Complex Source Time Functions Research Articles

Real‐Time Seismogeodetic Earthquake Magnitude Estimates for Local Tsunami Warnings

AbstractWe estimate a seismogeodetic earthquake moment magnitude using unclipped, broadband velocity and displacement waveforms from collocated Global Navigation Satellite Systems and seismic stations located within 800 km epicentral distance for nine 7.2 < < 9.1 earthquakes. We consider the vertical component of seismogeodetic displacement as an approximate source time function and integrate the associated time series to obtain the seismic moment. By continuing to integrate vertical displacement beyond the initial P‐waves, we obtain rapid estimates of that are within 0.2 magnitude units for 8 thrust faulting events and within 0.3 units for the single normal faulting event. Because our estimates of the seismic moment are based on the maximum value of integrated displacement, no regression against other source parameters, or distance, is necessary. Our new method shows promise for integration into earthquake and local tsunami early warning systems, including tsunami earthquakes characterized by relatively slow moment release over a longer rupture time, and earthquakes with complex source time functions.

From Interseismic Deformation With Near‐Repeating Earthquakes to Co‐Seismic Rupture: A Unified View of the 2020 Mw6.8 Sivrice (Elazığ) Eastern Turkey Earthquake

AbstractThe East Anatolian Fault (EAF) is a left‐lateral transform fault accommodating the relative motion between the Anatolian and Arabian plates. On January 24, 2020, Mw6.8 Sivrice (Elazığ) earthquake is the largest event that occurred along the EAF since the nineteenth century. The earthquake provides a unique opportunity to capture a critical stage of the seismic cycle from the interseismic deformation to co‐seismic rupture. In this study, we examine the relationship between the interseismic fault activity and co‐seismic behavior of the earthquake. A kinematic model of the earthquake obtained from strong‐motion, GNSS and teleseismic waveforms along with static displacements from GNSS and InSAR data shows that the mainshock ruptured only 45 km of the 95 km long Sivrice‐Pütürge segment. Rupture initiated adjacent to the interseismically weakly coupled northeastern section and propagated unilaterally toward southwest with a rupture velocity of ∼2.5 km/s, stopping ∼30 km before the southwestern segment boundary. The earthquake did not generate any surface offsets. We identified 4 long‐term near‐repeating earthquake clusters beneath the highest co‐seismic slip zone adjacent to the northeastern creeping section. The mainshock was dynamically triggered by a M ∼ 5.4 foreshock located within the zone of near‐repeating earthquakes. We suggest that creep along the weakly coupled section of the fault loaded the neighboring locked section leading to the repeating earthquakes below the locked zone. The earthquake partially ruptured a fault segment characterized by high rate of diffuse seismicity, structural complexities and heterogeneous coupling, leading to a source characterized by a complex source time function with relatively slow rupture velocity.

Modulation of fault strength during the seismic cycle by grain-size evolution around contact junctions

Earthquake prediction is fundamentally hampered by insufficient knowledge of the constitutive behavior of faults. Laboratory and theoretical work by Dieterich (1978) and Ruina (1983) formed the basis of rate-and-state friction, a constitutive framework that captures the healing and weakening of frictional contacts. The rate-and-state law successfully explains the frictional behavior of a wide range of rocks and many phenomena associated with fault slip. The physics underlying the rate-and-state friction law is not fully understood, but it is generally accepted that several micro-mechanisms operate depending on the texture, maturity, fluid content, and other physical and kinematic conditions of the fault. Here, I formulate a constitutive framework that unifies a number of laboratory experiments on silicate rocks, and the gouges derived from these rocks, that describe the effect of grain-size, temperature, real area of contact, gouge thickness, and fault roughness. Fundamental to the model is that fault strength is controlled by the area of the interfacial contact junctions that support the shear and normal loads. The curvature of asperities weakly controls the area of their contact junction and the resulting fault strength. As a result, the dynamics of grain-size evolution around micro-asperities enables the seismic cycle. The model quantitatively explains the correlation between the characteristic weakening distance and the gouge thickness (Marone and Kilgore, 1993), the correlation of the static friction coefficient with the rate-dependence parameters (Ikari et al., 2011), the dependence of frictional resistance to sub-solidus temperatures (Chester, 1994), and predicts a range of grain sizes that is compatible with observations at exposed fault zones. However, the dynamics of grain-size evolution is still challenging to verify experimentally. The model features a regularization at vanishing slip speeds that implies distinct dynamics at low strength. For friction coefficients lower than 0.1, seismic cycle simulations predict the emergence of frictional instabilities with complex source time-functions. This is a significant departure from the predictions of rate-and-state friction that is most relevant to the dynamics of décollements, serpentinized shear zones, and other low-strength faults.

Rupture characteristics of a small-sized earthquake (MW 4.2), onshore the south Red Sea, Saudi Arabia

The present study is based on the use of Empirical Green's Function (EGF) deconvolution technique to retrieve the slip distribution of the 2014 Mw 4.2 Jizan earthquake, Saudi Arabia. Two datasets of complex Source Time Functions (STFs) were retrieved using two appropriate EGF events. We inverted the STF datasets to recover the slip distribution over both nodal planes, using the Bayesian modeling followed by a linear least-squares method. The inversion was performed assuming both planes as the fault plane and examined the goodness of fit for each nodal plane. Based on a series of finite-source inversions using different rupture velocities, we resolved the rupture velocity at 2.7–2.8 km/s and the fault plane of NNW trending; paralleling the Red Sea rift. Using the estimated rupture velocities and the preferred fault plane, we imaged quite similar slip models, exhibiting two slip patches located to the updip and downdip directions from the hypocentre. The spatiotemporal slip distributions revealed a complex rupture history of such small-sized earthquake is likely to that reported for large-sized earthquakes. A seismic moment of 2.8–3.2E+15 NM and a corresponding moment magnitude of 4.2-4-3 are inferred. The stress drops obtained from the slip distribution models were 2.2–2.5 MPa; indicating a typical value that characterized the plate-boundary earthquakes.

The Cephalonia, Ionian Sea (Greece), sequence of strong earthquakes of January-February 2014: a first report

On 26.1.2014 and 3.2.2014 two strong earthquakes of M<sub>w</sub>6.0 and M<sub>w</sub>5.9 ruptured the western Cephalonia Isl., Ionian Sea (Greece), at the SSW-wards continuation of the Lefkada segment of the Cephalonia Transform Fault Zone (CTFZ), causing considerable damage and a variety of ground failures. High-precision relocation of the aftershocks implies that the seismogenic layer was of 35 km in length (L) striking NNE-SSW, of 10 km maximum in width and 15 km in thickness. Two aftershock spatial clusters were revealed at north (<em>L<sub>1</sub></em>~10 km) and at south (<em>L<sub>2</sub></em>~25 km). However, no time correlation was found between the two clusters and the two strong earthquakes. Fitting the temporal evolution of aftershocks to the Omori-law showed slow aftershock decay. Fault plane solutions produced by moment tensor inversions indicated that the strong earthquakes as well as a plenty of aftershocks (M<sub>w</sub>≥4.0) were associated with dextral strikeslip faulting with some thrust component and preferred fault planes striking about NNE-SSW. Average fault plane parameters obtained for the three largest events are: strike 21(±2)<sup>0</sup>, dip 65.5(±3)<sup>0</sup>, slip 173(±3)<sup>0</sup>. Broadband P-wave teleseismic records were inverted for understanding the rupture histories. It was found that the earthquake of 26.1.2014 had a complex source time function with c. 62 cm maximum slip, source duration of ~12 s and downwards rupture. Most of the slip was concentrated on a 13x9 km fault rupture. The earthquake of 3.2.2014 had a relatively simple source time function related with one big patch of slip with maximum slip c. 45 cm, with 10 s source duration. The rupture was directed upwards which along with the shallow focus (~5 km) and the simple source time function may explain the significantly larger (0.77 g) PGA recorded with the second earthquake with respect to the one recorded (0.56 g) with the first earthquake. Most of the slip was concentrated on a 12x6 km fault rupture. Maximum seismic intensity (<em>I<sub>m</sub></em>) of level VII and VIII to VIII+ was felt in Lixouri town and the nearby villages from the first and the second earthquake, respectively. The rupture histories and the increased building vulnerability after the damage caused by the first shock may account for the larger <em>I<sub>m</sub></em> caused by the second shock. However, the ground failures area of the second earthquake was nearly half of that of the first earthquake, which is consistent with the faster attenuation of ground acceleration away from the meizoseismal area caused by the second earthquake with respect to the first one. From that the 2014 earthquakes ruptured on land western Cephalonia we suggested to revise the CTFZ geometry in the sense that the Lefkada CTFZ segment does not terminates offshore NW Cephalonia but extends towards SSW in western Cephalonia.

Mantle Attenuation Estimated from Regional and Teleseismic P-waves of Deep Earthquakes and Surface Explosions

We estimated the network-averaged mantle attenuation t*(total) of 0.5 s beneath the North Korea test site (NKTS) by use of P-wave spectra and normalized spectral stacks from the 25 May 2009 declared nuclear test (mb 4.5; IDC). This value was checked using P-waves from seven deep (580–600 km) earthquakes (4.8 < M w < 5.5) in the Jilin-Heilongjiang, China region that borders with Russia and North Korea. These earthquakes are 200–300 km from the NKTS, within 200 km of the Global Seismic Network seismic station in Mudanjiang, China (MDJ) and the International Monitoring System primary arrays at Ussuriysk, Russia (USRK) and Wonju, Republic of Korea (KSRS). With the deep earthquakes, we split the t*(total) ray path into two segments: a t*(u), that represents the attenuation of the up-going ray from the deep hypocenters to the local-regional receivers, and t*(d), that represents the attenuation along the down-going ray to teleseismic receivers. The sum of t*(u) and t*(d) should be equal to t*(total), because they both share coincident ray paths. We estimated the upper-mantle attenuation t*(u) of 0.1 s at stations MDJ, USRK, and KSRS from individual and stacks of normalized P-wave spectra. We then estimated the average lower-mantle attenuation t*(d) of 0.4 s using stacked teleseismic P-wave spectra. We finally estimated a network average t*(total) of 0.5 s from the stacked teleseismic P-wave spectra from the 2009 nuclear test, which confirms the equality with the sum of t*(u) and t*(d). We included constraints on seismic moment, depth, and radiation pattern by using results from a moment tensor analysis and corner frequencies from modeling of P-wave spectra recorded at local distances. We also avoided finite-faulting effects by excluding earthquakes with complex source time functions. We assumed ω2 source models for earthquakes and explosions. The mantle attenuation beneath the NKTS is clearly different when compared with the network-averaged t* of 0.75 s for the western US and is similar to values of approximately 0.5 s for the Semipalatinsk test site within the 0.5–2 Hz range.

Moment tensor inversion of waveforms: a two-step time-frequency approach

SUMMARY We present a moment tensor inversion of waveforms, which is more robust and yields more stable and more accurate results than standard approaches. The inversion is performed in two steps and combines inversions in time and frequency domains. First, the inversion for the source-time function is performed in the frequency domain using complex spectra. Second, the time-domain inversion for the moment tensor is performed using the source-time function calculated in the first step. In this way, we can consider a realistic, complex source-time function and still keep the final moment tensor inversion linear. Using numerical modelling, we compare the efficiency and accuracy of the proposed approach with standard waveform inversions. We study the sensitivity of the retrieved double-couple and non-double-couple components of the moment tensors to noise in the data, to inaccuracies of the location and of the velocity model, and to the type of the focal mechanism. Finally, the proposed moment tensor inversion is tested on real data observed in a complex 3-D inhomogeneous geological environment: a production blast and a rockburst in the Pyhasalmi ore mine, Finland.

A frequency domain inversion code to retrieve time-dependent parameters of very long period volcanic sources

In this article, we present a new code for the inversion of source parameters of volcanic seismic events. The code, named VOLPIS (Volcanic Long Period Inversion for the Source), has as its main goal the retrieval of time-dependent source parameters, so that both amplitudes and time history of source components can be estimated. The methodology used has been designed to be extremely flexible, in order to be easily applied to different datasets and, eventually, used for the inversion of any other seismic source with a complex source time function. Certain physical phenomena occurring at volcanic focal regions can be expressed in terms of a generalized seismic point source, as the sum of single forces and dipole forces (moment tensor). Source time histories for different source components can be kept independent, so that superposition of different physical phenomena can be retrieved, even in cases where they occur with different timing. Otherwise, a constraint for partially or fully common source time history can be included. The algorithm is based on a frequency domain approach, where the inversion is solved separately for each complex sampled frequency. The code structure helps in simplifying the handling and processing of data, which can be adapted to data characteristics or user requirements. Significant examples of possible applications to synthetic data are included.

Improved Green’s functions for passive-source structural studies

Over the past two decades, teleseismic receiver functions have proved to be a useful tool to investigate crustal structure. Because they represent a first-order approximation to the [Formula: see text]-wave component of the teleseismic-[Formula: see text] Green’s function, receiver functions provide valuable information on physical properties related to shear modulus. However, the implicit use of the [Formula: see text]-component seismogram as a proxy for the source precludes the recovery of information on discontinuous structure involving contrasts in compressional modulus. By deconvolving improved estimates of complex source time functions generated by earthquakes, one may move beyond the conventional receiver function paradigm to a more accurate approximation of the earth’s Green’s function. Using a new deconvolution method, we present estimates of the [Formula: see text]-component of the teleseismic-[Formula: see text] Green’s functions at several stations of the Canadian National Seismic Network (CNSN) that clearly show the receiver-side pure [Formula: see text]-wave crustal multiple. The identification and characterization of these signals in studies of the lithosphere will afford better constraints on subsurface lithology and represent a narrowing of the gap between active- and passive-source seismic imaging.

The Southern Region of Peru Earthquake of June 23rd, 2001

The western border of South America is one of the most important seismogenic regions in the world. In this region the most damaging earthquake ever recorded occurred. In June 23rd, 2001, another very strong earthquake (Mw = 8.1–8.2) occurred and produced death and damages in the whole southern region of Peru. This earthquake was originated by a friction process between Nazca and South American plates and affected an area of about 300 km × 120 km defined by the distribution of more than 220 aftershocks recorded by a local seismic network that operated 20 days. The epicenter of the main shock was localized in the northwestern extremity of the aftershock area, which suggests that the rupture propagated towards the SE direction. The modeling of P-wave for teleseismic distances permitted to define a focal mechanism of reverse type with NW-SE oriented nodal planes and a possible fault plane moving beneath almost horizontally in NE direction. The source time function (STF) suggests a complex process of rupture during 85 sec with 2 successive sources. The second one of greater size, and located approximately 100–120 km toward the SE direction was estimated to have a rupture velocity of about 2 km/sec on a 28°-dipping plane to the SE (N135°). A second event happened 45 sec after the first one with an epicenter 130km farther to the SE and a complex STF. This event and the second source of the main shock caused a Tsunami with waves from 7 to 8 meters that propagated almost orthogonally to the coast line, by affecting mainly the Camana area.

Normal Mode Composition of Earthquake T Phases

Understanding the nature of the coupling between the underwater acoustic field and the land seismic field is important for evaluating the performance of the T-phase stations in the International Monitoring System for the Comprehensive Nuclear-Test-Ban Treaty. For upslope propagation in an ocean environment, the places where underwater acoustic field energy couples into the land seismic field are determined to first approximation by the local water depth and the normal mode composition of the acoustic energy. Therefore, the use of earthquake-generated Tphases as natural probes of water-to-land coupling characteristics is aided by knowledge of their modal composition. Data collected by a 200-element, 3000-m-aperture vertical hydrophone array during a 1989 experiment in the deep northeast Pacific Ocean are used to determine the mode composition of T-phase arrivals from two mb4.1 earthquakes near the west coast of the U.S., one occurring offshore and the other on land. Results from an eigenanalysis approach and conventional mode decomposition for the two events are consistent and show that at 5 Hz, the offshore event’s arrivals have higher-order mode content compared to those from the event on land. Single hydrophone recordings at Pt. Sur of two mb 4.4 Hawaiian events in 1996 and 1997, one occurring offshore and the second on land, display time-frequency arrival structures that are explainable by the dispersion characteristics over the oceanic path. Although other effects due to complex source time functions and shear wave and dispersive propagation effects along the initial land path cannot be separated with these single element data, differences in these two events’ arrival structures suggest differences in normal mode content consistent with those seen in the pair of 1989 events. Ocean-path dispersion also appears to play a significant role in determining the in-water arrival structure from a 1995 French nuclear test at Mururoa. Recordings of two Hawaiian events in 1997 by the T-phase station VIB and the seismic station at Berkeley illustrate that the water-land coupling confuses the relative timing between normal modes, resulting in apparent loss of information about the source.

Source mechanism of earthquakes in Peru

The source mechanism of 19 earthquakes that occurred in Peru (1990–1996) is studied using broad band data. Focal mechanisms are obtained using polarities of P wave and body wave form inversion. Shallow earthquakes show complex source time functions, intermediate and deep depth shocks have simpler ones. Stress distributions have been obtained from focal mechanisms estimated in this study and previous studies. Shallow earthquakes show reverse faulting with an ENE-WSW to E-W oriented pressure axes. Intermediate depth shocks indicate horizontal extension on E-W direction, normal to the Peru-Chilean trench. Earthquakes with foci at very deep depth show horizontal extension in the E-W direction in Peru-Brazil and N-S in Peru-Bolivia borders. This difference in stress orientation may indicate a different origin for deep activity at each region.

Source characteristics of historic earthquakes along the central Chile subduction Askew et alzone

We have analyzed four large to great historic earthquakes that occurred along the central Chile subduction zone from north to south on November 11, 1922 (M s=8.3), April 6, 1943 (M s=7.9), December 1, 1928 (M s=8.0) and January 25, 1939 (M s=7.8). Waveform modeling and P-wave first motions indicate that the 1922, 1928 and 1943 earthquakes are shallow and consistent with underthrusting of the Nazca Plate beneath the South American plate. In contrast, the 1939 earthquake is not an underthrusting event but rather a normal fault event within the down-going slab. The 1922 earthquake is by far the largest event with a complex source time function showing three pulses of moment release and a duration of 75 s. The 1943 earthquake has a simple source time function with one pulse of moment release and a duration of 24 s. This event had a local tsunami of 4 m and a far-field tsunami height in Japan of 10–30 cm. The 1928 earthquake also has a simple source time function with a duration of 28 s. The aftershocks and highest intensities are south of the epicenter indicating a southward rupture with most of the seismic moment release occurring 50–80 km south of the 1928 epicenter but still north of the adjacent 1939 earthquake region. The 1939 Chillan earthquake was not an underthrusting but rather a complex normal fault earthquake. Our preferred model is a normal fault mechanism at a depth of 80 to 100 km with two pulses of moment release and a total duration of approximately 60 s. The high intensities, lack of a tsunami, and inland location associated with the 1939 event are all consistent with an intraplate event within the down-going slab. The 1939 earthquake was clearly more destructive than the other similar size or larger events. This may in part be due to the intraplate nature of the event but also due to high amplification of the sites in the Central Valley of south central Chile.

Source mechanism of intermediate and deep earthquakes in southern Spain

Focal mechanisms of 10 intermediate-depth earthquakes (30<h<150 km) and one very deep (h 650 km) which occurred in southern Spain and the Alboran Sea are studied in this paper. Distribution of epicenters with foci at intermediate depth shows a N–S alignment with a geometry parallel to the east of the Arc of Gibraltar. Focal mechanisms have been determined from first motion of P-wave and modeling wave forms of direct P arrivals. Most of the intermediate depth events present single source time functions with small time duration (smaller than 0.3 s) and only for the event of March 27, 1987 a complex source time function has been found. The very deep earthquake of March 8, 1990 has a complex focal mechanism with a long source time function (1.2 s) and two different fault-plane orientations. Scalar seismic moments and dimensions have been obtained from modeling and spectral analysis. The results are interpreted in terms of the seismotectonic framework of the region and suggest the existence of a vertical slab extending from 50 km to 150 km with strike N–S produced by a lithospheric delamination process. The existence of the very deep activity at 650 km of depth may be explained in terms of a block of lithospheric material that still cold and rigid, generates the very deep earthquakes.

Rupture characteristics of the 1982 Tonga and 1986 Kermadec earthquakes

Source time functions are obtained for the December 19, 1982, Tonga and October 20, 1986, Kermadec earthquakes from deconvolution of P waves, using a model including a significant dip of the ocean bottom. In such a geometry, the amplitudes of subsequent water multiples are large and irregular. As a result, we can model these two events as consisting of a single pulse of short duration (23 s in the case of the 1982 event, and 19 s for the 1986 one), despite complex teleseismic waveshapes. This is in contrast to inversions using flat‐layered models, which required very complex source time functions, which were mutually inconsistent at different stations. The anomalously large tsunami generated by the 1982 event is modeled as the result of propagation of the rupture into the sedimentary wedge at the trench, rather than due to a long‐lived complex source time function, the latter being also incompatible with the total lack of excitation of the Earth's normal modes in the low‐frequency limit. For the 1986 event, we derive a source duration (19 s) very short in relation to its moment (8×1027 dyn‐cm). This earthquake, a high stress drop rupture featuring a unique combination of strike‐slip and thrust motion on a plane dipping oceanwards of the trench, is interpreted as expressing internal deformation of the subducting Pacific slab at a depth of 40 km, under the compressional stresses resulting from the collision.

Numerical simulations of earthquake source process by a three-dimensional crack model. Part II. Seismic waves and spectrum.

We study the seismic waves and spectrum generated by a dynamical rupture process by a three-dimensional crack model obtained in Part I (MIYATAKE, 1980). The fracture condition used here is a finite stress fracture criterion, while the strength is assumed to be distributed irregularly. It is found that complex seismic waves are generated by the irregular rupture process rather than by the complex source-time functions. Near-field and far-field seismic wave radiated from the fault with the above distribution are also calculated. Seismic waves from the fault with a line barrier and a block barrier have almost similar wave shapes. Random barrier distribution radiates high frequency seismic waves, and the spectral amplitudes tend to decay inversely in proportion to the frequency in the range beyond the corner frequency.