Waveform inversion for microseismic velocity analysis and event location in VTI media

Geophysics ◽  
2017 ◽  
Vol 82 (4) ◽  
pp. WA95-WA103 ◽  
Author(s):  
Oscar Jarillo Michel ◽  
Ilya Tsvankin
Keyword(s):  
Moment Tensor ◽  
Source Parameters ◽  
Waveform Inversion ◽  
Transversely Isotropic ◽  
Velocity Analysis ◽  
Origin Time ◽  
Velocity Models ◽  
Trade Offs ◽  
Receiver Array ◽  
Event Location

Waveform inversion (WI), which has been extensively used in reflection seismology, could provide improved velocity models and event locations for microseismic surveys. Here, we develop an elastic WI algorithm for anisotropic media designed to estimate the 2D velocity field along with the source parameters (location, origin time, and moment tensor) from microseismic data. The gradient of the objective function is obtained with the adjoint-state method, which requires just two modeling simulations at each iteration. In the current implementation the source coordinates and velocity parameters are estimated sequentially at each stage of the inversion to minimize trade-offs and improve the convergence. Synthetic examples illustrate the accuracy of the inversion for layered VTI (transversely isotropic with a vertical symmetry axis) media, as well as the sensitivity of the velocity-analysis results to noise, the length of the receiver array, errors in the initial model, and variability in the moment tensor of the recorded events.

scholarly journals The waveform inversion of mainshock and aftershock data of the 2006 M6.3 Yogyakarta earthquake

2021 ◽  
Vol 8 (1) ◽  
Author(s):  
Hijrah Saputra ◽  
Wahyudi Wahyudi ◽  
Iman Suardi ◽  
Ade Anggraini ◽  
Wiwit Suryanto
Keyword(s):  
Joint Inversion ◽  
Body Wave ◽  
Moment Tensor ◽  
Near Field ◽  
Source Parameters ◽  
Waveform Inversion ◽  
Moment Tensor Inversion ◽  
Slip Distribution ◽  
Aftershock Distribution ◽  
Velocity Models

AbstractThis study comprehensively investigates the source mechanisms associated with the mainshock and aftershocks of the Mw = 6.3 Yogyakarta earthquake which occurred on May 27, 2006. The process involved using moment tensor inversion to determine the fault plane parameters and joint inversion which were further applied to understand the spatial and temporal slip distributions during the earthquake. Moreover, coseismal slip distribution was overlaid with the relocated aftershock distribution to determine the stress field variations around the tectonic area. Meanwhile, the moment tensor inversion made use of near-field data and its Green’s function was calculated using the extended reflectivity method while the joint inversion used near-field and teleseismic body wave data which were computed using the Kikuchi and Kanamori methods. These data were filtered through a trial-and-error method using a bandpass filter with frequency pairs and velocity models from several previous studies. Furthermore, the Akaike Bayesian Information Criterion (ABIC) method was applied to obtain more stable inversion results and different fault types were discovered. Strike–slip and dip-normal were recorded for the mainshock and similar types were recorded for the 8th aftershock while the 9th and 16th June were strike slips. However, the fault slip distribution from the joint inversion showed two asperities. The maximum slip was 0.78 m with the first asperity observed at 10 km south/north of the mainshock hypocenter. The source parameters discovered include total seismic moment M0 = 0.4311E + 19 (Nm) or Mw = 6.4 with a depth of 12 km and a duration of 28 s. The slip distribution overlaid with the aftershock distribution showed the tendency of the aftershock to occur around the asperities zone while a normal oblique focus mechanism was found using the joint inversion.

Location and source parameters of the 19 June 1994 (MW = 5.0) offshore Petrolia, California, earthquake

1997 ◽  
Vol 87 (1) ◽  
pp. 272-276
Author(s):  
Jochen Braunmiller ◽  
Beate Leitner ◽  
John Nábělek ◽  
Anne M. Tréhu
Keyword(s):  
Moment Tensor ◽  
Source Parameters ◽  
Low Frequency ◽  
P Wave ◽  
Ocean Bottom ◽  
Time Data ◽  
Origin Time ◽  
Depth Estimate ◽  
Moment Tensor Solution ◽  
Velocity Models

Abstract The MW = 5.0, 19 June 1994 offshore Petrolia, California, earthquake was well recorded by nine ocean-bottom hydrophones (OBH) and seismometers (OBS), providing an opportunity to precisely locate an earthquake in the tectonically active Mendocino triple junction region. Adding the offshore data improves the azimuthal station coverage and essentially removes the epicenter's sensitivity to the choice of inversion parameters and velocity models. The hypocentral parameters, assuming an oceanic upper-mantle velocity of 7.9 km/sec, are 10:39:33.2 UTC for origin time, 40.376° N latitude, 124.441° W longitude, and a depth of 18.8 km. The moment-tensor solution obtained by modeling of low-frequency regional waveforms indicates predominantly strike-slip faulting with a north-south-trending P axis, as is typical for Gorda plate earthquakes, and confirms the depth estimate from the P-wave travel-time data.

scholarly journals Passive seismic event estimation using multiscattering waveform inversion

Geophysics ◽  
2019 ◽  
Vol 84 (3) ◽  
pp. KS59-KS69 ◽  
Author(s):  
Chao Song ◽  
Zedong Wu ◽  
Tariq Alkhalifah
Keyword(s):  
Seismic Data ◽  
Waveform Inversion ◽  
Velocity Model ◽  
Source Image ◽  
Time Inversion ◽  
Multiple Sources ◽  
Origin Time ◽  
Secondary Sources ◽  
Velocity Models ◽  
Passive Seismic

Passive seismic monitoring has become an effective method to understand underground processes. Time-reversal-based methods are often used to locate passive seismic events directly. However, these kinds of methods are strongly dependent on the accuracy of the velocity model. Full-waveform inversion (FWI) has been used on passive seismic data to invert the velocity model and source image, simultaneously. However, waveform inversion of passive seismic data uses mainly the transmission energy, which results in poor illumination and low resolution. We developed a waveform inversion using multiscattered energy for passive seismic to extract more information from the data than conventional FWI. Using transmission wavepath information from single- and double-scattering, computed from a predicted scatterer field acting as secondary sources, our method provides better illumination of the velocity model than conventional FWI. Using a new objective function, we optimized the source image and velocity model, including multiscattered energy, simultaneously. Because we conducted our method in the frequency domain with a complex source function including spatial and wavelet information, we mitigate the uncertainties of the source wavelet and source origin time. Inversion results from the Marmousi model indicate that by taking advantage of multiscattered energy and starting from a reasonably acceptable frequency (a single source at 3 Hz and multiple sources at 5 Hz), our method yields better inverted velocity models and source images compared with conventional FWI.

A user-friendly probabilistic earthquake source inversion framework for joint inversion of seismic, geodetic, and gravitational signals - The Grond toolkit

2020 ◽  
Author(s):  
Sebastian Heimann ◽  
Marius Isken ◽  
Daniela Kühn ◽  
Hannes Vasyura-Bathke ◽  
Henriette Sudhaus ◽  
...  
Keyword(s):  
Open Source ◽  
Joint Inversion ◽  
Moment Tensor ◽  
Source Parameters ◽  
Seismic Source ◽  
Magma Ascent ◽  
Caldera Collapse ◽  
Waveform Data ◽  
Trade Offs ◽  
Finite Fault

<p>Seismic source and moment tensor waveform inversion is often ill-posed or non-unique if station coverage is poor or signals are weak. Three key ingredients can help in these situations: (1) probabilistic inference and global search of the full model space, (2) joint optimisation with datasets yielding complementary information, and (3) robust source parameterisation or additional source constraints. These demands lead to vast technical challenges, on the performance of forward modelling, on the optimisation algorithms, as well as on visualisation, optimisation configuration, and management of the datasets. Implementing a high amount of automation is inevitable.</p><p>To tackle all these challenges, we are developing a sophisticated new seismic source optimisation framework, Grond. With its innovative Bayesian bootstrap optimiser, it is able to efficiently explore large model spaces, the trade-offs and the uncertainties of source parameters. The program is highly flexible with respect to the adaption to specific source problems, the design of objective functions, and the diversity of empirical datasets.</p><p>It uses an integrated, robust waveform data processing, and allows for interactive visual inspection of many aspects of the optimisation problem, including visualisation of the result uncertainties. Grond has been applied to CMT moment tensor and finite-fault optimisations at all scales, to nuclear explosions, to a meteorite atmospheric explosion, and to volcano-tectonic processes during caldera collapse and magma ascent. Hundreds of seismic events can be handled in parallel given a single optimisation setup.</p><p>Grond can be used to optimise simultaneously seismic waveforms, amplitude spectra, waveform features, phase picks, static displacements from InSAR and GNSS, and gravitational signals.</p><p>Grond is developed as an open-source package and community effort. It builds on and integrates with other established open-source packages, like Kite (for InSAR) and Pyrocko (for seismology).</p>

Determination of anisotropic velocity models from walkaway VSP data acquired in the presence of dip

Geophysics ◽  
1997 ◽  
Vol 62 (3) ◽  
pp. 723-729 ◽  
Author(s):  
Colin M. Sayers
Keyword(s):  
Transversely Isotropic ◽  
Seismic Profile ◽  
Slowness Surface ◽  
Velocity Models ◽  
Receiver Array ◽  
Strike Direction ◽  
Vsp Data ◽  
Walkaway Vsp ◽  
Oriented Parallel

Wide‐aperture walkaway vertical seismic profile (VSP) data acquired through transversely isotropic horizontal layers can be used to determine the P phase‐slowness surface, local to a receiver array in a borehole. In the presence of dip, errors in the slowness surface may occur if the medium is assumed to be layered horizontally. If the acquisition plane is oriented parallel to the dip direction, the derived slowness is too large for sources offset from the well in the down‐dip direction and too small for sources offset from the well in the up‐dip direction. For acquisition parallel to the strike of the layers, the recovery of the P phase‐slowness in the vicinity of the receiver array is excellent. It is therefore preferable to orient the walkaway VSP in the strike direction to estimate the anisotropic parameters of the medium in the vicinity of a receiver array. However, this may not be possible. If the dip direction of all layers has the same azimuth, the variation of walkaway traveltimes with azimuth has a simple form. This allows data from a single walkaway VSP extending both sides of a well to be inverted for the local anisotropic P phase‐slowness surface at the receivers even in the presence of dip. If data are acquired at more than one azimuth, the dip direction can be determined.

Depth‐domain velocity analysis in VTI media using surface P-wave data: Is it feasible?

Geophysics ◽  
2001 ◽  
Vol 66 (3) ◽  
pp. 897-903 ◽  
Author(s):  
Yves Le Stunff ◽  
Vladimir Grechka ◽  
Ilya Tsvankin
Keyword(s):  
Vertical Velocity ◽  
Heterogeneous Media ◽  
Wave Reflection ◽  
P Wave ◽  
Model Parameters ◽  
Velocity Analysis ◽  
Depth Migration ◽  
Velocity Models ◽  
Trade Offs ◽  
Vti Media

The main difficulties in anisotropic velocity analysis and inversion using surface seismic data are associated with the multiparameter nature of the problem and inherent trade‐offs between the model parameters. For the most common anisotropic model, transverse isotropy with a vertical symmetry axis (VTI media), P-wave kinematic signatures are controlled by the vertical velocity V0 and the anisotropic parameters ε and δ. However, only two combinations of these parameters—NMO velocity from a horizontal reflector Vnmo(0) and the anellipticity coefficient η—can be determined from P-wave reflection traveltimes if the medium above the reflector is laterally homogeneous. While Vnmo(0) and η are sufficient for time‐domain imaging in VTI media, they cannot be used to resolve the vertical velocity and build velocity models needed for depth migration. Here, we demonstrate that P-wave reflection data can be inverted for all three relevant VTI parameters (V0, ε and δ) if the model contains nonhorizontal intermediate interfaces. Using anisotropic reflection tomography, we carry out parameter estimation for a two‐layer medium with a curved intermediate interface and reconstruct the correct anisotropic depth model. To explain the success of this inversion procedure, we present an analytic study of reflection traveltimes for this model and show that the information about the vertical velocity and reflector depth was contained in the reflected rays which crossed the dipping intermediate interface. The results of this work are especially encouraging because the need for depth imaging (such as prestack depth migration) arises mostly in laterally heterogeneous media. Still, we restricted this study to a relatively simple model and constrained the inversion by assuming that one of the layers is isotropic. In general, although lateral heterogeneity does create a dependence of P-wave reflection traveltimes on the vertical velocity, there is no guarantee that for more complicated models all anisotropic parameters can be resolved in a unique fashion.

Generation and processing of pseudo‐shear‐wave data: Theory and case study

Geophysics ◽  
2003 ◽  
Vol 68 (6) ◽  
pp. 1807-1816 ◽  
Author(s):  
Vladimir Grechka ◽  
Pawan Dewangan
Keyword(s):  
Shear Wave ◽  
Pure Shear ◽  
Transversely Isotropic ◽  
Velocity Model ◽  
Velocity Analysis ◽  
Ocean Bottom ◽  
S Wave ◽  
Velocity Models ◽  
Conversion Point

Processing of converted (PS) waves currently adopted by the exploration industry is essentially based on resorting the PS data into common‐conversion‐point gathers and using them for velocity analysis. Here, we explore an alternative procedure. Our key idea is to generate the so‐called pseudo‐shear (ΨS) seismograms from the recorded PP and PS traces and run conventional velocity analysis on the reconstructed ΨS data. This results in an effective S‐wave velocity model because our method creates data that possess kinematics of pure shear‐wave primaries. We never deal with such complexities of converted waves as moveout asymmetry, reflection point dispersal, and polarity reversal; therefore, these generally troublesome features become irrelevant. We describe the details of our methodology and examine its behavior both analytically and numerically. We apply the developed processing flow to a four‐component ocean‐bottom cable line acquired in the Gulf of Mexico. Since the obtained stacking velocities of P‐ and ΨS‐waves indicate the presence of effective anisotropy, we proceed with estimating a family of kinematically equivalent vertical transversely isotropic (VTI) velocity models of the subsurface.

scholarly journals The waveform inversion of mainshock and aftershock data of the 2006 M6.3 Yogyakarta earthquake

2021 ◽  
Author(s):  
Hijrah Saputra ◽  
Wahyudi Wahyudi ◽  
Iman Suardi ◽  
Ade Anggraini ◽  
Wiwit Suryanto
Keyword(s):  
Joint Inversion ◽  
Body Wave ◽  
Moment Tensor ◽  
Near Field ◽  
Source Parameters ◽  
Moment Tensor Inversion ◽  
Slip Distribution ◽  
Coseismic Slip ◽  
Aftershock Distribution ◽  
Velocity Models

Abstract This study comprehensively investigates the source mechanisms associated with the mainshock and aftershocks of the Yogyakarta earthquake of magnitude Mw = 6.3 on May 27, 2006. Therefore, this study is to provide a more precise answer to the controversial source mechanism. This study uses moment tensor inversion to obtain fault plane parameters and joint inversion to obtain spatial and temporal slip distributions during an earthquake. The coseismic slip distribution is overlaid with the relocated aftershock distribution to see the stress field variations around the tectonic area of the study. Moment tensor inversion uses near-field data, and joint inversion uses near-field and teleseismic body wave data. The data is filtered by trial and error using a bandpass filter with frequency pairs and velocity models from several previous studies. The green's function for moment tensor inversion calculated using the extended reflectivity method and joint inversion computed using the Kikuchi and Kanamori methods. In this study, we apply the Akaike Bayesian Information Criterion (ABIC) method to obtain more stable inversion results. The results of the mainshock and aftershock moment tensor inversion show different fault types. The mainshock fault types are strike-slip and dip-normal types, while the 8th aftershock is of the same type as the mainshock, while the 9th and 16th June are strike slips. The joint inversion results show two asperities. The maximum slip is 0.78 m, with the first asperity 10 km south of the mainshock and the second asperity 10 km north of the mainshock. The obtained source parameters are total seismic moment M0 = 0.4311E + 19 (Nm) or Mw = 6.4, with a source depth of 12 km and a source duration of 28 seconds. Slip distribution overlay with aftershock distribution shows compatibility. The type of focus mechanism that results from this joint inversion is the oblique.

Sign in / Sign up

Export Citation Format

Share Document