In geophysical inverse problems, the distribution of physical properties in an Earth model is inferred from a set of measured data. A necessary step is to select data that are best suited to the problem at hand. This step is performed ahead of solving the inverse problem, generally on the basis of expert knowledge. However, expert-opinion can introduce bias based on pre-conceptions. Here we apply a trans-dimensional algorithm to automatically weigh data on the basis of how consistent they are with the fundamental assumptions made to solve the inverse problem. We demonstrate this approach by inverting arrival times for the location of a seismic source in an elastic half space, under the assumptions of a point source and constant velocities. The key advantage is that the data do no longer need to be selected by an expert, but they are assigned varying weights during the inversion procedure.
The rockfall risk due to mining-induced seismicity reduces by installing appropriate rock support to absorb the kinetic energy from a seismic event, which is calculated by assuming the mass of ejected rock and its ejection velocity. Estimation of ejection velocity is normally based on scaling laws that do not consider the effect of the excavation free-surface and existing fractures near the excavation free-surface. Field monitoring studies have shown that the peak particle velocity on the free-surface can be much larger than the velocity in deep solid rock. The interaction between the fractures and the free-surface under incident S-wave is investigated by using a two-dimensional UDEC model with fractured zone characterized as one, two, three and four sets of parallel fractures with varied intersecting angles. The results show that wave amplification factor varies according to the incident wave angle, the number of fracture sets and fracture spacing.
Using original seismograph records and bulletin data were-determined the origin time, location, seismic moment (M0) and magnitudes (MS and Mw) for the four earthquakes in the beginning of the 20th century. These are two strong earthquakes April 4, 1904 near Krupnik, Bulgaria (Mw = 6.8, MS = 7.2 respectively), the April 23 1909 earthquake near Benavente, Portugal (MS = 6.3), and the June 14, 1913 earthquake near Gorna Orjahovitza, Bulgaria (MS = 6.3). Twenty-nine traces from original records have been analysed, a large number of original station bulletins have been consulted and a consistent methodology for analysing these early 20th century instrumental information is presented. In spite of a thorough effort in re-assembling and quality control of the original data, large inaccuracies remain in the improved instrumental epicentre locations and origin times. The seismic moment estimates we obtained (2.3 1018 ≤ M0 ≤ 3.9 1019 Nm) are the first ever determined for these events. The magnitude estimates (6.3 ≤ MS ≤ 7.2 and 6.2 ≤ Mw ≤ 7.0) are robust and systematically lower than most of previous estimates for all earthquakes (Gutenberg and Richter, 1954; Christoskov and Grigorova, 1968; Karnik, 1969). For the largest Krupnik event our estimates agree with those of Abe and Noguchi (1983b) and Pacheco and Sykes (1992). The studied earthquakes all occur in moderately seismic active regions, therefore our results may have significant consequences for hazard estimates in those regions.
Kiirunavaara (Kiruna) iron ore mine owned by LKAB (Sweden) is one of the largest underground mines. Miningstarted in 1898 as an open pit mine. In mid-1950, the mine started a transition to underground mining andpassed to only underground mining in 1962. More substantial problems with seismicity started in 2007-2008when the deepest mining level was 907 m (ca. 670 m below surface). By 2016, the mining production is at1,022–1,079 m Level (ca. 785–845 m below surface). More than one billion tonnes of ore have been extractedsince the beginning of mining. The average yearly production in recent years is 28 million tonnes.By 2016 the mine has the largest underground seismic system in the world with 204 operational geophones.The number of the sensors (geophones with natural frequencies of 4.5, 14, and a few of 30 Hz) changed withthe increasing of production depth. The major stages with seismic system upgrades are: August 2008–June2009 with 112 installed geophones, and July 2012–September 2013 with 95 installed geophones. During2016–2017 it is planned to install some additional 45 geophones.The study was carried out to identify some trends in seismicity as the mining goes deeper and to find thecorrelation with some main controlling parameters – volume and depth of the production in order to obtaininformation for future seismic hazard and risk analysis. Custom made applications within mXrap were utilisedto carry out the spatial variations of seismicity.The analysis showed substantial difference between the seismicity in the three studied blocks – 15/16, 28/30,and 33-37/34, with the weakest seismic activity in Block 15/16 (Mmax 1.6, maximum observed magnitude),followed by Block 28/30 (Mmax 2.2), and then largest seismicity in Block 33-37/34 (Mmax 2.2). The dailyseismicity rate increased substantially through the years only for Block 33-37/34. The seismicity correlatesstrongly with the production depth. In general a straightforward correlation between the production volumeand number of larger events (M > 0) was not found for the three studied blocks, assuming there are otherfactors affecting the seismicity, e.g. geological structures, areas with contrast in geomechanical properties,etc. The spatial variations of some seismic source parameters were traced for varying periods of time,depending on the major production stages (opening of new levels, full production, closing) for the threeblocks. The distributions of the cumulative seismic energy showed a maximum around and below theproduction. The cumulative seismic moment and number of events in most cases showed a maximum aroundand above the production, indicating caving in these areas. The static stress drop shows the largest valuesaround and below the production on the footwall side, corresponding also to the areas with increased stress.The energy index showed increased stresses in the same areas (EI > 1).This study is only the first overview of the seismicity in Kiruna Mine. For seismic hazard assessment and riskanalysis further more detailed studies with smaller time intervals need to be carried out to obtain more precisecorrelations between the seismic parameters and the production volume and depth, and other possible factorsaffecting seismicity (geological structures, areas with contrast geomechanical properties, etc.).
Over the past decades a number of small earthquakes have been recorded in Leamington - Ridgetown area along the north shore of Lake Erie (southwestern Ontario). A new seismic cluster is forming in this area, away from the already known clusters in Ontario. The new seismic area lays across the seismic area south of Lake Erie (along the Pennsylvania- Ohio border), known for some moderate induced events related to the oil production there. Another cluster related to the oil/gas production- in the region of Gobles, north of Lake Erie-has been documented and studied by Mereu et al. (1986). The induced seismicity is usually related not to the oil/gas production itself but to the water injection accompanying this production. The water injection is used in southern Ontario in Leamington - Ridgetown area to increase the oil/gas recovery from the existing reservoirs. The relationship between the new forming cluster and the ongoing oil/gas production north of Lake Erie is studied here. The parameters of the earthquakes in the area (hypocenter location, magnitudes, seismic moment, stress drop, and focal mechanism and/or seismic moment tensors for some events) are calculated using the POLARIS and Canadian National Seismic Network (CNSN) data. A temporary seismic network, consisting of four high-frequency three-component stations, has been installed in the fall of 2008 to record data from possible smaller events, not recorded by the permanent stations. The lithology, structural geology, and hydrology of the site are critical for determining if the water injection can induce seismic events. This type of data as well as data about the local tectonics (the existing faults) have been collected and analyzed. The main goal of this work was to find if any spatial or temporal correlations between the seismicity pattern and oil production/water injection exist. The preliminary results of the study suggest a correlation between the seismic activity and the oil/gas production. The study provides also additional information about the tectonic regime in southern Ontario and on throws some light on the hypothesis for induced seismicity due to the oil/gas production north of Lake Erie.
On October 20 2005 at 21:16 GMT, a magnitude mN 4.3 earthquake occurred in the southern part of Georgian Bay, approximately 12 km north of Thornbury, Ontario, Canada (latitude 44.67° N and longitude 80.46° W). This earthquake is the largest one in southern Ontario recorded by a local seismograph network and is of particular interest due to its location 90 km from a proposed long-term storage facility for high-level nuclear waste. The earthquake was felt along the southern shore of Georgian Bay with maximum intensity of IV MM. During the first 24 hours after the earthquake occurred, four portable ORION seismograph systems were installed to record possible aftershocks. The main shock on October 20 2005 was preceded by a foreshock 30 sec before it, and was followed by 5 aftershocks within 4-day period. All the epicenters of the foreshock and aftershocks were within 2.5 km from the epicenter of the main shock. The large amount of available data from the recently installed broad-band POLARIS seismograph stations, as well as the permanent CNSN stations and the temporary stations, gave us a unique opportunity to study the parameters of this event. The analysis of the foreshock-main shock-aftershock sequence indicated focal depths around 7 to 12 km. The focal mechanism calculated from the polarities of P-arrivals showed predominantly thrust mechanism of the main shock, with nodal planes oriented almost NW-SE. The focal mechanism is very similar to the predominant focal mechanism of the earthquakes in Western Quebec Seismic Zone but different from the predominant strike-slip focal mechanisms south of Lake Erie and the oblique slip mechanisms in western Lake Ontario. Aeromagnetic data reveal a prominent NW-SE structural fabric for the basement rocks beneath Georgian Bay, in good agreement with the orientation of the nodal planes. This structural fabric probably reflects mafic dykes (the Matachewan dyke swarm). The spectra of S-waves, recorded at 13 bedrock stations, were fitted with Brune’s model and used to calculate the seismic moment (3.6e+14 N.m), source radius (~ 400 m), stress drop (~ 20 bars), and moment magnitude (Mw 3.7). This seismic moment and calculated focal mechanism were used as initial approximation for seismic moment tensor inversion. The results of the inversion showed correspondence between the seismic moment and double-couple focal mechanism calculated from the moment tensor
Using data from 27 seismograph stations for the period 1990-2001, we have relocated 106 hypocenters of earthquakes with magnitudes from 0.9 to 5.4 in the region of the southern Great Lakes. Two complementary methods were used for relocation: a conventional least-squares approach (Lienert and Havskov, 1995) and joint hypocentral determination (Pujol, 2000). These two methods yielded mutually consistent spatial patterns of seismicity with an average difference of 3.7 km in epicentral locations and 1.1 km in focal depths. We show that the hypocenter locations are not very sensitive to realistic uncertainties in 1D crustal velocity. Our sharpened definition of zones of seismicity delineates several clusters beneath Lake Ontario, around Niagara Falls, and near the south shore of Lake Erie. These seismicity zones appear to correlate with areas where the regional magnetic data exhibit prominent short-wavelength (<5 km) linear anomalies. The magnetic anomalies are associated with basement structures that formed during the Precambrian (Mesoproterozoic) Grenville orogen. Both the seismicity and magnetic anomalies exhibit statistically significant preferred orientations at N40°E-N45°E, but the correlation of the earthquake clusters with specific aeromagnetic lineaments remains uncertain. Three preliminary focal mechanisms of earthquakes with magnitudes mN 3.1 to 3.8 show unusual normal faulting, with nodal planes in almost the same direction as the magnetic trends, N42°E-N52°E. Proximity of the earthquake clusters to large bodies of water, coupled with colinearity with magnetic anomaly trends, suggests that both surface water and pre-existing basement structures may play significant roles in controlling intraplate seismicity in the southern Great Lakes region
On 20 October 2005 at 21:16 UTC, a moderate earthquake (mN 4.3) occurred in an area of low seismicity within Georgian Bay, approximately 12 km north of Thornbury, Ontario (44.67° N, 80.46° W). Despite its moderate magnitude, it was exceptionally well recorded and is of particular interest because of its location 90 km from a proposed long-term storage facility for low- and medium-level nuclear waste. No damage was reported, but ground shaking was felt to a distance of 100 km. Within 24 hours after the mainshock, four portable seismograph systems were installed in the epicentral region. In total, eight events were recorded over a 4-day period, including a foreshock and six aftershocks. The unusually rich dataset from this moderate earthquake sequence enabled robust determination of hypocentral parameters, including well-constrained focal depths for most events. For the mainshock, we estimated a seismic moment of M0 4.5 × 1014 N m and corner frequency of 3.7 Hz, based on a spectral fit using Brune's source model. Least-squares waveform inversion of P and S phases yielded a double-couple focal mechanism with a reverse-sense of slip and northwest-striking nodal planes. The reverse mechanism and mid-crustal focal depths (10-12 km) are characteristic, in general, of more abundant seismicity located ∼200 km northeast of this event in the western Quebec seismic zone. These parameters differ, however, from shallow (2-6 km) earthquakes, with predominantly strike-slip mechanisms, observed near Lake Erie ∼200 km to the south. We attribute this north-south change in rupture mechanism to variations in crustal stress induced by postglacial isostatic rebound. Aeromagnetic data in and around the epicentral region reveal prominent northwest-striking lineations caused by Precambrian mafic dykes. Under midcrustal conditions, the dyke material is mechanically stronger than generally more felsic upper-crustal host rocks. We propose that where large dykes are favorably oriented with respect to the stress field, they may strongly influence the locations of intraplate earthquake rupture in Shield regions.
The conventional methods for determining the magnitude of an earthquake such as Richter magnitude, Nuttli magnitude etc. are based mainly on peak-to-peak amplitudes of different phases on the seismic trace. These magnitude scales were developed in the past during the days when we had only paper seismic recordings. The conventional method for determining moment magnitude obtained from seismic moment is measured actually as the low frequency level of the displacement spectrum. Both of the above mentioned methods ignore a lot of the information, which is provided in a modern digital 3-component seismogram. In this study we explore the applicability of total signal energy as the main parameter to be used in magnitude estimation. Over 2100 three-component seismic traces from 258 local and regional earthquakes recorded on the Southern Ontario Seismic /POLARIS networks were used in this study. To relate the new energy magnitude scale to the old ones, including Mw, we have calculated most known magnitude types and the seismic moment of the earthquakes. To carry out this work, an automatic procedure was developed for measuring the peak-to peak amplitudes, periods, duration, and signal energy for each seismic trace. For calculation of the seismic moment, an iterative technique was developed to separate the effects of source functions from site response and geometrical spreading and attenuation effects. We have compared our energy magnitude measurements with the other well-known magnitude measurements by monitoring the solution errors. Our results show that the measurements of total seismic signal energy in both the P- and S-wave trains can improve the precision of the earthquake magnitude significantly and reduce much of the scatter found in conventional magnitude measurements.
The conventional methods for rapid determination of earthquake magnitude are based mainly on peak-to-peak amplitudes of specific phases on a seismic trace. Today, broadband digital records are readily accessible in real time, enabling the use of more information from a seismogram for rapid magnitude calculation. The aim of this work is to introduce a new magnitude scale for routine seismological analysis, denoted ME (P-wave, S-wave+coda, or both). This magnitude scale uses the signal energy and is illustrated here with a case study from southern Ontario/western Quebec (Canada). Traditional types of magnitude scales, based on the estimated maximum velocity (mb) and Richter local magnitude (ML), as well as the moment magnitude (MW), and some other magnitude types, based on the coda energy (MCoda) and ehvelope area (MEnv) are also computed for the study area for comparative purposes. Ihe proposed approach employed for this study can be easily applied to any other region of the world. The developed automatic procedure allowed the simultaneous computation of different magnitudes and different trace components and types of waves. The data used for this research are from 238 well-recorded earthquakes between 1991 and 2006 in southern Ontario/western Quebec/northern Ohio/northern NY State (1.0 < mN < 5.5). The results of our work show that, in general, magnitude values based on signal energy ME give less scattered estimates than magnitude values based on peak-to-peak measurements. We recommend using ME (S + coda) scale (vertical component) for quantifying the earthquakes in the study area in the future. The magnitude formula for this scale is given by ME = 0.5log ẼS + 0.92logD + 3.56 + S, where ẼS is the signal energy defined here (∑vs2 Δt, vS is measured in mm/s, Δt is the sample interval in seconds), D is the epicentral distance in km, and S is the station correction. The new ME magnitude can be used for a quick estimate of the MW magnitude for the study region using the relationship: MW = ME - 0.51 (for earthquakes with ME ≥ 2.6), obtained here.
Three local seismic systems were installed by August 2015 in deep underground mines in Sweden – Zinkgruvan Mine (Lundin Mining AB), Garpenberg Mine (Boliden AB), and Kiirunavaara Mine (LKAB) as part of a project for developing new methods for Evaluating the Rock Support Performance (ERSP, Vinnova). The areas were chosen within the most probable volumes where large rockbursts can be expected. The local systems were installed at mine levels between 730 and 1150 m in different mines. The horizontal extend of each instrumented areas is between 70 and 100 m. The seismic system in each mine is a combination of uni-axial and three-axial 4.5 Hz geophones installed on the surface, in shallow (~0.5 m) and deeper (6-9 m) boreholes in profiles across drifts. These profiles are in close proximity to profiles with extensometers, instrumented bolts, and observation holes. The seismic systems are manufactured and installed by the Institute of Mine Seismology (IMS). The aim of the seismic systems is to record the seismic events that occur in the vicinity of the instrumented areas and provide valuable data about the variability of seismic waveforms around the underground openings and changes when seismic waves approach them. Data is used to study: 1) the attenuation/decrease of the maximum ground velocity (PPV) with the distance, especially at small distances; 2) site effects, including maximum amplitudes, predominant frequency, and duration of the seismic signals, 3) the attenuation/amplification of the seismic waves approaching the underground opening. The final aim is to obtain new information that can be used for improved requirements for the rock support design in rockburst prone areas.The installation of the seismic systems started in May 2015 (Zinkgruvan Mine) and was completed by August 2015. They run mostly in triggered mode with initial automatic arrival time picking and source parameter calculation and subsequent manual processing of seismic event of interest. More than 200,000 seismic events with magnitude from -4.5 to 2.0 were recorded by December 2015. At present only a small portion of all data was processed manually and the procedures for processing of the events were developed on this subset. The first results from the monitoring showed that there are differences in the amplitudes and shape of the seismic signals recorded by the sensors installed in deeper borehole (behind the most blast-damaged zone (6 – 9 m)) and close to the surface (0.5 m) or on the surface of the openings. There are also differences between the waveforms recorded on the walls and the roof along the same profiles or on nearby profiles. Data from the investigated rockbursts showed maximum velocity recorded from a seismic events at close distances with magnitude larger than 0.5 in the order of 10 cm/s with clipping levels 10 – 20 cm/s.
Earthquakes manifested in a small volume of the earth's crust are analysed, presuming that the orientation of the rupture disturbances is conditioned by the local neotectonic conditions. A relation is sought between the seismic manifestations and the modern crushing structures in the earthquake region. In this very case, the earthquakes in the Velingrad district (Bulgaria) are analysed, which began with an earthquake of the 5,3 magnitude and a maximum intensity of the VII-VIII degree by the Medvedev-Schponheuer- Karnik scale and the followed 504 earthquakes for a period of 179 days. From the data obtained, a presence of a complex ridge-like structure can be presumed, which develops in a valley depression on the background of a vault block rising in the limits of the granite layer of the earth crust, to a 15 km depth. The frequency of the earthquakes is obtained as a function of the difference (tsg - tpg) and the greatest earthquake frequency is shown to be in the interval of 0,8-1,4 sec and the greatest difference of times corresponds to a 20 km hypocentral distance
Earthquakes that occurred between 1985-1991 in SW Bulgaria and border regions with Greece and the Former Yugoslav Republic of Macedonia (bounded by 22.0°E-22.4°E and 40.8°N-42.4°N) are located to define the characteristics of the present seismicity. The hypocenters of 2657 earthquakes with magnitudes 0.3 < ML < 5.1 are determined using the program HYPOLOC (modification of the HYPO71 program) and a velocity model is obtained for the local area of SW Bulgaria. Earthquake tend to cluster in seven areas. The seismic activity in the clusters changes with time and correlates mainly with the occurrence of the strongest events in each cluster. The clusters are classified by the density of earthquake epicenters. Following this classification, those with the 'highest' seismicity are characterised as the Krupnik cluster, and those with the 'lowest' seismicity, the Butkovsko (Kerkini) Lake cluster. The location and shape of the cluster areas are related to boundaries between areas of subsidence (grabens, depressions) and areas of uplift (horsts, swells). A few areas without seismicity, which are related to horst structures are outlined. Historical data exists for each of the obtained cluster areas (from 52) for earthquakes of magnitude larger than 5.0.
We have developed a simple semblance-weighted stacking technique to estimate crustal thickness and average V P/V S ratio using teleseismic receiver functions. We have applied our method to data from 32 broadband seismograph stations that cover a 700 × 400 km 2 region of the Grenville orogen, a 1.2-0.98 Ga Himalayan-scale collisional belt in eastern North America. Our seismograph network partly overlaps with Lithoprobe and other crustal refraction surveys. In 8 out of 9 cases where a crustal-refraction profile passes within 30 km of a seismograph station, the two independent crustal thickness estimates agree to within 7%. Our regional crustal-thickness model, constructed using both teleseismic and refraction observations, ranges between 34.0 and 52.4 km. Crustal-thickness trends show a strong correlation with geological belts, but do not correlate with surface topography and are far in excess of relief required to maintain local isostatic equilibrium. The thickest crust (52.4 ± 1.7 km) was found at a station located within the 1.1 Ga mid-continent (failed) rift. The Central Gneiss Belt, which contains rocks exhumed from deep levels of the crust, is characterized by V P/V S ranging from 1.78 to 1.85. In other parts of the Grenville orogen, V P/V S is found to be generally less than 1.80. The thinnest crust (34.5-37.0 km) occurs northeast of the 0.7 Ga Ottawa-Bonnechere graben and correlates with areas of high intraplate seismicity
Aftershock series of even comparatively small seismic events can pose a risk to the mining operation or the personnel in deep underground mines as the main shocks and some of the aftershocks can cause damage in the rock mass. Stochastic modeling was applied in this study for the analysis of the temporal evolution of aftershock occurrence probability during a M1.85 aftershock sequence in Kiirunavaara Mine, Sweden. The Restricted Epidemic-Type Aftershock Sequence (RETAS) model was chosen for estimation of the aftershock occurrence probability. This model considers all events with magnitude above the magnitude of completeness M0 and has the advantage of including the Modified Omori Formula (MOF) model and Epidemic-Type Aftershock Sequence (ETAS) model as its end versions, considering also all intermediate models. The model was applied sequentially to data samples covering cumulative periods of time, starting from the first 2 h after the main event and increasing them by 2 h until the period covered the entire 72-h sequence. For each sample, the best-fit RETAS version was identified and the probability of a M ≥ 0.5 aftershock for every next 2 h was determined through Monte Carlo simulation. The feasibility of the resulting probability evolution for suspension and re-starting of the mining operations was discussed together with possible prospects for future development of the methodology.
Mining induced seismicity is an undesired consequence of mining operations, which poses significant hazard to miners and infrastructures and requires an accurate analysis of the rupture process. Seismic moment tensors of mining-induced events help to understand the nature of mining-induced seismicity by providing information about the relationship between the mining, stress redistribution and instabilities in the rock mass. In this work, we adapt and test a waveform-based inversion method on high frequency data recorded by a dense underground seismic system in one of the largest underground mines in the world (Kiruna mine, Sweden). A stable algorithm for moment tensor inversion for comparatively small mining induced earthquakes, resolving both the double-couple and full moment tensor with high frequency data, is very challenging. Moreover, the application to underground mining system requires accounting for the 3-D geometry of the monitoring system. We construct a Green's function database using a homogeneous velocity model, but assuming a 3-D distribution of potential sources and receivers. We first perform a set of moment tensor inversions using synthetic data to test the effects of different factors on moment tensor inversion stability and source parameters accuracy, including the network spatial coverage, the number of sensors and the signal-tonoise ratio. The influence of the accuracy of the input source parameters on the inversion results is also tested. Those tests show that an accurate selection of the inversion parameters allows resolving the moment tensor also in the presence of realistic seismic noise conditions. Finally, the moment tensor inversion methodology is applied to eight events chosen from mining block #33/34 at Kiruna mine. Source parameters including scalar moment, magnitude, double-couple, compensated linear vector dipole and isotropic contributions as well as the strike, dip and rake configurations of the double-couple term were obtained. The orientations of the nodal planes of the double-couple component in most cases vary from NNW to NNE with a dip along the ore body or in the opposite direction.
A Mw 3.1 earthquake occured in Lake Ontario along the United States-Canada border, about 30 km south from Port Hope, Ontario, Canada, on 4 August 2004. Despite its small size, the shock was very well recorded by broadband seismographic stations deployed in recent years in Ontario, Canada, and in New York State. More than 40 broadband stations at local and regional ranges provided high-quality digital data. Waveform data analysis constrained the source at a depth of 4 (±2) km, which places the shock in the shallow Precambrian basement beneath Paleozoic platform deposits. The source mechanism from the regional waveform inversion for the double-couple moment tensor is predominantly strike-slip faulting. A NS striking (8°) nodal plane dipping to the east (dip = 59°) is the likely fault plane which represents right-lateral strike-slip motion. The subhorizontal P-axis orientation (trend = 234° and plunge = 12°) is consistent with the maximum horizontal compressional stress (SHmax) direction in eastern North America. Although the 4 August 2004 event is a small shock and has the seismic moment of M0 = 4.45 (±2.30) × 1013 Nm, it is the largest instrumentally recorded earthquake that has occurred in Lake Ontario. This and other significant earthquakes in the region suggest a broad-scale strike-slip faulting stress regime with a shallow seismogenic layer in the Erie-Ontario Lowlands region. The shallow focal depths of earthquakes in the region increase the risk of higher ground shaking compared to other seismic zones in northeastern North America with a deeper seismogenic layer.
Strong mining‐induced earthquakes are often followed by aftershocks, similar to natural earthquakes. Although the magnitudes of such in‐mine aftershocks are not high, they may pose a threat to mining infrastructure, production, and primarily, people working underground. The existing post‐earthquake mining procedures usually do not consider any aspects of the physics of the mainshock. This work aims to estimate the rate and distribution of aftershocks following mining‐induced seismic events by applying the rate‐and‐state model of fault friction, which is commonly used in natural earthquake studies. It was found that both the pre‐mainshock level of seismicity and the coseismic stress change following the mainshock rupture have strong effects on the aftershock sequence. For mining‐induced seismicity, however, we need to additionally account for the constantly changing stress state caused by the ongoing exploitation. Here, we attempt to model the aftershock sequence, its rate, and distribution of two M≈2 events in iron ore Kiruna mine, Sweden. We could appropriately estimate the aftershock sequence for one of the events because both the modeled rate and distribution of aftershocks matched the observed activity; however, the model underestimated the rate of aftershocks for the other event. The results of modeling showed that aftershocks following mining events occur in the areas of pre‐mainshock activity influenced by the positive coulomb stress changes, according to the model’s assumptions. However, we also noted that some additional process not incorporated in the rate‐and‐state model may influence the aftershock sequence. Nevertheless, this type of modeling is a good tool for evaluating the risk areas in mines following a strong seismic event.
Studying the deterioration of shotcrete due to freezing and thawing is important for improvement of the understanding of the failure mechanisms/debonding of shotcrete in cold regions. Water leakage in a tunnel leads to ice growth during freezing temperature and ultimately creates favorable environment for fallouts of shotcrete and rock. Repeated freezing and thawing of shotcrete lead to development of new micro cracks and propagation of pre-existing micro cracks. In this study, test panels of granite with dimension 800 x 800 x 80 mm covered with 50-mm thick shotcrete were subjected to freezing and thawing action in a controlled environment. The initiation and the development of freeze-induced micro cracks in shotcrete-rock interface were studied by continuously monitoring acoustic emissions (AE) and temperature. The clustering of the AE events during freezing and thawing indicates that micro cracks appeared in the shotcrete-rock interface and caused adhesion failure. The larger number of AE events in the panels, with access to water during freezing, confirmed that water contributes to material deterioration and also reduces the adhesive strength. The test results showed that most of the acoustic emission occurred during the freezing cycle and the number of acoustic emission events did not increase with the successive increase of the number of freezing and thawing cycles.
Road and railway tunnels in cold regions are often affected by problems related to water leakage and freezing temperatures. Water leakage in a tunnel leads to ice growth when the temperature goes below freezing and creates favourable environment for fallouts of shotcrete and rock. This paper presents results and observations from laboratory freezing – thawing experiments on rock blocks covered with shotcrete and focuses on the degradation of the shotcrete-rock interface due to ice growth.. The initiation and the development of freeze-induced micro cracks in shotcrete-rock interface were studied by continuously monitoring acoustic emissions (AE) and temperature. The clustering of the AE events during freezing and thawing indicates that micro cracks appeared in the shotcrete-rock interface and caused adhesion failure. The larger number of AE events in the panels, with access to water during freezing, confirmed that water contributes to material deterioration and also reduces the adhesive strength.
The Southern Ontario Seismic Network (SOSN) consists of eleven three-component short-period seismic stations, located mainly in the Toronto-Hamilton-Niagara area of Ontario, Canada. The network has been in operation by the University of Western Ontario (UWO) for Ontario Power Generation (OPG) since 1991 with the purpose of obtaining information on the seismicity and seismic hazards of a region of southern Ontario in which a number of nuclear power stations are located. Over the past decade, an average of more than ten local earthquakes per year in the western Lake Ontario area was detected by the SOSN. Most of the events were in the 2–3 magnitude (MN) range. The largest events during this time took place in the surrounding regions—Pymatuning, northwestern Pennsylvania (285 km southwest from Toronto, just south of Lake Erie, 25 September 1998, MN 5.4), northern Ontario/Quebec border (325 km north of Toronto, 1 January 2000, MN 5.2), Ashtabula, Ohio (262 km southwest of Toronto, 26 January 2001, MN 4.4), and Au Sable Forks, New York (436 km east of Toronto, 20 April 2002, MN 5.1). The largest earthquake (MN 3.8) in the western Lake Ontario region during the past ten years occurred on 26 November 1999 in Lake Ontario, 16 km southeast of the town of Pickering, which lies just east of Toronto. The estimated location uncertainty (±2 km) is significantly better than that which was possible before 1991. The focal depths, though poorly constrained for most events, are shown to lie in the 3–15 km range, well within the Grenvillian rocks of the Precambrian Shield. The new seismicity map shows that a definite pattern is emerging in the SOSN data set in Lake Ontario, one which is significantly different from the past historical earthquake patterns obtained when the instrumental coverage was poor. Most events occur in scattered clusters in the western part of Lake Ontario and the northwestern corner of New York State. The area of seismicity does not extend significantly to the north of western Lake Ontario and appears to end to the west rather abruptly along a 30 km small fault line running from south of Hamilton in a north-northeasterly direction to Burlington, Ontario. Although the area of seismicity coincides with a region of linear magnetic anomaly trends (suggesting a strong structural fabric in the basement rocks), the correlation of seismicity of the new SOSN data set with magnetic lineaments is still unclear. The cause of the seismicity is speculated to be related to water flows along various fissures below the lake. It is known from induced seismicity studies of reservoirs that the presence of fluids can cause earthquakes by changing the pore pressure and reducing the friction along any faults which may be present. From seismic reflection studies, dipping structures and shear zones have been imaged to extend southeastward under Lake Ontario. This may explain why most of the earthquakes are occurring under the lake or southeast of the lake.
We analyzed over 3000 Fourier spectra from 370 earthquakes of energy magnitude (M-E) 1.1-6.0 recorded by the Southern Ontario Seismic Network (SOSN)/POLARIS networks during the period 1991-2010 in the area of southern Ontario and western Quebec. We employed a range of velocity stacking methods to significantly reduce the problem of variability due to wave scattering. This enabled us to determine underlying nonrandom spectral features, including source effects, site effects, and anelastic attenuation effects on spectral shape. The analysis technique is that we stack the velocity spectra of the whole observed data set into one or two bins and then compare that sum (the observed stack) with the theoretical expectation for corresponding stacks of simulated signals (the theoretical stack) for a given set of input parameters. A grid-search technique is used to find the input-parameter combination that optimizes the agreement between the observed and theoretical stacks. By stacking the spectra in different ways, different underlying spectral features are explored. We find the method works surprisingly well, allowing us to determine the apparent anelastic attenuation effects on the spectral shape, the average effect of site response, and some basic features of the source spectra. The key results of our paper: (1) there is no unique pair of values of the coefficients Q(0) and n of the frequency-dependent Quality factor relationship Q=Q(0)f(n), but there exist pairs of Q(0) and n along a curve in Q(0)-n space that are equivalent in terms of their effect on spectral shape; (2) the relationship between log corner frequency and energy magnitude (M-E) is linear, with a slope close to (-0.22) that is consistent with constant-width faulting for the studied small-to-moderate events; (3) the relation between moment magnitude M and energy magnitude M-E was found to be M = 8/9 M-E.
Damage caused by the earthquake of 7 December 1986 (M = 5.7) and its aftershocks in low-storey residential buildings in the town of Strazhitza, Bulgaria, situated in the epicentral zone are systematized. A scale of damages is compiled for two types of buildings which is coordinated with the MSK-64 scale. The territorial distribution of the seismic intensity is compared with the results of the detailed engineering-geological study. A correlation between the Quarternary deposit thickness and the observed seismic intensity is obtained
The Nakamura method, which utilizes the Horizontal to Vertical Spectral Ratio (HVSR) analysis, is widely used for seismic microzonation studies. The HVSR is an easy tool for estimation of site response resonances based on recorded ambient noise; however, it gives amplifications at resonant frequencies that are poorly correlated to the actual amplifications during strong ground motion.
Generally, the site response, including any resonant effects, depends on the amplitude, frequency and duration of ground motion. An approach was proposed previously by McGuire [1], in which the transfer function of the soil response was approximated as a Single Degree of Freedom (SDOF) oscillator with one resonant frequency, obtained from the maximum in HVSR. A new approach is developed here, in which the entire HVSR curve is approximated by a manageable set of parallel band-pass resonators. Each individual oscillator is defined by three parameters: center frequency, gain, and steepness (Q factor). This approximation allows for the development and use of an analytical model of the HVSR curve.
The application of the new approach is demonstrated on data recorded by the stations of the Southern Ontario Seismic Network (SOSN/Polaris), which have well studied characteristics and site response [2,3]. Data collected at each site consists of noise recordings to obtain the HVSR, as well as earthquake records. The analytical HVSR curves for each station are used to remove the site effect component from the recorded seismograms.
A three-station broadband network was installed around the Bruce Nuclear site at the beginning of August 2007, to monitor microearthquakes within a 50-km radius of the plant. The seismic network was equipped with borehole stations installed at cased boreholes at depths of 25, 27 and 40 m, and temporary surface stations at the same sites. The aim of the doubled identical equipment (Geotech Instruments' KS2000) was to compare the records of local, regional and teleseismic events, and seismic noise and to obtain results about the noise reduction, attenuation and the site response at each station. During the design and installation of the seismic network different geophysical surveys were carried out: refraction seismic profiles, vertical seismic profiling, and noise level measurements at different depths along the borehole. The obtained velocity models were used for modeling of the site response and finally comparison with the real data from the parallel borehole-surface recordings, and measured predominant frequencies using the Nakamura's HVSR method. The real noise reduction estimated from the parallel recordings was compared with the predicted 10 dB noise reduction. A practical conclusion was drawn out about the optimum borehole depths for instrument installation based on the noise reduction / attenuation balance and signal-to-noise ratio with the depth. The seismic threshold magnitude for the monitored area estimated at the design stage was compared with the threshold magnitude obtained from the real data.
The Nakamura method, which involves horizontal-to-vertical spectral ratio (HVSR) analysis, is widely used for seismic microzonation studies. The noise from local traffic in city conditions presents a challenge for the application ofHVSR analysis. This article presents a technique developed for separation of the transient noise due to local traffic (high-level noise) and background ambient noise (low-level noise) and the application of theHVSR analysis to both partitions of the noise. This approach is applied to identify the predominant frequencies for almost 200 noise samples from the Greater Toronto area. The results demonstrated that the developed technique is effective and allows estimation of the fundamental resonant frequency in theHVSR in urban environment, even in the presence of intensive nearby traffic. The interpretation of the obtained results showed that, most probably, the lower (fundamental) frequency appears due to multiple reflections from the overburden/bedrock boundary. In some cases, a resonance with higher amplitude is dominant, and it is due to a contrast boundary between soil layers in the overburden
Back analysis for evaluation of the merits of the short-term seismic hazard indicators (precursors) used in the mines and their potential application for early warning was carried out for fourteen seismic events that potentially caused damage in Kiirunavaara Mine, Sweden, selected according to our designed criteria. Five short-term hazard indicators: Seismic Activity Rate (SAR), Cumulative Seismic Moment (CSM), Energy Index (EI), Cumulative Apparent Volume (CAV) and Seismic Apparent Stress Frequency (ASF) were tested. The behaviour of the indicators was studied using the parameters of all seismic events within a sphere around the hypocenter location of the analyzed seismic source within one month before the main (damaging) event. The size of the sphere equals the estimated radius of the analyzed seismic source (area of inelastic deformation). mXrap software (Australian Centre for Geomechanics) was used for data visualization, manipulation, analysis and extraction. The results from the main analysis showed a good agreement between the expected and actual behaviour of the SAR, CSM and CAV indicators. In overall, CSM and CAV ranked the highest positive/expected behaviour followed by SAR (Table 3). The EI and ASF ranked lowest and showed to be sensitive to the number of events within the source sphere. The rate of false warnings and missed warnings was also investigated for the 25 days-long period before the damaging events. A similar trend was observed as for the main analysed event. The results from this study can be used for further improvement of the short-term hazard estimations and early warning system in deep underground mines.
Forty-six mining-induced seismic events with moment magnitude between −1.2 and 2.1 that possibly caused damage were studied. The events occurred between 2008 and 2013 at mining level 850–1350 m in the Kiirunavaara Mine (Sweden). Hypocenter locations were refined using from 6 to 130 sensors at distances of up to 1400 m. The source parameters of the events were re-estimated using spectral analysis with a standard Brune model (slope −2). The radiated energy for the studied events varied from 4.7 × 10−1 to 3.8 × 107 J, the source radii from 4 to 110 m, the apparent stress from 6.2 × 102 to 1.1 × 106 Pa, energy ratio (Es/Ep) from 1.2 to 126, and apparent volume from 1.8 × 103 to 1.1 × 107 m3. 90% of the events were located in the footwall, close to the ore contact. The events were classified as shear/fault slip (FS) or non-shear (NS) based on the Es/Ep ratio (>10 or <10). Out of 46 events 15 events were classified as NS located almost in the whole range between 840 and 1360 m, including many events below the production. The rest 31 FS events were concentrated mostly around the production levels and slightly below them. The relationships between some source parameters and seismic moment/moment magnitude showed dependence on the type of the source mechanism. The energy and the apparent stress were found to be three times larger for FS events than for NS events.