scholarly journals Accounting for theory errors with empirical Bayesian noise models in nonlinear centroid moment tensor estimation

2021 ◽  
Author(s):  
H Vasyura-Bathke ◽  
J Dettmer ◽  
R Dutta ◽  
P M Mai ◽  
S Jónsson
Keyword(s):  
Parameter Estimation ◽  
Uncertainty Quantification ◽  
Measurement Errors ◽  
Velocity Structure ◽  
Moment Tensor ◽  
Velocity Model ◽  
Model Uncertainties ◽  
Centroid Moment Tensor ◽  
Uncertainty Estimates ◽  
Noise Models

Summary Centroid moment-tensor (CMT) parameters can be estimated from seismic waveforms. Since these data indirectly observe the deformation process, CMTs are inferred as solutions to inverse problems which are generally under-determined and require significant assumptions, including assumptions about data noise. Broadly speaking, we consider noise to include both theory and measurement errors, where theory errors are due to assumptions in the inverse problem and measurement errors are caused by the measurement process. While data errors are routinely included in parameter estimation for full CMTs, less attention has been paid to theory errors related to velocity-model uncertainties and how these affect the resulting moment-tensor (MT) uncertainties. Therefore, rigorous uncertainty quantification for CMTs may require theory-error estimation which becomes a problem of specifying noise models. Various noise models have been proposed, and these rely on several assumptions. All approaches quantify theory errors by estimating the covariance matrix of data residuals. However, this estimation can be based on explicit modelling, empirical estimation, and/or ignore or include covariances. We quantitatively compare several approaches by presenting parameter and uncertainty estimates in non-linear full CMT estimation for several simulated data sets and regional field data of the Ml 4.4, 13 June 2015 Fox Creek, Canada, event. While our main focus is at regional distances, the tested approaches are general and implemented for arbitrary source model choice. These include known or unknown centroid locations, full MTs, deviatoric MTs, and double-couple MTs. We demonstrate that velocity-model uncertainties can profoundly affect parameter estimation and that their inclusion leads to more realistic parameter uncertainty quantification. However, not all approaches perform equally well. Including theory errors by estimating non-stationary (non-Toeplitz) error covariance matrices via iterative schemes during Monte Carlo sampling performs best and is computationally most efficient. In general, including velocity-model uncertainties is most important in cases where velocity structure is poorly known.

Source Parameter Sensitivity of Earthquake Simulations assisted by Machine Learning 

2021 ◽  
Author(s):  
Marisol Monterrubio-Velasco ◽  
J. Carlos Carrasco-Jimenez ◽  
Otilio Rojas ◽  
Juan E. Rodriguez ◽  
David Modesto ◽  
...  
Keyword(s):  
Machine Learning ◽  
High Performance ◽  
Moment Tensor ◽  
Source Parameters ◽  
Velocity Model ◽  
Source Parameter ◽  
Earthquake Simulation ◽  
Centroid Moment Tensor ◽  
Ground Shaking ◽  
Earthquake Simulations

<p>After large magnitude earthquakes have been recorded, a crucial task for hazard assessment is to quickly estimate Ground Shaking (GS) intensities at the affected region. Urgent physics-based earthquake simulations using High-Performance Computing (HPC) facilities may allow fast GS intensity analyses but are very sensitive to source parameter values. When using fast estimates of source parameters such as magnitude, location, fault dimensions, and/or Centroid Moment Tensor (CMT), simulations are prone to errors in their computed GS. Although the approaches to estimate earthquake location and magnitude are consolidated, depth location estimates are largely uncertain. Moreover, automatic CMT solutions are not always provided by seismological agencies, or such solutions are available at later times after waveform inversions allow the determination of moment tensor components. The uncertainty on these parameters, especially a few minutes after the earthquake has been registered, strongly affects GS maps resulting from simulations.</p><p>In this work, we present a workflow prototype to produce an uncertainty quantification method as a function of the source parameters. The core of this workflow is based on Machine Learning (ML) techniques. As a study case, we consider a domain of 110x80 km centered in 63.9ºN-20.6ºW in Southern Iceland, where the 17 best-mapped faults have hosted the historical events of the largest magnitude. We generate synthetic GS intensity maps using the AWP-ODC finite-difference code for earthquake simulation and a one-dimensional velocity model, with 40 recording surface stations. By varying a few source parameters (e.g. event magnitude, CMT, and hypocenter location), we finally model tens of thousands of hypothetical earthquakes. Our ML analog will then be able to relate GS intensity maps to source parameters, thus simplifying sensitivity studies.</p><p>Additionally, the results of this workflow prototype will allow us to obtain ML-based intensity maps a few seconds after an earthquake occurs exploiting the predictive power of ML techniques. We will evaluate the accuracy of these maps as standalone complements to GMPEs and simulations.</p>

scholarly journals Centroid moment tensor inversions of offshore earthquakes using a three-dimensional velocity structure model: slip distributions on the plate boundary along the Nankai Trough

2020 ◽  
Vol 222 (2) ◽  
pp. 1109-1125 ◽  
Author(s):  
Shunsuke Takemura ◽  
Ryo Okuwaki ◽  
Tatsuya Kubota ◽  
Katsuhiko Shiomi ◽  
Takeshi Kimura ◽  
...  
Keyword(s):  
Wave Propagation ◽  
Seismic Wave ◽  
Velocity Structure ◽  
Nankai Trough ◽  
Moment Tensor ◽  
Plate Boundary ◽  
Seismic Wave Propagation ◽  
Structure Model ◽  
Centroid Moment Tensor ◽  
Heterogeneous Distribution

SUMMARY Due to complex 3-D heterogeneous structures, conventional 1-D analysis techniques using onshore seismograms can yield incorrect estimation of earthquake source parameters, especially dip angles and centroid depths of offshore earthquakes. Combining long-term onshore seismic observations and numerical simulations of seismic wave propagation in a 3-D model, we conducted centroid moment tensor (CMT) inversions of earthquakes along the Nankai Trough between April 2004 and August 2019 to evaluate decade-scale seismicity. Green's functions for CMT inversions of earthquakes with moment magnitudes of 4.3–6.5 were evaluated using finite-difference method simulations of seismic wave propagation in the regional 3-D velocity structure model. Significant differences of focal mechanisms and centroid depths between previous 1-D and our 3-D catalogues were found in the solutions of offshore earthquakes. By introducing the 3-D structures of the low-velocity accretionary prism and the Philippine Sea Plate, dip angles and centroid depths for offshore earthquakes were well-constrained. Teleseismic CMT also provides robust solutions, but our regional 3-D CMT could provide better constraints of dip angles. Our 3-D CMT catalogue and published slow earthquake catalogues depicted spatial distributions of slip behaviours on the plate boundary along the Nankai Trough. The regular and slow interplate earthquakes were separately distributed, with these distributions reflecting the heterogeneous distribution of effective strengths along the Nankai Trough plate boundary. By comparing the spatial distribution of seismic slip on the plate boundary with the slip-deficit rate distribution, regions with strong coupling were clearly identified.

Earthquake-based Full-Waveform Inversion at the Exploration Scale from Dense Broadband Array Data

2020 ◽  
Author(s):  
Qiancheng Liu ◽  
Frederik Simons ◽  
Fuchun Gao ◽  
Paul Williamson ◽  
Jeroen Tromp
Keyword(s):  
Moment Tensor ◽  
Regional Scale ◽  
Waveform Inversion ◽  
Three Dimensional ◽  
Velocity Model ◽  
Hydrocarbon Exploration ◽  
Full Waveform Inversion ◽  
Centroid Moment Tensor ◽  
Shear Wave Speed ◽  
Full Waveform

<p>In challenging environments with natural seismicity and where active source acquisition is expensive and dangerous, the question arises whether naturally occurring earthquakes offer useful information for hydrocarbon exploration. Here, we report on an experiment that installed 252/247 receivers to acquire data for two periods of 7 months, in the presence of significant rugged topography (elevations from 500 m to 3500 m). The station density is about 1 per 25 km^2 (compared to, e.g., USArray, where the average station spacing was 70 km). Data were recorded in a frequency band from 0.2 Hz to 50 Hz. Several thousand seismic events originating within the array bounds were identified in these data. A compressional-speed tomographic velocity model was derived using first-arriving phases. Centroid moment tensor (CMT) solutions have been obtained for about 4% of the identified events using Green’s function-based multicomponent waveform inversion, assuming a layered velocity model. We are now working to improve that model by performing elastic full-waveform inversion for three-dimensional compressional and shear-wave speed perturbations, honoring the topography, after a prior full-wavefield-based reassessment of the earthquake source mechanisms. We are also aiming to increase the number of events considered in the inversion while weighting the data based on estimates of data quality. This is assessed with a flexible automated procedure that considers a variety of data attributes over a range of frequencies. We run simulations using the spectral-element package SPECFEM3D on a cluster that employs 4 GPU cards per simulation. We identify the promising areas of good initial fit from the highest-quality seismic traces and gradually bring the predictions in line with the observations via LBFGS model optimization. We review the results of our work so far, discuss how to continue to bring best practices from global seismology down to the regional scale, and consider the implications for using such passive experiments to complement or replace active exploration in such challenging zones.</p>

Comparison of magnitudes estimated by the Japan Meteorological Agency with moment magnitudes for intermediate and deep earthquakes

1996 ◽  
Vol 86 (3) ◽  
pp. 832-842 ◽  
Author(s):  
Akio Katsumata
Keyword(s):  
Velocity Structure ◽  
Epicentral Distance ◽  
Moment Tensor ◽  
Focal Depth ◽  
Moment Magnitude ◽  
Japan Meteorological Agency ◽  
Maximum Displacement ◽  
Centroid Moment Tensor ◽  
Attenuation Function ◽  
Shallow Earthquakes

Abstract The magnitude, MJMA, estimated by the Japan Meteorological Agency (JMA) is generally referred to in Japan for the regional seismicity in the area. MJMA is determined from maximum displacement amplitudes of the total seismic wave traces. For earthquakes shallower than 60 km, MJMA is determined by Tsuboi's formula, and for earthquakes deeper than 60 km, by Katsumata's formula. These relations were designed to give almost the same magnitude value as that of Gutenberg and Richter. We compared MJMA with moment magnitude, MW, which can be calculated from the centroid moment tensor (CMT) solutions. It was found that the average difference between MJMA and MW is not significant for shallow earthquakes in the magnitude range from 5 to 7, but it is significant at a low level for the earthquakes of deeper foci. The averaged difference reaches about 0.4 magnitude units for the focal depth of 600 km. We derived an attenuation function for the maximum displacement amplitude assuming the validity of the moment magnitude. This relation between epicentral distance and amplitude for shallow earthquakes is almost identical to the one calculated from Tsuboi's formula. It is suggested that the estimated attenuation function for deep-focus earthquakes reflects the specific Q and velocity structure that is peculiar to the subduction zone.

Sign in / Sign up

Export Citation Format

Share Document