported precursors.     

Forecasting rupture dimension using the pattern informatics technique

The Pattern Informatics (PI) technique [Tiampo, K.F., Rundle, J.B., McGinnis, S., Gross, S., Klein, W., 2002. Mean-field threshold systems and phase dynamics: An application to earthquake fault systems, Europhys. Lett., 60, 481–487] is founded on the premise that changes in the seismicity rate are a proxy for changes in the underlying stress. This new approach to the study of seismicity quantifies its local and regional space–time patterns and identifies regions of local quiescence or activation. Here we use a modification of the PI method to quantify localized changes surrounding the epicenters of large earthquakes in California in an attempt to objectively quantify the rupture zones of these upcoming events. We show that this method can be used to forecast the size and magnitude of future earthquakes.     

A statistical procedure for the analysis of seismotectonically induced hydrochemical signals: A case study from the Eastern Carpathians, Romania

Changes in the stress field of an aquifer system induced by seismotectonic activity may change the mixing ratio of groundwaters with different compositions in a well, leading to hydrochemical signals which in principle could be related to discrete earthquake events. Due to the complexity of the interactions and the multitude of involved factors the identification of such relationships is a difficult task. In this study we present an empiric statistical approach suitable to analyse if there is an interdependency between changes in the chemical composition of monitoring wells and the regional seismotectonic activity of a considered area. To allow a rigorous comparison with hydrochemistry the regional earthquake time series was aggregated into an univariate time series. This was realized by expressing each earthquake in form of a parameter “e”, taking into consideration both energetic (magnitude of a seismic event) and spatial parameters (position of epi/hypocentrum relative to the monitoring site). The earthquake and the hydrochemical time-series were synchronised aggregating the e-parameters into “earthquake activity” functions E, which takes into account the time of sampling relative to the earthquakes which occurred in the considered area. For the definition of the aggregation functions a variety of different “e” parameters were considered. The set of earthquake functions E was grouped by means of factor analysis to select a limited number of significant and representative earthquake functions E to be used further on in the relation analysis with the multivariate hydrochemical data set. From the hydrochemical data a restricted number of hydrochemical factors were extracted. Factor scores allow to represent and analyse the variation of the hydrochemical factors as a function of time. Finally, regression analysis was used to detect those hydrochemical factors which significantly correlate with the aggregated earthquake functions.This methodological approach was tested with a hydrochemical data set collected from a deep well monitored for two years in the seismically active Vrancea region, Romania. Three of the hydrochemical factors were found to correlate significantly with the considered earthquake activities. A screening with different time combinations revealed that correlations are strongest when the cumulative seismicity over several weeks was considered. The case study also showed that the character of the interdependency depends sometimes on the geometrical distribution of the earthquake foci. By using aggregated earthquake information it was possible to detect interrelationships which couldn't have been identified by analysing only relations between single geochemical signals and single earthquake events. Further on, the approach allows to determine the influence of different seismotectonic patterns on the hydrochemical composition of the sampled well. The method is suitable to be used as a decision instrument in assessing if a monitoring site is suitable or not to be included in a monitoring net within a complex earthquake prediction strategy.     

Earthquake prediction: 20 years of global experiment

Earthquake professionals have for many decades recognized the benefits to society from reliable earthquake predictions, but uncertainties regarding source initiation, rupture phenomena, and accuracy of both the timing and magnitude of the earthquake occurrence have oftentimes seemed either very difficult or impossible to overcome. The problem is that most of these methods cannot be adequately tested and evaluated either because of (a) lack of a precise definition of “prediction” and/or (b) shortage of data for meaningful statistical verification. This is not the case for the pattern recognition algorithm M8 designed in 1984 for prediction of great, Magnitude 8, earthquakes, hence its name. By 1986, the algorithm was rescaled for applications aimed at smaller magnitude earthquakes, down to M5+ range, and since then it has become a useful tool for systematic monitoring of seismic activity in a number of test seismic regions worldwide. After confirmed predictions of both the 1988 Spitak (Armenia) and the 1989 Loma Prieta (California) earthquakes, a “rigid test” to evaluate the efficiency of the intermediate-term middle-range earthquake prediction technique has been designed. Since 1991, each half-year, the algorithm M8 alone and in combination with its refinement MSc has been applied in a real-time prediction mode to seismicity of the entire Earth, and this test outlines, where possible, the areas in the two approximations where magnitude 8.0+ and 7.5+ earthquakes are most likely to occur before the next update. The results of this truly global 20-year-old experiment are indirect confirmations of the existing common features of both the predictability and the diverse behavior of the Earth’s naturally fractal lithosphere. The statistics achieved to date prove (with confidence above 99 %) rather high efficiency of the M8 and M8-MSc predictions limited to intermediate-term middle- and narrow-range accuracy. These statistics support the following general conclusions—(1) precursory seismic patterns do exist; (2) the size of an area where precursory seismic patterns show up is much larger than that of the source zone of the incipient target earthquake; (3) many precursory seismic patterns appear to be similar, even in regions of fundamentally different tectonic environments; and (4) some precursory seismic patterns are analogous to those in advance of extreme catastrophic events in other complex nonlinear systems (e.g., magnetic storms, solar flares, “starquakes”, etc.)—that are of high importance for further searches of the improved earthquake forecast/prediction algorithms and methods.     

Continued seismic activity at the Koyna reservoir site, India

Following the impounding of the Shivaji Sager Lake in 1962, tremors became prevalent in the Koyna region, considered previously to be aseismic. During ensuing years the tremor frequency appears to have been dependent on the rate of increase of water level, maximum water level reached, and the period for which high levels were retained. This culminated in a burst of seismic activity from September 1967 to January 1968 following the record water levels in the reservoir and included the earthquake of September 13, 1967 with magnitude 5.5 and the damaging December 10, 1967 earthquake of magnitude 6.0. During the next five years water levels were kept low and no significant earthquakes occurred subsequent to the October 29, 1968 earthquake of magnitude 5.

The reservoir was filled to maximum capacity during September 1973 and this was followed by a conspicuous increase in seismic activity which included an earthquake of magnitude 5.1 on October 17, 1973. However, seismic activity during 1973 was much less severe than that of 1967. This relative decrease in seismicity may indicate that (a) the “threshhold level” for relatively large magnitude earthquakes had increased; (b) a major portion of the accumulated strains had been released; and/or (c) the importance of the longer period of high loading in 1967. Similar observations have been made at other seismically active reservoir sites.     

The 1998 Hida Mountain, Central Honshu, Japan, earthquake swarm: Double-difference event relocation, frequency–magnitude distribution and Coulomb stress changes

By using the double-difference relocation technique, we have determined the fine structure of seismicity during the 1998 Hida Mountain earthquake swarm. The distribution of seismic activity defines two main directions (N–S and E–W) that probably correspond to the regional stress pattern. The detailed structure of seismicity reveals intense spatio-temporal clustering and earthquake lineations. Each cluster of events contains a mainshock and subsequent aftershock activity that decays according to the Omori law. The seismicity and the b-value temporal and spatial patterns reflect the evolution of the static stress changes during the earthquake swarm. About 80% of the swarm's best-relocated events occur in regions of increased ΔCFF. The smaller value of b found in the northern part of the swarm region and a larger b-value observed to the south, for the same period of time, could be well explained by the static stress changes caused by the larger events of the sequence. We argue that the state of stress in the crust is the main factor that controls the variation of b-value.     

Finite element models to represent seismic activity of the Indian plate

Quantification of seismic activity is one of the most challenging problems faced by earthquake engineers in probabilistic seismic hazard analysis. Currently, this problem has been attempted using empirical approaches which are based on the regional earthquake recurrence relations from the available earthquake catalogue. However, at a specified site of engineering interest, these empirical models are associated with large number of uncertainties due to lack of sufficient data. Due to these uncertainties, engineers need to develop mechanistic models to quantify seismic activity. A wide range of techniques for modeling continental plates provides useful insights on the mechanics of plates and their seismic activity. Among the different continental plates, the Indian plate experiences diffused seismicity. In India, although Himalaya is regarded as a plate boundary and active region, the seismicity database indicates that there are other regions in the Indian shield reporting sporadic seismic activity. It is expected that mechanistic models of Indian plate, based on finite element method, simulate stress fields that quantify the seismic potential of active regions in India. This article explores the development of a finite element model for Indian plate by observing the simulated stress field for various boundary conditions, geological and rheological conditions. The study observes that the magnitude and direction of stresses in the plate is sensitive to these conditions. The numerical analysis of the models shows that the simulated stress field represents the active seismic zones in India.     

Global-scale analysis of seismic activity prior to 2004 Sumatra-Andaman mega-earthquake

Following our experience in intermediate-term monitoring of seismicity before large earthquakes worldwide and in some regions we apply a similar approach to the analysis of seismicity on global scale several decades before the December 26, 2004 Sumatra–Andaman mega-earthquake. The Earth lithosphere is being considered as a single whole, representing the ultimate level of the complex Earth's hierarchy. The study brings up the following questions: Are there any anomalies of the global seismicity behaviour observed in advance of the mega-earthquake? Do these anomalies correspond to the global scale seismic patterns similar to those detected on the regional scale before great, major, and strong events? In other words, does the Earth lithosphere considered as a single whole show up an approach of the mega-earthquake in the way typical for events on the smaller levels of hierarchy? The results of our investigation favour the positive answer to the questions. Specifically we found that during the decade before the Sumatra–Andaman mega-earthquake the Earth lithosphere reveals, at least in intermediate-term scale, classical symptoms of instability, which can be depicted by known precursory seismic patterns. These are: (i) transformation of frequency–magnitude distribution, (ii) change in the rate of seismic activity, and (iii) depth redistribution of activity. Moreover changes of dependencies between magnitudes of different types are detected at the same time. The observed global scale patterns of collective behaviour of seismicity may indicate the state of criticality of the Earth lithosphere before the Sumatra–Andaman mega-earthquake.     

Temporal analysis of croatian seismogenic zones to improve earthquake magnitude prediction

The prediction of earthquakes is a task of utmost difficulty that has been addressed in many different ways. However, an initial definition of the area of interest is needed, with adequate catalogs. In this work, different seismogenic zones proposals in the Republic of Croatia are studied, in terms of predictability. Such zones have been characterized with widely used seismicity parameters. Later, studies based on training and test sets properties as well as the quality of the data involved are carried out. The studies presented in this work analyze the prediction performance across the earthquake magnitude time series of the target seismogenic zones. Results show that specific prediction techniques could be used in some zones to improve earthquake magnitude prediction.     

Seismic hazard and probability assessment of Kashmir valley,northwest Himalaya,India

Seismic hazard analysis of the northwest Himalayan belt was carried out by using extreme value theory (EVT). The rate of seismicity (a value) and recurrence intervals with the given earthquake magnitude (b value) was calculated from the observed data using Gutenberg–Richter Law. The statistical evaluation of 12,125 events from 1902 to 2017 shows the increasing trend in their inter-arrival times. The frequency–magnitude relation exhibits a linear downslope trend with negative slope of 0.8277 and positive intercept of 4.6977. The empirical results showed that the annual risk probability of high magnitude earthquake M?≥?7.7 in 50 years is 88% with recurrence period of 47 years, probability of M?≤?7.5 in 50 years is 97% with recurrence period of 27 years, and probability of M?≤?6.5 in 50 years is 100% with recurrence period of 4 years. Kashmir valley, located in the NW Himalaya, encompasses a peculiar tectonic and structural setup. The patterns of the present and historical seismicity records of the valley suggest a long-term strain accumulation along NNW and SSE extensions with the decline in the seismic gap, posing a potential threat of earthquakes in the future. The Kashmir valley is characterized by the typical lithological, tectono-geomorphic, geotechnical, hydrogeological and socioeconomic settings that augment the earthquake vulnerability associated with the seismicity of the region. The cumulative impact of the various influencing parameters therefore exacerbates the seismic hazard risk of the valley to future earthquake events.     

Earthquake magnitude prediction by adaptive neuro-fuzzy inference system (ANFIS) based on fuzzy C-means algorithm

This paper investigates the prediction of future earthquakes that would occur with magnitude 5.5 or greater using adaptive neuro-fuzzy inference system (ANFIS). For this purpose, the earthquake data between 1950 and 2013 that had been recorded in the region with 2°E longitude and 4°N latitude in Iran has been used. Thereupon, three algorithms including grid partition (GP), subtractive clustering (SC) and fuzzy C-means (FCM) were used to develop models with the structure of ANFIS. Since the earthquake data for the specified region had been reported on different magnitude scales, suitable relationships were determined to convert the magnitude scales into moment magnitude and all records uniformed based on the relationships. The uniform data were used to calculate seismicity indicators, and ANFIS was developed based on considered algorithms. The results showed that ANFIS-FCM with a high accuracy was able to predict earthquake magnitude.     

Seismic and electrical precursors to the 17-1-1983, M7 kefallinia earthquake,Greece: Signatures of a SOC system

Prior to the 17-1-1983 event, the seismicity of the broader area of the Ionian islands and western Greece exhibited several phenomena interpretable in the context of a self-organised critical system with long range interactions. The regional seismic energy release exhibited power law acceleration towards the time of rupture, the numerical modelling of which yields a time-to-failure of 1983.1 ± 0.2. Time dependent changes were also observed in the b-values, assuming the form of monotonic increase that promptly reversed after the earthquake. This indicates the induction of instability to the region due to the earthquake preparation process, which is consistent with the critical point earthquake model. The critical point model predicts that failure is a co-operative effect occurring at small scale, and cascading from the microscopic to the macroscopic scale. This involves a crack propagation avalanche at the terminal phase of the seismic cycle, the time function of which has been modelled with a limited class of characteristic transient bay-like shapes, featuring a corner frequency and inverse power energy distribution law. Electrification processes due to crack propagation may generate an electrical precursor with similar characteristics. Such a potential precursor has been observed independently on 15-1-1983, approx. 120km from the epicentre. In consequence of our observations, we discuss a model relating seismicity and electrical precursors.