Search

We monitored the time history of the velocity change (dv/v) from 2002 to 2022 to investigate temporal changes in the physical state near the Parkfield Region of the San Andreas Fault throughout the interseismic period. Following the coseismic decrease in dv/v caused by the 2003 San Simeon (SS) and the 2004 Parkfield earthquakes, the dv/v heals logarithmically and shows a net long‐term increase in which the current dv/v level is equivalent to, or exceeding, the value before the 2003 SS earthquake. We investigated this long‐term trend by fitting the model accounting for the environmental and coseismic effects to the channel‐weighted dv/v time series. We confirmed with the metrics of Akaike information criterion and Bayesian information criterion that the additional term of either a linear trend term, or a residual healing term for the case where the healing had not been completed before the SS earthquake occurred, robustly improved the fit to the data. We eventually evaluated the sensitivity of the dv/v time history to the GNSS‐derived strain field around the fault. The cumulative dilatational strain spatially averaged around the seismic stations shows a slight extension, which is opposite to what would be expected for an increase in dv/v. However, the cumulative rotated axial strain shows compression in a range near the maximum contractional horizontal strain (azimuth of N35°W to N45°E), suggesting that the closing of pre‐existing microcracks aligned perpendicular to the axial contractional strains would be a candidate to cause the long‐term increase observed in the multiple station pairs. Plain Language Summary: We monitored the temporal change of velocity (dv/v) around the Parkfield Region from 2002 to 2022 to investigate the healing of dv/v after the 2003 San Simeon (SS) and the 2004 Parkfield earthquakes. Following those events, the dv/v recovers logarithmically with time and shows a net long‐term increase in which the current dv/v level is equivalent to, or exceeding, the value before the 2003 SS earthquake. We investigated this long‐term trend by fitting the model accounting for the environmental factors, the coseismic effects, and the additional term explaining the long‐term increase. Models with the additional term provide a statistically robust improved fit to the data, suggesting the long‐term increase observed in the dv/v time history is a non‐negligible factor. We eventually evaluated the sensitivity of the dv/v to the GNSS‐derived strain field around the fault. The spatially averaged dilatational strain could not explain the increase in dv/v as it shows a slight extension, whereas the rotated axial strain shows compression in a range near the maximum contractional horizontal strain. The closing of pre‐existing microcracks aligned perpendicular to the contractional strains thus would be a candidate to cause the increase in the material's rigidity and the average seismic wave velocity. Key Points: Ambient seismic noise monitoring at Parkfield measures relative velocity change (dv/v) from 2002 to 2022The post‐seismic healing from the 2004 Parkfield earthquake continues today, with the current dv/v exceeding the level before the eventA statistically significant long‐term increase correlates with the contractional axial strains caused by tectonic loading [ABSTRACT FROM AUTHOR]

Cross-correlations of ambient seismic noise are widely used for seismic velocity imaging, monitoring and ground motion analyses. A typical step in analysing noise cross-correlation functions (NCFs) is stacking short-term NCFs over longer time periods to increase the signal quality. Spurious NCFs could contaminate the stack, degrade its quality and limit its use. Many methods have been developed to improve the stacking of coherent waveforms, including earthquake waveforms, receiver functions and NCFs. This study systematically evaluates and compares the performance of eight stacking methods, including arithmetic mean or linear stacking, robust stacking, selective stacking, cluster stacking, phase-weighted stacking, time–frequency phase-weighted stacking, N th-root stacking and averaging after applying an adaptive covariance filter. Our results demonstrate that, in most cases, all methods can retrieve clear ballistic or first arrivals. However, they yield significant differences in preserving the phase and amplitude information. This study provides a practical guide for choosing the optimal stacking method for specific research applications in ambient noise seismology. We evaluate the performance using multiple onshore and offshore seismic arrays in the Pacific Northwest region. We compare these stacking methods for NCFs calculated from raw ambient noise (referred to as Raw NCFs) and from ambient noise normalized using a one-bit clipping time normalization method (referred to as One-bit NCFs). We evaluate six metrics, including signal-to-noise ratios, phase dispersion images, convergence rate, temporal changes in the ballistic and coda waves, relative amplitude decays with distance and computational time. We show that robust stacking is the best choice for all applications (velocity tomography, monitoring and attenuation studies) using Raw NCFs. For applications using One-bit NCFs, all methods but phase-weighted and N th-root stacking are good choices for seismic velocity tomography. Linear, robust and selective stacking methods are all equally appropriate choices when using One-bit NCFs for monitoring applications. For applications relying on accurate relative amplitudes, the linear, robust, selective and cluster stacking methods all perform well with One-bit NCFs. The evaluations in this study can be generalized to a broad range of time-series analysis that utilizes data coherence to perform ensemble stacking. Another contribution of this study is the accompanying open-source software package, StackMaster , which can be used for general purposes of time-series stacking. [ABSTRACT FROM AUTHOR]

The subsurface materials of Ciudad de México have unique mechanical properties that give rise to strong site effects. We investigated temporal changes in the seismic velocity at strong-motion and broad-band seismic stations throughout Mexico City, including sites with different geologic characteristics ranging from city center locations situated on lacustrine clay to hillsite locations on volcanic bedrock. We used autocorrelations of urban seismic noise, enhanced by waveform clustering, to extract subtle seismic velocity changes by coda wave interferometry. We observed and modeled seasonal, co-, and postseismic changes, as well as a long-term linear trend in seismic velocity. Seasonal variations can be explained by self-consistent models of thermo-elastic and poro-elastic changes in the subsurface shear wave velocity. Overall, sites on lacustrine clay-rich sediments appear to be more sensitive to seasonal surface temperature changes, whereas sites on alluvial and volcaniclastic sediments and on bedrock are sensitive to precipitation. The 2017 Mw 7.1 Puebla and 2020 Mw 7.4 Oaxaca earthquakes both caused a clear drop in seismic velocity followed by a time-logarithmic recovery that may still be ongoing for the 2017 event at several sites, or that may remain incomplete. The slope of the linear trend in seismic velocity is correlated with the downward vertical displacement of the ground measured by Interferometric Synthetic Aperture Radar, suggesting a causative relationship and supporting earlier studies on changes in the resonance frequency of sites in the Mexico City basin due to groundwater extraction. Our findings show how sensitively shallow seismic velocity, and in consequence, site effects, react to environmental, tectonic and anthropogenic processes. They also demonstrate that urban strong-motion stations provide useful data for coda-wave monitoring given sufficiently high-amplitude urban seismic noise. [ABSTRACT FROM AUTHOR]

Seismograms contain multiple sources of seismic waves, from distinct transient signals such as earthquakes to continuous ambient seismic vibrations such as microseism. Ambient vibrations contaminate the earthquake signals, while the earthquake signals pollute the ambient noise's statistical properties necessary for ambient-noise seismology analysis. Separating ambient noise from earthquake signals would thus benefit multiple seismological analyses. This work develops a multitask encoder–decoder network named WaveDecompNet to separate transient signals from ambient signals directly in the time domain for 3-component seismograms. We choose the active-volcanic Big Island in Hawai'i as a natural laboratory given its richness in transients (tectonic and volcanic earthquakes) and diffuse ambient noise (strong microseism). The approach takes a noisy 3-component seismogram as input and independently predicts the 3-component earthquake and noise waveforms. The model is trained on earthquake and noise waveforms from the STandford EArthquake Dataset (STEAD) and on the local noise of seismic station IU.POHA. We estimate the network's performance by using the explained variance metric on both earthquake and noise waveforms. We explore different neural network designs for WaveDecompNet and find that the model with long-short-term memory (LSTM) performs best over other structures. Overall, we find that WaveDecompNet provides satisfactory performance down to a signal-to-noise ratio (SNR) of 0.1. The potential of the method is (1) to improve broad-band SNR of transient (earthquake) waveforms and (2) to improve local ambient noise to monitor the Earth's structure using ambient noise signals. To test this, we apply a short-time average to a long-time average filter and improve the number of detected events. We also measure single-station cross-correlation functions of the recovered ambient noise and establish their improved coherence through time and over different frequency bands. We conclude that WaveDecompNet is a promising tool for a broad range of seismological research. [ABSTRACT FROM AUTHOR]

Distributed Acoustic Sensing (DAS) is increasingly recognised as a valuable tool for glaciological seismic applications, although analysing the large data volumes generated in acquisitions poses computational challenges. We show the potential of active-source DAS to image and characterise subglacial sediment beneath a fast-flowing Greenlandic outlet glacier, estimating the thickness of sediment layers to be 20–30 m. However, the lack of subglacial velocity constraint limits the accuracy of this estimate. Constraint could be provided by analysing cryoseismic events in a counterpart 3-day record of passive seismicity through, for example, seismic tomography, but locating them within the 9 TB data volume is computationally inefficient. We describe experiments with data compression using the frequency-wavenumber (f-k) transform ahead of training a convolutional neural network, that provides a ~300-fold improvement in efficiency. In combining active and passive-source and our machine learning framework, the potential of large DAS datasets could be unlocked for a range of future applications. [ABSTRACT FROM AUTHOR]