V. B. Smirnov

Physical modeling of the formation and evolution of seismically active fault zones

Acoustic emission (AE) in rocks is studied as a model of natural seismicity. A special technique for rock loading has been used to help study the processes that control the development of AE during brittle deformation. This technique allows us to extend to hours fault growth which would normally occur very rapidly. In this way, the period of most intense interaction of acoustic events can be studied in detail. Characteristics of the acoustic regime (AR) include the Gutenberg-Richter b-value, spatial distribution of hypocenters with characteristic fractal (correlation) dimension d, Hurst exponent H, and crack concentration parameter Pc. The fractal structure of AR changes with the onset of the drop in differential stress during sample deformation. The change results from the active interaction of microcracks. This transition of the spatial distribution of AE hypocenters is accompanied by a corresponding change in the temporal correlation of events and in the distribution of event amplitudes as signified by a decrease of b-value. The characteristic structure that develops in the low-energy background AE is similar to the sequence of the strongest microfracture events. When the AR fractal structure develops, the variations of d and b are synchronous and d = 3b. This relation which occurs once the fractal structure is formed only holds for average values of d and b. Time variations of d and b are anticorrelated. The degree of temporal correlation of AR has time variations that are similar to d and b variations. The observed variations in laboratory AE experiments are compared with natural seismicity parameters. The close correspondence between laboratory-scale observations and naturally occurring seismicity suggests a possible new approach for understanding the evolution of complex seismicity patterns in nature.

Generation and evolution of subduction-related batholiths from the central Urals: constraints on the P-T history of the Uralian orogen

Abstract Batholiths from the accreted terranes in the Urals were generated by repeated episodes of melting and intrusion. Verkhisetsk, the largest and one of the most complex subduction-related batholiths from the Urals, comprises an outer envelope of older tonalites, trondhjemites and granodiorites dated at 315–320 Ma and equilibrated at 6 kbar, intruded by an inner core of younger granodiorites, adamellites and granites dated at 275–290 Ma and equilibrated at 4 kbar. Older rocks have a high-A1 TTD/adakite chemistry, ϵ320Machur(ND) ≈ 2–5, and initial 87 Sr 86 Sr ≈ 0.7043 , with pronounced lateral zoning marked by the increase of LREE/HREE, Cr, Ni, and Mg eastwards. They were originated by melting of metabasalts in a subducted slab of young hot oceanic lithosphere with a temperature/depth trajectory that intersected the garnet-in univariant at ≈ 1050°C and 13 kbar, thus causing the lateral zoning. We believe that such a ‘warm’ trajectory can be attributed to the oblique subduction of a young lithosphere. Younger rocks range from Na-rich metaluminous granodiorites to K-rich peraluminous two-mica granites with relict epidote crystals, and have almost the same isotopic signature as older rocks, from which they were generated by anatexis as a consequence of a melting event that occurred throughout the Urals at 275–290 Ma and produced most of the huge continental-type batholiths on the eastern side. This melting event is tentatively supposed to be the result of underplating by mafic magmas, which also caused the growth of the Urals crust from the Moho downwards until it reached the increased thickness revealed by seismological studies.

Regularities in transient modes in the seismic process according to the laboratory and natural modeling

Regularities in the excitation and relaxation of rock failure were revealed in a series of laboratory experiments. Similar regularities are found also in natural conditions. A physical idea and its mathematical description are suggested for explaining the obtained experimental data. The aim of the experiments was to understand the character of excitation of the failure, triggered by the external impact, and its relaxation after the cessation of the pressure, depending on the intensity of the acting stresses. Different rates of increase in the initiating strains result in different acoustic responses that reflect the development of failure. At the higher rates of deformation, the observed process was similar to the aftershock sequences, and at the lower, to the seismic swarms. The character and parameters of the acoustic response change with the increase in the acting strains. The patterns of the changes exhibit several regularities. In case of the swarm-like activity, the time of maximum activity (and, correspondingly, the beginning in its decay) increases with the increase in acting strains. In case of the aftershock-like activity, the level of applied strains determines the parameters of the Omori’s law. The delay in the power-law’s decrease in activity increases with the growth of the load (similar to the increasing time until the beginning of the decay in the swarm-like activity). Similar regularities are defined in natural conditions in the experiments on the rock’s failure induced by water infusion into a borehole (Soultz-sous-Forêts, France). A hypothesis of competitive excitation and relaxation is suggested for explaining the observed experimental data. Mathematical modeling has confirmed the validity of this hypothesis.

Simulation of Triggered Earthquakes in the Laboratory

The experiments were conducted for the study of stick-slip at the contact between two granite blocks. Three cases were studied under the following conditions: 1) the increase of load at a constant rate; 2) the additional application of sinusoidal oscillations in the frequency range from 1 to 30 Hz; 3) subjection to the impulse in the kilohertz frequency range. The imposition of sinusoidal oscillations with the amplitude of 15% of the maximal load caused the reduction of time by 10% for the discussed in terms of durability. The high frequency impulse influence increased this effect and also caused essential changes in the amplitude of elastic oscillations generating during the stick-slip. The trigger phenomena should be integrated in prediction models of the time and magnitude of earthquakes.

Joint Inversion of the Differential Satellite Interferometry and GPS Data: a Case Study of Altai (Chuia) Earthquake of September 27, 2003

Based on the data of differential satellite interferometry, the field of displacements of the Earth’s surface in the line-of-sight direction is determined for the region of the Altai Earthquake that struck on September 27, 2003. The displacements are estimated for unforested areas of Chuia and the Kurai depressions, and for a part of their mountainous surroundings. In that part of the region where unwrapping of the data was possible, the amplitude of displacements amounts up to 150 cm for Chuia and 100 cm for the Kurai depressions. In order to locate the surface of the seismic rupture and to find the field of displacements on this surface, the method for the combined inversion of the displacements data, provided by satellite interferometry (the present work) and geodesy [Gol’din et al., 2005], is suggested and applied. The admissible range of the parameters of the rupture was specified from the seismology and seismotectonics data.The combined use of geodetic and satellite interferometry data makes the solution of the inverse problem more stable and yields a seismic momentum estimate, which is consistent with the seismological determinations. We discuss the possible contributions of various postseismic processes; in particular, based on analyzing the energy of the aftershocks, we assess the contribution of the postseismic creep to the displacements, determined from the interferometry and geodesy data, for different coseismic and postseismic time intervals.

Reservoir induced seismicity in the Koyna–Warna region, India: Overview of the recent results and hypotheses

The state of the art in the geological and geophysical study of the region of Koyna and Warna water reservoirs is reviewed. The probable geodynamical factors of induced seismicity are discussed. The detailed geophysical surveys, satellite geodetic data, and time history of the seismicity in the region reveal a complicated pattern of the structure and recent geodynamics of the region. The existing data suggest that the induced seismicity is here most likely to be caused by the regional (intraplate) stresses driving the displacements along the orthogonal network of the faults whose strength has dropped and continues decreasing due to the reservoir impoundment and operation processes. The evolution of the seismicity which started immediately after the rapid filling of the Koyna reservoir in the region of the dam, then rapidly expanded southwards and eventually became concentrated in the region of the subsequently constructed Warna reservoir shows that seismic events can be initiated by a number of factors whose contributions may vary with time. The key ones among them include reservoir loading and its seasonal variations; water saturation of the faults which guide the propagation of the front of fracture, increased permeability, and, probably, mineral transformations (hydrolysis) under the water level fluctuations in the reservoirs; and displacement of the front of the high pore pressure down to the main source zone of the earthquakes at a depth of 6–8 km. Based on the analysis presented in the paper, we outline the directions of the future research aimed at studying the nature and dynamics of induced seismicity in the region of large water reservoirs.

Seismic response to electromagnetic sounding of the Earth’s lithosphere

The seismic catalogues of 1967–2008 for the Bishkek geodynamical test site are analyzed for the purpose of studying the response of seismic activity to the electromagnetic sounding of the Earth’s crust during two series of field experiments with high-power controlled sources. The first series of the experiments, which were carried out in 1982–1990, utilized the pulses provided by a magnetohydrodynamic (MHD) generator. The sounding signals in the second series of the experiments (2000–2005) were generated by the capacitor-thyristor source ERGU 600-2. In these experiments, temporal variations of the set of statistical parameters characterizing the seismicity, which are typically used in the studies of the background and transient modes of seismicity, were investigated in a selected spatial domain within 150 km from the current electrodes. In terms of time, the analysis was conducted on two levels of detail. The study on a temporal scale of a few years was focused on the variations that preceded, accompanied, and followed the series of the experiments, while the day-scale analysis considered variations that were observed within 10 days after each sounding event. The day-scale analysis yields the following results. The slope of the frequency-magnitude diagram of the earthquakes (b value) during the sounding events is substantially larger than its background value. The slope of the graph gradually becomes gentler within about a day and a half after termination of sounding. The seismic activity slightly enhances during the interval of sounding and abates after its termination to a minimum, which corresponds to the interval of decreasing b value. This character of variations in seismicity differs from the scenario previously established for other transitional seismic regimes. The analysis on a temporal scale of a few years revealed variations in the studied parameters of the seismicity, some of which fall in both sounding intervals of 1983–1990 and 2000–2005. However, these variations are not unique; their character and durations suggest their being associated with the processes of preparation and after effects of the strong earthquakes that occurred in the vicinity of the sounding dipole.

Triggered and tectonic driven earthquakes in the Koyna–Warna region, western India

Two strong M > 5.0 earthquakes within a span of six months occurred in a triggered seismicity environment in the Koyna–Warna region in western India in 2000. The region is experiencing continued seismicity since the last five decades indicating that this region is close to critical stresses and minor perturbations in the stresses due to reservoir loading and unloading can trigger earthquakes. In the present study we applied the technique developed for identification of prognostic anomalies for tectonic earthquakes to the Koyna–Warna catalogue prior to these two earthquakes with an aim to study the process of source preparation for triggered earthquakes. In case of tectonic earthquakes, unstable conditions in a source zone develop gradually leading to a metastable zone which shows variations in certain seismicity parameters known as prognostic anomalies. Our results indicate that the variations in seismicity parameters before the two strong earthquakes in the Koyna region have a pattern of prognostic anomalies typical of tectonic earthquakes. We conclude that initiation of failure in a metastable zone can be caused both, by external impacts, reservoir loading and unloading in our case, and internal processes of avalanche-like failure development.

Patterns of seismic swarm activity in the Corinth Rift in 2000–2005

Based on the data of the detailed earthquake catalog provided on the website of the Corinth Rift Laboratory, zones of swarm activity are revealed and the variations in the statistical parameters of seismic swarms that occurred in the western part of the Gulf of Corinth are calculated. The preliminary analysis of the catalogue is carried out; the magnitude of completeness and the accuracy of the location of the earth-quake are estimated; the changes in these parameters associated with the development of the observational network are assessed. The b-value (b-values) and the cluster dimension of the set of hypocenters are estimated, and time variations in these parameters in the course of the evolution of swarm activity are revealed. The style of changes in the parameters characterizing the seismic regime during intervals of swarm activity indicates that the process of failure exhibits scale redistribution over the course of time, changing from upscaling (progression from smaller to larger scales) at the stage of increasing seismicity to downscaling (progression from larger to lower scales) at the stage of decay. These particular features of enhancement and reduction of swarm seismicity are qualitatively similar to the scenarios of source preparation and aftershock relaxation of strong earthquakes. The pattern of variations of the swarm seismicity studied is similar to those identified in the previous laboratory and field modeling of various transient modes of seismicity. This fact confirms the relevancy of the retrieved results and conclusions based on the laboratory studies of transient modes, and suggests that the latter have a universal governing mechanism.

Bayesian estimation of the Modified Omori Law parameters for the Iranian Plateau

The forecasting of large aftershocks is a preliminary and critical step in seismic hazard analysis and seismic risk management. From a statistical point of view, it relies entirely on the estimation of the properties of aftershock sequences using a set of laws with well-defined parameters. Since the frequentist and Bayesian approaches are common tools to assess these parameter values, we compare the two approaches for the Modified Omori Law and a selection of mainshock–aftershock sequences in the Iranian Plateau. There is a general agreement between the two methods, but the Bayesian appears to be more efficient as the number of recorded aftershocks decreases. Taking into account temporal variations of the b-value, the slope of the frequency-size distribution, the probability for the occurrence of strong aftershock, or larger main shock has been calculated in a finite time window using the parameters of the Modified Omori Law observed in the Iranian Plateau.