The present status of reservoir induced seismicity investigations with special emphasis on Koyna earthquakes

The status of Reservoir Induced Seismicity (RIS) has been reviewed periodically (Rothé, 1968, 1973; Gupta and Rastogi, 1976; Simpson, 1976; Packer et al., 1979). In the present paper, the significant work carried out during the last three years on RIS is reviewed.An earthquake of magnitude occurred on November 14, 1981 in the vicinity of Aswan Lake, Egypt, 17 years after the filling started in 1964. This event occurred 4 days after the seasonal maximum in the reservoir water level and was followed by a long sequence of aftershocks. Another event of magnitude occurred in the vicinity of Aswan Lake on August 20, 1982. Results of preliminary investigations indicate that this seismic activity is reservoir induced. Recent analyses of induced seismic events at Nurek Reservoir U.S.S.R., show that the second stage of filling during August to December 1976, increasing the maximum depth from 120 m to 200 m, was accompanied by an intense burst of shallow seismic activity. An outward migration from the centre of the reservoir, possibly associated with diffusion of pore pressure, is revealed by the temporal distribution of earthquake foci. A variety of investigations including the in situ measurement of tectonic stress, pore pressure, permeability, distribution of faults, etc., in addition to monitoring seismicity, have been undertaken in the vicinity of the Monticello Reservoir, South Carolina. The largest reservoir induced earthquake is predicted not to exceed magnitude 5.The Koyna Reservoir, India, continues to be the most outstanding example of RIS. Three earthquakes of magnitude 5 occurred in September 1980. Earthquakes of magnitude 4 occur frequently in the vicinity of Koyna, the latest being on February 5, 1983. Events that occurred during the period 1967–1973 have been relocated using better procedures and are found to be much shallower and the epicentres less diffused. Location of 12 earthquakes of M_s 4.0, their foreshocks and aftershocks, that occurred during 1973–1976, composite focal mechanism solutions and related studies are consistent with the delineation of a N-S trending fault through the reservoir area. In a couple of interesting studies it has been demonstrated that earthquakes of magnitude 5.0 in the Koyna region are usually preceded by several magnitude 4 earthquakes in the preceding fortnight. Also, a rate of loading of Koyna reservoir of at least 40 ft/week appears to be a necessary, although not sufficient, condition for the occurrence of magnitude 5 earthquakes. Smooth filling/emptying appears to be the key to reduce the hazard of RIS.A map and a table of the reported cases of reservoir induced changes in seismicity through 1982 have been compiled.     

Premonitory crustal deformations, strains and seismotectonic features (b-values) preceding Koyna earthquakes

There have been instances of premonitory variations in tilts, displacements, strains, telluric current, seismomagnetic effects, seismic velocities ( V_p, V_s) and their ratio (V_p/V_s), b-values, radon emission, etc. preceding large and moderate earthquakes, especially in areas near epicentres and along faults and other weak zones. Intensity and duration (T) of these premonitory quantities are very much dependent on magnitude (M) of the seismic event. Hence, these quantities may be utilised for prediction of an incoming seismic event well in advance of the actual earthquake. In the recent past, tilts, strain in deep underground rock and crustal displacements have been observed in the Koyna earthquake region over a decade covering pre- and postearthquake periods; and these observations confirm their reliability for qualitative as well as quantitative premonitory indices. Tilt began to change significantly one to two years before the Koyna earthquake of December 10, 1967, of magnitude 7.0. Sudden changes in ground tilt measured in a watertube tiltmeter accompanied an earthquake of magnitude 5.2 on October 17, 1973 and in other smaller earthquakes in the Koyna region, though premonitory changes in tilt preceding smaller earthquakes were not so much in evidence. However, changes in strains in deep underground rock were observed in smaller earthquakes of magnitude 4.0 and above. Furthermore, as a very large number of earthquakes (M = 1–7.0) were recorded in the extensive seismic net in the Koyna earthquake region during 1963–1975, precise b-value variations as computed from the above data, could reveal indirectly the state of crustal (tectonic) strain variations in the earthquake focal region and consequently act as a powerful premonitory index, especially for the significant Koyna earthquakes of December 10, 1967 (M = 7.0) and October 17, 1973 (M = 5.2). The widespread geodetic and magnetic levelling observations covering the pre- and postearthquake periods indicate significant vertical and horizontal crustal displacements, possibly accompanied by large-scale migration of underground magma during the large seismic event of December 10, 1967 in the Koyna region (M = 7.0). Duration (T) of premonitory changes in tilt, strains, etc., is generally governed by the equation of the type logT = A + BM (A and B are statistically determined coefficients). Similar other instances of premonitory evidences are also observed in micro-earthquakes (M = − 1 to 2) due to activation of a fault caused by nearby reservoir water-level fluctuations.     

Water level fluctuations due to earthquakes in Koyna-Warna region,India

Earthquakes cause a variety of hydrological phenomena, including changes in the ground water levels in bore wells. The Koyna region in the peninsular shield of India, hitherto considered stable in terms of seismic activity, has been active since 1967. More recently, the earthquakes have been localized to the newly impounded Warna reservoir, which is located south of Koyna, where a burst of seismicity occurred in 1993. The region continues to remain seismically active even after four decades. Twenty-one bore wells were drilled around the seismic source volume in the region to observe water level changes resulting from earthquake phenomena. Our studies have shown coseismic anomalous water level changes to be associated with the moderate earthquakes of April 25, 1997 and February 11, 1998. Our results show that changes in the ground water level in bore wells are correlated with micro-earthquake activity, both preceding and following moderately sized earthquakes. The results have implications in enhancing our understanding of earthquake mechanisms.     

Earthquakes associated with the Clark Hill reservoir, South Carolina — A case of induced seismicity

Following the McCormick County, South Carolina, local magnitude 4.3 earthquake on August 2, 1974, continuous seismic activity has been observed in the area. The epicenters are located within 3 km from the Clark Hill reservoir. The shallow seismic activity appears to be related to water level fluctuations in the reservoir, and follows them by about two days. The frequency of earthquakes is related to the rate of change of water level, while the energy release is seen to depend on the water level itself. Pore pressure fluctuations at focal depths in a regime of high tectonic stresses is the suggested cause for the triggering of earthquakes.     

Koyna,India, an ideal site for near field earthquake observations

The Koyna earthquake of M 6.3 on December 10, 1967 is the largest artificial water reservoir triggered earthquake globally. It claimed ～ 200 human lives and devastated the Koyna township. Before the impoundment of the Shivaji Sagar Lake created by the Koyna Dam, there were no earthquakes reported from the region. Initially a few stations were operated in the region by the CentralWater and Power Research Station (CWPRS). The seismic station network grew with time and currently the National Geophysical Research Institute (NGRI), Hyderabad is operating 23 broadband seismographs and 6 bore hole seismic stations. Another reservoir, Warna, was created in 1985, which provided a further impetus to Reservoir Triggered Seismicity (RTS). Every year following the monsoon, water levels rise in the two reservoirs and there is an immediate increase in triggered earthquakes in the vicinity of Koyna-Warna reservoirs in the months of August–September. Peak RTS is observed in September and later during December.Another spurt in triggered earthquakes is observed during the draining of the reservoirs in the months of April- May. A comparative study of RTS earthquake sequences and the ones occurring in nearby regions made it possible to identify four common characteristics of RTS sequences that discriminate them from normal earthquake sequences. As the RTS events continue to occur at Koyna in a large number in a limited area of 20 km x 30 km, at shallow depths (mostly 2 to 9 km), the region being accessible for all possible observations and there being no other source of earthquakes within 100 km of Koyna Dam, it was suggested to be an ideal site for near field observations of earthquakes. This suggestion was discussed by the global community at an ICDP sponsored workshop held at Hyderabad and Koyna in 2011. There was an unanimous agreement about the suitability of the site for deep scientific drilling; however, a few additional observations/experiments were suggested. These were carried out in the following three years and another ICDP workshop was held in 2014, which totally supported setting up a borehole laboratory for near field investigations at Koyna. Location of a Pilot Bore-hole was decided on the basis of seismic activity and other logistics. The 3 km deep Pilot Borehole was spudded on December 20, 2016 and completed on June 11, 2017.     

Seismic activity in the McNaughton Lake area, Canada

Filling of McNaughton Lake, with a capacity at full load of25 · 10⁹m³ and maximum depth 191 m, was initiated on March 29, 1973. An earthquake swarm of 747 events (M_L > 0) with largest eventM_L = 4.7 occurred within 17 km of the reservoir just prior to loading. Subsequent to this, three swarms of 194, 292 and 22 events with maximumM_L = 4.1 occurred in the same region; however, no earthquakes have occurred between the reservoir and the swarm area. The level of regional seismic activity is similar to that observed prior to loading. The distribution of this activity, excluding swarm events, exhibits a spatial pattern similar to that recorded earlier by regional seismic stations, except that several events appear to be associated with the fault underlying the Rocky Mountain Trench in which the lake is formed. During a loading-unloading cycle in which the maximum water depth varied from 98 m to 171 m to 131 m, the change ofv_p was less than 2%. This indicates that no significant change in dilatancy or degree of water saturation occurred in the upper crustal layer during this cycle.     

South Australian seismicity 1967–1971

The South Australian seismograph network has been extended since 1968 so that most earthquakes of Richter magnitude ML ≥ 1.9 within the state are located accurately. Recurrence relations have been derived which define the seismicity of the known active areas. The seismic energy release has also been used to indicate the spatial variation in earthquake activity. Epicentre trends have been noted, and may be related to hypothesized movements on intracontinental plate boundaries.     

Investigations of earthquake probabilities in the Indian Ocean seismic belts

Gumbel's extreme-value theory is used to estimate the probability of occurrence and average return periods for earthquakes in the Indian Ocean seismic belts. The nature of seismic activity, and annual and 50 year maximum magnitudes of earthquakes are also discussed. The earthquake occurrence model of autocorrelation lends support for the periodicity of the most probable earthquake in these belts. The percentage probability of recurrence of earthquakes of magnitude 8 and above has been estimated for the region mentioned.     

2010年新西兰M7.1地震前的长波辐射变化特征

以2001—2011年美国NOAA长波辐射数据为背景数据,利用涡度背景场法研究2010年9月3日新西兰南岛M7.1级地震前后卫星长波辐射数据变化特征。结果表明,地震当月在震中西南侧出现显著的长波辐射异常变化,这种变化在全年各月及2001—2011年11年历年同月变化中都是最为显著的,认为其是本次地震的1次映震表现。用同样方法对区内2001—2011年11年间发生6次7.0级以上地震的长波辐射数据进行分析,结果在3次陆地地震发震前都检测到了长波辐射异常变化,而海域地震前则未发现这种现象。检索前人的相关研究结果,发现仅有2次海域地震(2004年12月26日印度尼西亚苏门答腊西北海域8.7级地震和2010年1月12日海地7.0级地震)前有长波辐射异常变化的报道,而根据全球云量分布资料显示,这2个地震所发生的区域是全球洋面云量分布最少的2个区域,而新西兰地震发生的区域位于全球洋面云量分布最多的区域。因此,认为由于水汽和云层对地表红外辐射的强吸收作用,长波辐射捕捉陆地地震红外辐射异常变化更加灵敏,对陆地地震的映震效能要强于海洋地震。     

三峡库区仙女山和九畹溪断裂带水库地震变化规律

三峡库区蓄水以来,长江周边出现了大量的水库地震,主要集中于巴东—泄滩—仙女山区域,目前已达上万次,最大震级为5.1级。通过野外地质调查及对已有水库地震数据进行研究,采用构造地质分析方法,对三峡库区仙女山和九畹溪断裂带水库地震空间上的迁移规律、时间上的周期规律以及微地震群的成因机理进行分析。结果表明：从时间上看,水库地震具有周期性,表现为长周期（与库水位相关）和短周期（与库水位快速波动相关）;从空间上看,水库地震具有迁移规律,受九畹溪断裂控制的触发型水库地震存在着逐渐向南迁移的特征,受仙女山断裂控制的触发型水库地震局限分布于仙女山断裂北延端点处,分布于仙女山断裂西侧（周坪乡附近）带状分布的水库诱发地震逐渐呈点状向南迁移。     

The Anderson Reservoir seismic gap — Induced aseismicity?

A persistent 10-km seismicity gap along the Calaveras fault appears to be related to the presence of the Leroy Anderson Reservoir in the Calaveras-Silver Creek fault zones southeast of San Jose, California. A magnitude-4.7 earthquake occurred at a depth of 5 km in the centre of the gap on October 3, 1973. The sequence of immediate aftershocks usually accompanying shallow earthquakes of this magnitude in central California did not occur. A bridge crossing the reservoir near its southeast end has been severely deformed, apparently the result of tectonic creep on the Calaveras fault. The occurrence of creep and absence of small earthquakes along the Calaveras in the vicinity of the reservoir suggest a transition from stick slip to stable sliding, possibly brought about by increased pore pressure.     

Seismic hazard monitoring in Kamchatka and its applications to the M = 6.6 earthquake of 6 October 1987

The earthquake of 6 October 1987 (M = 6.6), which occurred near the Shipunsky Cape, Kamchatka, was the largest crustal event in the vicinity of the main city of Kamchatka — Petropavlovsk-Kamchatsky — during the last three decades. It was followed by numerous aftershocks. This earthquake allowed us to test the effectiveness of the seismic hazard monitoring in Kamchatka, including the seismological, geodetic and hydrogeochemical surveys. The seismic survey provided the location and source nature of the main shock and aftershocks and the seismic environment of the main shock. The geodetic and hydrogeochemical surveys have yielded data on the response to earthquakes of the Earth's surface deformations, water level, and chemical elements concentration in the underground water. As a result, the following data were obtained:

u

The earthquake of 6 October had a seismic moment 4–10 E18 Nm, thrust type of faulting and the source volume of 20 × 20 × 10 km³. The maximum intensity was VI–VII (MSK-64 scale) and maximum acceleration 88 cm/s².

Before this event, a relative increase in the number of the upper mantle (depth more than 100 km) moderate magnitude earthquakes during 5 years and a one-year period of seismic quiescence for small shallow earthquakes, were recognized. Significant anomalies in HCO₃ and H₃BO₃ concentrations in the underground waters were observed in the wells a week before the main shock.

     

Radiant flux as a guide to relative seismic efficiency

The parameters “radiant flux” (energy radiated per unit area of an earthquake fault in unit time) and “radiant flux per unit displacement” reflect the power dissipated on a fault during slip. Values for moderate-to-large earthquakes range over two orders of magnitude, implying considerable variations in seismic efficiency, even for events of similar magnitude occurring on faults of the same type.     

Lineament fabric from airborne LiDAR and its influence on triggered earthquakes in the Koyna-Warna region,western India

We present the results of the first airborne LiDAR survey flown in the Koyna-Warna region and examine the relationship between the lineament fabric and the ongoing seismicity in the region. Our studies reveal that earthquakes of M≥4.0 for the period 1968 to 2016 are strongly correlated with a 10 km wide N-S fracture zone, which not only represents the surface expression of seismically active basement faults, but also act as conduits for water percolation between the Koyna and Warna reservoirs. A decreasing trend in the annual distribution of earthquakes was observed from 1985. A new burst of seismic activity in 1993 followed the impoudment of the Warna reservoir. We report a change in annual seismicity pattern, where seismicity peaks during September and December in the pre-Warna period, with a new peak emerging during March-April subsequent to the impoundment of Warna reservoir. A model is proposed to explain the seismicity along dominant N-S lineaments and the impact of Warna reservoir impounding which altered the hydrogeologic regime in the region.     

Earthquake magnitude — recent research and current trends

The magnitude recommendations adopted by the IASPEI Assembly at Zürich in 1967 have had both a stabilizing and a stimulating effect on magnitude determinations and related research. From 1967 onwards, one magnitude research paper has appeared on the average almost every week, thus making this parameter the most studied one in seismology. New facilities and more accurate methods, e.g., concerning instrumental equipment and interpretation techniques, have made it possible to improve earlier achievements. The application of magnitude scales has been extended in all respects, e.g., with regard to epicentral distances, focal depths, wave types and wave periods. Magnitude—frequency relations have become the most investigated equations within seismology, observationally as well as theoretically. They have wide applications, e.g., for estimating the maximum magnitudes of future earthquakes — an important item in earthquake prediction. The magnitudes provide significant information on other source parameters, such as wave energy, fault length, seismic moment. Relations between different types of magnitude yield valuable information on source properties. For instance, relations between magnitudes based on body and surface waves are used for efficient discrimination between earthquakes and underground explosions. It is our purpose to review the magnitude development in the post-Zürich period (1967–1980), partly for geologists, tectonophysicists and engineers, who need an overview, partly for seismologists, who need an introduction to an overwhelmingly comprehensive literature.     

Space–time distribution of interplate moment release including slow earthquakes and the seismo-geodetic coupling in the Sanriku-oki region along the Japan trench

Along the Japan trench where some Mw8 class interplate earthquakes occurred in the past century such as the 1896 Sanriku tsunami earthquake (M6.8, Mt8.6, 12×10²⁰ N m) and the 1968 Tokachi-oki earthquake (Mw8.2, 28×10²⁰ N m), the Pacific plate is subducting under northeast Japan at a rate of around 8 cm/year. The seismic coupling coefficient in this region has been estimated to be 20–40%. In the past decade, three ultra-slow earthquakes have occurred in the Sanriku-oki region (39°N–42°N): the 1989 Sanriku-oki (Mw7.4), the 1992 Sanriku-oki (Mw6.9), and the 1994 Sanriku-oki (Mw7.7) earthquakes. Integrating their interplate moments released both seismically and aseismically, we have the following conclusions. (1) The sum of the seismic moments of the three ultra-slow earthquakes was (4.8–6.6)×10²⁰ N m, which was 20–35% of the accumulated moment (18.6–23.0)×10²⁰ N m, in the region (39°N–40.6°N, 142°E–144°E) for the 21–26 years since the 1968 Mw8.2 Tokachi-oki earthquake. This is consistent with the previous estimates of the seismic coupling coefficient of 20–40%. On the other hand, the sum of the interplate moments including aseismic faulting is (11–16)×10²⁰ N m, leading to a “seismo-geodetic coupling coefficient” of 50–85%, which is an extension of the seismic coupling coefficient to include slow events. (2) The time constants showed a large range from 1 min (10² s) for the 1968 Tokachi-oki earthquake to 10–20 min (10³ s) for the 1896 Sanriku tsunami earthquake, to one day (10⁵ s) for the 1992 Sanriku-oki ultra-slow earthquake, to on the order of one year (10⁷ s) for the 1994 Sanriku-oki ultra-slow earthquakes. (3) Based on the space–time distribution, three “gaps of moment release,” (40.6°N–42°N, 142°E–144°E) 39°N–40°N, 142°E–143°E) and (39°N–40°N, 142°E–144°E), are identified, instead of the gaps of seismicity.     

Seismic hazard map of Coimbatore using subsurface fault rupture

This study presents the future seismic hazard map of Coimbatore city, India, by considering rupture phenomenon. Seismotectonic map for Coimbatore has been generated using past earthquakes and seismic sources within 300 km radius around the city. The region experienced a largest earthquake of moment magnitude 6.3 in 1900. Available earthquakes are divided into two categories: one includes events having moment magnitude of 5.0 and above, i.e., damaging earthquakes in the region and the other includes the remaining, i.e., minor earthquakes. Subsurface rupture character of the region has been established by considering the damaging earthquakes and total length of seismic source. Magnitudes of each source are estimated by assuming the subsurface rupture length in terms of percentage of total length of sources and matched with reported earthquake. Estimated magnitudes match well with the reported earthquakes for a RLD of 5.2% of the total length of source. Zone of influence circles is also marked in the seismotectonic map by considering subsurface rupture length of fault associated with these earthquakes. As earthquakes relive strain energy that builds up on faults, it is assumed that all the earthquakes close to damaging earthquake have released the entire strain energy and it would take some time for the rebuilding of strain energy to cause a similar earthquake in the same location/fault. Area free from influence circles has potential for future earthquake, if there is seismogenic source and minor earthquake in the last 20 years. Based on this rupture phenomenon, eight probable locations have been identified and these locations might have the potential for the future earthquakes. Characteristic earthquake moment magnitude (M _w ) of 6.4 is estimated for the seismic study area considering seismic sources close to probable zones and 15% increased regional rupture character. The city is divided into several grid points at spacing of 0.01° and the peak ground acceleration (PGA) due to each probable earthquake is calculated at every grid point in city by using the regional attenuation model. The maximum of all these eight PGAs is taken for each grid point and the final PGA map is arrived. This map is compared to the PGA map developed based on the conventional deterministic seismic hazard analysis (DSHA) approach. The probable future rupture earthquakes gave less PGA than that of DSHA approach. The occurrence of any earthquake may be expected in near future in these eight zones, as these eight places have been experiencing minor earthquakes and are located in well-defined seismogenic sources.     

Teleseismic P-wave residual investigations at Shillong,India

The northeast India region is seismically very active and it has experienced two large earthquakes of magnitude 8.7 during the last eight decades (1897 and 1950). We have analysed teleseismic P-wave residuals at Shillong, the only reliable seismic station operating in the region, to investigate a possible association of travel-time residual anomaly with earthquake occurrence. The period covered is from October 1964 through March 1976. The total number of events is 9479, including 1767 events with depth >/ 100 km. Six-monthly average residuals have been calculated. The standard deviations are less than 0.10 sec for these data sets. During the period of investigations, no major earthquake took place close to Shillong. The earthquake of June 1, 1969 with a magnitude (M_b) of 5.0, at an epicentral distance of 20 km from Shillong is the only significant event. This earthquake is found to be associated with a travel-time increase with a maximum amplitude of 0.4 sec. It appears that, in general, the P-wave velocity has decreased in the neighbourhood of Shillong since 1969. A quadrant-wise analysis of residuals indicates that the residual anomaly is most prominent in the SE quadrant from Shillong.     

Spatial-temporal variability of seismic hazard in peninsular India

This paper examines the variability of seismic activity observed in the case of different geological zones of peninsular India (10°N–26°N; 68°E–90°E) based on earthquake catalog between the period 1842 and 2002 and estimates earthquake hazard for the region. With compilation of earthquake catalog in terms of moment magnitude and establishing broad completeness criteria, we derive the seismicity parameters for each geologic zone of peninsular India using maximum likelihood procedure. The estimated parameters provide the basis for understanding the historical seismicity associated with different geological zones of peninsular India and also provide important inputs for future seismic hazard estimation studies in the region. Based on present investigation, it is clear that earthquake recurrence activity in various geologic zones of peninsular India is distinct and varies considerably between its cratonic and rifting zones. The study identifies the likely hazards due to the possibility of moderate to large earthquakes in peninsular India and also presents the influence of spatial rate variation in the seismic activity of this region. This paper presents the influence of source zone characterization and recurrence rate variation pattern on the maximum earthquake magnitude estimation. The results presented in the paper provide a useful basis for probabilistic seismic hazard studies and microzonation studies in peninsular India.