Earthquake swarms in continental rifts — A comparison of selected cases in America, Africa and Europe

The occurrence of earthquake swarms is typically related to magmatic activity in volcanoes, yet swarms are also common in other intracontinental regions such as continental rifts. We present here a summary of geophysical observations that have been made in earthquake swarm areas of the Rio Grande, Kenya, and Eger rifts, focusing on characteristic parameters for the origin and generation of the swarm earthquakes.Our compilation of seismological parameters such as spatial distribution and focal parameters of hypocenters, magnitude statistics, and the location of the swarm centres in the rift environments reveals major similarities. The earthquake swarms take place at shallow depth between 0 and 10 km. The maximum magnitudes are mostly less than 4.5. The b-values, indicating the magnitude frequency relation of the seismicity, are about 0.8. They are hence not deviating from a normal non-volcanic intraplate environment, but are considerably lower than those of volcanic earthquake swarms. Focal mechanism studies give uniform pictures of stress field orientation and faulting style for the swarm areas. In all three rifts, the centres of swarm activity seem to be restricted to rift valley sections that may be influenced by large-scale fracture or shear zones that intersect the rifts. We conclude that these deep-reaching zones of weakness allow intrusions of upper mantle material into crustal layers, where magma-related fluids or fluctuations of the magma bodies themselves cause the generation of earthquake swarms.     

Weak tidal correlation of NW-Bohemia/Vogtland earthquake swarms

We analyze the possible effect of solid Earth tidal stresses upon a vertical strike-slip fault in NW-Bohemia/Vogtland, central Europe, typical by occurrence of swarm earthquakes. The horizontal components of solid Earth tidal stresses were found strongly to prevail and to reach the level of 2 kPa. We examined tidal triggering as influence of tidal stresses to launching the swarm activity in relative absence of other stress disturbances. The onset times of 46 swarms of mostly M_L < 3 earthquakes that occurred in the period 1991–2005 displayed an increased occurrence near the fortnightly maximum of tidal extensive normal stress. The statistical test however did not prove a statistically significant correlation indicating a triggering effect of fault extension due to tidal loading. We also examined tidal effects to the already running seismic activity of the prominent 2000 swarm by comparing the tidal stress distribution in the investigated period with the distribution of tidal stresses in the occurrence times of each earthquake. The results show that these distributions are almost similar, which indicates that individual earthquakes occur independent of tidal stresses. The unclear tidal correlation of the swarm seismicity may be interpreted by small amplitudes and rates of tidal stress changes compared to the amplitudes and rates of coseismic stress perturbations and of pressure bursts of deep generated fluids.     

Seismically induced variations in Mariánské Lázně fault gas composition in the NW Bohemian swarm quake region, Czech Republic — A continuous gas monitoring

A continuously operated gas monitoring station was emplaced within the epicentral area of the NW Bohemian swarm earthquakes overlying directly the active Mariánské Lázně fault. The recordings of 8-month continuous monitoring period are presented. The variations in radon concentrations are similarly to variations in CO₂, i.e. CO₂ is considered to be the carrier gas for radon. Very small diurnal variations in gas concentration are caused by the earth tides, as daily variations in meteorological conditions cannot explain a short daily minimum at midday times. Sudden changes in gas concentration, which clearly exceed these diurnal variations occur and are always linked with seismic activities. Decreased gas concentration may indicate compression resulting in reduced fault permeability as is implied by negative peaks following local earthquake swarms. A sudden increase in CO₂ and Rn concentration may indicate an increased fault permeability caused by stress redistribution, giving rise to opening of migration pathways. This implies a repeatedly sudden rise in gas concentration before local earthquake swarms. Several variations in gas concentration were monitored linked with remote earthquakes of ground motion amplitudes  >1 μm. These seismic events are accompanied by an interference of the diurnal gas concentration–stress-cycle along the Mariánské Lázně fault. However, if shocks of remote earthquake can alter properties of the migrating fluids or the fault properties it can be suggested that these are able to trigger local seismicity, as indicated in the case of the Slovenia earthquake on 12th July 2004.     

Tidal triggering of earthquakes in the subducting Philippine Sea plate beneath the locked zone of the plate interface in the Tokai region, Japan

We found a characteristic space–time pattern of the tidal triggering effect on earthquake occurrence in the subducting Philippine Sea plate beneath the locked zone of the plate interface in the Tokai region, central Japan, where a large interplate earthquake may be impending. We measured the correlation between the Earth tide and earthquake occurrence using microearthquakes that took place in the Philippine Sea plate for about two decades. For each event, we assigned the tidal phase angle at the origin time by theoretically calculating the tidal shear stress on the fault plane. Based on the distribution of the tidal phase angles, we statistically tested whether they concentrate near some particular angle or not by using Schuster's test. In this test, the result is evaluated by p-value, which represents the significance level to reject the null hypothesis that earthquakes occur randomly irrespective of the tidal phase angle. As a result of analysis, no correlation was found for the data set including all the earthquakes. However, we found a systematic pattern in the temporal variation of the tidal effect; the p-value significantly decreased preceding the occurrence of M ≥ 4.5 earthquakes, and it recovered a high level afterwards. We note that those M ≥ 4.5 earthquakes were considerably larger than the normal background seismicity in the study area. The frequency distribution of tidal phase angles in the pre-event period exhibited a peak at the phase angle where the tidal shear stress is at its maximum to accelerate the fault slip. This indicates that the observed small p-value is a physical consequence of the tidal effect. We also found a distinctive feature in the spatial distribution of p-values. The small p-values appeared just beneath the strongly coupled portion of the plate interface, as inferred from the seismicity rate change in the past few years.     

Triggered seismicity in the Andean arc region via static stress variation by the MW = 8.8, February 27, 2010, Maule earthquake

We localized crustal earthquakes in the Andean arc, between 35°S and 36°S, from December 2009 to May 2010. This research shows a seismicity increase, in a narrow longitudinal area, of more than nine times after the great M_w 8.8 Maule earthquake.The localized seismicity defines an area of ∼80 km long and ∼18 km wide and NNW to NNE trend. The M_d magnitudes varied from 0.7 to 3.1 except for two earthquakes with M_w of 3.9 and 4.5, located in the northern end of the area. The focal mechanisms for these two last events were normal/strike-slip and strike-slip respectively.During 2011, a network of 13 temporary stations was installed in the trasarc region in Malargüe, Argentina. Sixty earthquakes were localized in the study region during an 8 month period.We explored how changes in Coulomb conditions associated with the mega-thrust earthquake triggered subsequent upper-plate events in the arc region. We assumed the major proposed structures as receiver faults and used previously published earthquake source parameters and slip distribution for the Maule quake. The largest contribution to static stress change, up to 5 bars, derives from unclamping resulting consistent with co-seismic dilatational deformation inferred from GPS observations in the region and subsidence in nearby volcanoes caused by magma migration.Three different Quaternary tectonic settings–extensional, strike-slip and compressional-have been proposed for the arc region at these latitudes. We found that the unclamping produced by the Maule quake could temporarily change the local regime to normal/strike-slip, or at least it would favor the activation of Quaternary NNE to N-trending dextral strike-slip faults with dextral transtensional movement.     

Seismicity of the Baikal rift system from regional network observations

In the paper we report the state-of-the-art of seismicity study in the Baikal rift system and the general results obtained. At present, the regional earthquake catalog for fifty years of the permanent instrumental observations consists of over 185,000 events. The spatial distribution of the epicenters, which either gather along well-delineated belts or in discrete swarms is considered in detail for different areas of the rift system. At the same time, the hypocenters are poorly constrained making it difficult to identify the fault geometry. Clustered events like aftershock sequences or earthquake swarms are typical patterns in the region; moreover, aftershocks of M ⩾ 4.7 earthquakes make up a quarter of the whole catalog. The maximum magnitude of earthquakes recorded instrumentally is M_LH7.6 for a strike-slip event in the NE part of the Baikal rift system and M_LH6.8 for a normal fault earthquake in the central part of the rift system (Lake Baikal basin). Predominant movement type is normal faulting on NE striking faults with a left lateral strike-slip component on W–E planes. In conclusion, some shortcomings of the seismic network and data processing are pointed out.     

Statistical occurrence analysis and spatio-temporal distribution of earthquakes in the Apennines (Italy)

We present two examples of statistical analysis of seismicity conducted by integrating geological, geophysical and seismological data with the aim to characterize the active stress field and to define the spatio-temporal distribution of large earthquakes. Moreover, our data will help to improve the knowledge of the “seismogenic behavior” of the areas and to provide useful information for seismic hazard evaluation.The earthquakes are described by two non-parametric statistical procedures integrating also tectonic-physical parameters to study the spatio-temporal variability.The results show that the areas are characterized by: 1) a stress regime with mainly extensional kinematics; 2) tectonic structures mainly oriented with the active stress field (S_hmin = N44° ± 18° in the southern Apennines and S_hmin = N50° ± 17° in the central Apennines); 3) cluster distribution of seismicity and 4) a high probability of earthquake occurrence (M > 5.5) in the next 10 years.     

Power law distributions of seismic rates

We report an empirical determination of the probability density functions P_data(r) of the number r of earthquakes in finite space–time windows for the California catalog for different space (5 × 5 to 50 × 50 km²) and time intervals (0.1 to 1000 days). The data can be represented by asymptotic power law tails together with several cross-overs reasonably explained by one of the most used reference model in seismology (ETAS), which assumes that each earthquake can trigger other earthquakes according to complex cascades. These results are useful to constrain the physics of earthquakes and to estimate the performance of forecasting models of seismicity.     

On the swelling potential of compacted high plasticity clays

We conducted a study of the spatial distributions of seismicity and earthquake hazard parameters for Turkey and the adjacent areas, applying the maximum likelihood method. The procedure allows for the use of either historical or instrumental data, or even a combination of the two. By using this method, we can estimate the earthquake hazard parameters, which include the maximum regional magnitude Mˆ_max, the activity rate of seismic events and the well-known bˆ value, which is the slope of the frequency-magnitude Gutenberg-Richter relationship. These three parameters are determined simultaneously using an iterative scheme. The uncertainty in the determination of the magnitudes was also taken into consideration. The return periods (RP) of earthquakes with a magnitude M ≥ m are also evaluated. The whole examined area is divided into 24 seismic regions based on their seismotectonic regime. The homogeneity of the magnitudes is an essential factor in such studies. In order to achieve homogeneity of the magnitudes, formulas that convert any magnitude to an M_S-surface scale are developed. New completeness cutoffs and their corresponding time intervals are also assessed for each of the 24 seismic regions. Each of the obtained parameters is distributed into its respective seismic region, allowing for an analysis of the localized seismicity parameters and a representation of their regional variation on a map. The earthquake hazard level is also calculated as a function of the form Θ = (Mˆ_max,RP_6.0), and a relative hazard scale (defined as the index K) is defined for each seismic region. The investigated regions are then classified into five groups using these parameters. This classification is useful for theoretical and practical reasons and provides a picture of quantitative seismicity. An attempt is then made to relate these values to the local tectonics.     

Evaluation of deep crustal earthquakes in northern Germany – Possible tectonic causes

We present new evidence for seven deep crustal, intraplate earthquakes in northern Germany, a region regarded as an area of low seismicity. From 2000 to 2018, seven earthquakes with magnitudes of M_L 1.3–3.1, were detected at depths of 17.0–31.4 km. By placing the earthquake hypocentres in a geological three‐dimensional model, we can correlate two of the earthquakes with the Thor Suture, a major fault zone in this area. Five of the earthquakes group in the lower crust near the Moho, which implies that parts of the lower crust and the crust/mantle boundary in northern Germany act as a structural discontinuity on which deformation localizes. Numerical simulation implies that stress changes due to glacial isostatic adjustment most likely triggered these deep crustal earthquakes.     

On the possibilities of premonitory swarms for three sequences of earthquakes of the Burma—Szechwan region

Short term spatial and temporal variations in seismicity prior to the three sequences of earthquakes of m_b 5.8 of the Burma—Szechwan region are studied. Six years (1971–1976) of ISC seismicity data, as reported in the Regional Catalogue of Earthquakes, are considered. During the period, six earthquakes of body wave magnitude m_b 5.8 occurred in four sequences. Of these, three sequences are preceded by swarm activity in the epicentral regions. Evison (1977b) suggested that the swarm before the sequences of large shocks is a possible long-term precursor. He derived the conclusion by analyzing earthquakes in New Zealand and California. The analysis of the seismicity data for the region under investigation supports Evison's view and suggests that a relation between swarms and sequences of large events exists. The precursory time period (i.e. the time from beginning of the swarm to the main shock) for the Szechwan earthquakes of m_b = 5.9 (Feb. 6, 1973) and m_b = 5.8 (May 10, 1974) and the Burma earthquake of m_b = 6.2 (Aug. 12, 1976) are 305, 317 and 440 days, respectively.     

First-order rupture features of the 2011 MW 9.0 Tohoku (Japan) earthquake from surface waves

This study analyzed the rupture directivity of the 2011 Tohoku earthquake by using 100-s Rayleigh-wave travel-times, influenced by the finite source, to derive the fault parameters of the earthquake. The results demonstrated that the earthquake exhibited a slow rupture propagation with a rupture velocity of approximately 1.5–2.0 km/s and asymmetric bilateral faulting. The two rupture directions were N60°E and N127°E, with rupture lengths of approximately 276 km and 231 km, respectively. The rupture toward N60°E had a source duration of approximately 183 s, longer than that toward N127°E (approximately 156 s). Overall, the entire source duration of the earthquake faulting lasted approximately 183 s. Regarding historical seismicity in eastern Japan, the 2011 Tohoku earthquake not only ruptured a locked area in which large earthquakes have rarely occurred, but also ruptured the source regions of several historical earthquakes. With the exception of its slow rupture velocity and generation of a tsunami, the rupture features of the 2011 Tohoku earthquake were inconsistent with those of typical tsunami earthquakes.     

Usage of an early warning and information system Web-site for real-time seismicity in Iceland

Iceland has been subjected to destructive earthquakes and volcanic eruptions throughout history. Such events are often preceded by changes in earthquake activity over varying timescales. Although most seismicity is confined to micro-earthquakes, large earthquakes have occurred within populated regions. Following the most recent hazardous earthquakes in 2000, the Icelandic Meteorological Office (IMO) developed an early warning and information system (EWIS) Web-site for viewing near-real-time seismicity in Iceland. Here we assess Web-site usage data in relation to earthquake activity, as recorded by the South Iceland Lowland (SIL) seismic network. Between March 2005 and May 2006 the SIL seismic network recorded 12,583 earthquakes. During this period, the EWIS Web-site logged a daily median of 91 visits. The largest onshore event (M _L 4.2) struck 20 km from Reykjavík on 06 March 2006 and was followed by an immediate, upsurge in usage resulting in a total of 1,173 unique visits to the Web-site. The greatest cluster of large (≥M _L 3) events occurred 300 km offshore from Reykjavík in May 2005. Within this swarm, 9 earthquakes ≥M _L 3 were detected on 11 May 2005, resulting in the release of a media bulletin by IMO. During the swarm, and following the media bulletin, the EWIS Web-site logged 1,234 unique visits gradually throughout the day. In summary, the data reveal a spatial and temporal relationship between Web-site usage and earthquake activity. The EWIS Web-site is accessed immediately after the occurrence of a local earthquake, whereas distant, unfelt earthquakes generate gradual interest prompted by media bulletins and, possibly, other contributing factors. We conclude that the Internet is a useful tool for displaying seismic information in near-real-time, which has the capacity to help increase public awareness of natural hazards.     

Seismic swarms,fluid flow and hydraulic conductivity in the forearc offshore North Costa Rica and Nicaragua

At the continental margin of north Costa Rica and Nicaragua, the strongly hydrated Cocos Plate subducts beneath the Caribbean Plate. From the downgoing Cocos plate fluids are released through extensional fractures in the overriding plate. At the seafloor, they form fluid seeps, mounds and other types of fluid expulsion. Using an offshore temporary seismic network, we investigated seismicity possibly related to these processes and observed several swarms of earthquakes located on the continental slope trenchward of the seismogenic zone of S Nicaragua. The seismicity occurred within the downgoing plate, near the plate interface and in the overriding plate. We interpret these swarm events as an expression of pore pressure propagation under critical stress conditions driven by fluid release from the downgoing plate. In order to estimate hydraulic diffusivity and permeability values, we applied a theory developed for injection test interpretation to the spatio-temporal development of the swarms. The resulting diffusivity and permeability values are in the ranges of 28–305 m²/s and 3.2 × 10^?14 m²–35.1 × 10^?14 m², respectively, applying to the continental and oceanic crust near the plate interface. These values are somewhat larger than observed in drill logs on the margin wedge off north Costa Rica, but of comparable magnitude to values estimated for the Antofagasta 1995 earthquake aftershock sequence.     

Monitoring of earthquake precursors by multi-parameter stations in Eskisehir region (Turkey)

The objective of this study was to investigate the geochemical and hydrogeological effects of earthquakes on fluids in aquifers, particularly in a seismically active area such as Eskisehir (Turkey) where the Thrace–Eskisehir Fault Zone stretches over the region. The study area is also close to the North Anatolian Fault Zone generating devastating earthquakes such as the ones experienced in 1999, reactivating the Thrace–Eskisehir Fault. In the studied area, Rn and CO₂ gas concentrations, redox potential, electrical conductivity, pH, water level, water temperature, and the climatic parameters were continuously measured in five stations for about a year. Based on the gathered data from the stations, some ambiguous anomalies in geochemical parameters and Rn concentration of groundwater were observed as precursors several days prior to an earthquake. According to the mid-term observations of this study, well-water level changes were found to be a good indicator for seismic estimations in the area, as it comprises naturally filtered anomalies reflecting only the changes due to earthquakes. Also, the results obtained from this study suggest that both the changes in well-water level and gas–water chemistry need to be interpretated together for more accurate estimations. Valid for the studied area, it can be said that shallow earthquakes with epicentral distances of <30 km from the observation stations have more influence on hydrochemical parameters of groundwater and well-water level changes. Although some hydrochemical anomalies were observed in the area, it requires further observations in order to be able to identify them as precursors.     

Synthesis of seismicity studies for western Alaska

An examination of the Alaskan earthquake catalogs revealed that from 1928 through 1965, twelve earthquakes of magnitudes (M) in the range 5.6 to 7.3 were located in and around Seward Peninsula region of western Alaska. Moreover, a number of earthquakes of M < 5.0 were found to distribute over the same area. Further investigation of the seismicity employing a local seismographic network in the above area showed a higher level of onshore and offshore seismic activity than had been previously recognized. A number of clusters of earthquakes have been identified. Some of them are located in the epicentral areas of past strong earthquakes (M > 5.5) and some are located in areas traversed by mapped faults. The nature of the seismicity as identified with the local network data is primarily crustal over the entire study area. Investigation of focal mechanisms of isolated strong earthquakes or clusters of small earthquakes show normal faulting as the dominant mode of strain energy release in the western part of Alaska. Moreover, in areas lying, approximately, south and north of Kotzebue Sound, the principal components of horizontal stresses tend to orient in the NW-SE and WNW-ESE directions, respectively.     

Background seismicity in the Central Apennines of Italy: The Abruzzo region case study

We investigate background seismic activity of the Abruzzo region, a 5000 km² area located within the Central Apennines of Italy, where in the past 600 years at least 5 large earthquakes (I = XI–X) have occurred.Between April 2003 and September 2004, a dense temporary seismic network composed of 30 digital three-component seismic stations recorded 850 earthquakes with 0.9 < M_L < 3.7. We present earthquake locations and focal mechanisms obtained by standard procedures and an optimized velocity model computed with a search technique based on genetic algorithms.The seismicity occurs at a low and constant rate of  2.6 e− 04 events/daykm² and is sparsely distributed within the first 15 km of the crust. Minor increases in the seismicity rate are related to the occurrence of small and localised seismic sequences that occur at the tip of major active normal faults along secondary structures.We observe that during the 16 months of study period, the Fucino fault system responsible for the 1915 Fucino earthquake (M_S = 7.0), and the major normal faults of the area, did not produce significant seismic activity.Fault plane solutions evaluated using P-wave polarity data show the predominance of normal faulting mechanisms ( 55%) with NE-trending direction of extension coherent with the regional stress field active in this sector of the Apennines. Around 27% of the focal solutions have pure strike–slip mechanisms and the rest shows transtensional faulting mechanisms that mainly characterise the kinematics of the secondary structures activated by the small sequences.We hypothesize that the largest known NW-trending normal faults are presently locked and we propose that in the case of activation, the secondary structures located at their tips may act as transfer faults accommodating a minor part of the extensional deformation with strike–slip motion.     

Delineation and interpretation of spatial coseismic response of groundwater levels in shallow and deep parts of an alluvial plain to different earthquakes: A case study of the Kumamoto City area,southwest Japan

Coseismic changes in groundwater levels have been investigated throughout the world, but most studies have focused on the effects of one large earthquake. The aim of this study was to elucidate the spatial patterns of level changes in response to several earthquakes, and the relationship of the patterns to shallow and deep groundwater in the same area. We selected the Kumamoto City area in southwest Japan, a region with one of the richest groundwater resources in Japan, as our study site. Data from hourly measurements of groundwater levels in 54 wells were used to characterize the coseismic responses to four earthquakes that occurred in 2000, 2001, 2005, and 2008. Although the distance to the hypocenter (12–2573 km), and seismic energy (M_w = 5.0–8.0) of these earthquakes varied, systematic groundwater level changes were observed in the range of 0.01–0.67 m. Spatial patters of the level changes were clarified by interpolating the point data by a spline method. The zones where coseismic rises were observed were generally wider for deep groundwater than for shallow groundwater, probably as a result of an increase in compressive stress. General trends in the changes in groundwater levels, and calculated pressure changes, were clarified to be consistent in the deep groundwater, but the coseismic increases or decreases in compressive stress in the shallow groundwater were variable, depending on the distance to the earthquake epicenter. We developed a conceptual model of the mechanism underlying this phenomenon by assuming permeability enhancement induced by elastic strain and pore-pressure change over the depth range. In addition, the importance of local geology was identified, because levels in the area of Togawa lava (a porous andesite) tended to change more in magnitude, and more quickly, with a shorter recovery time, than levels measured in the area outside the lava.