An official website of the United States government
Abstract Urban basin investigation is crucial for seismic hazard assessment and mitigation. Recent advances in robust nodal‐type sensors facilitate the deployment of large‐N arrays in urban areas for high‐resolution basin imaging. However, arrays typically operate for only one month due to the instruments' battery life, and hence, only record a few teleseismic events. This limits the number of available teleseismic events for traditional receiver function (RF) analysis‐the primary method used in sediment‐basement interface imaging in passive source seismology. Insufficient stacking of RFs from a limited number of earthquakes could, however, introduce significant biases to the results. In this study, we present a novel Bayesian array‐based Coherent Receiver Function (CRF) method that can leverage datasets from short‐term dense arrays to constrain basin geometry. We cast the RF deconvolution as a sparsity‐promoted inverse problem, in which the deconvolution at a single‐station involves the constraints from neighboring stations and multiple events. We solve the inverse problem using a trans‐dimensional Markov chain Monte Carlo Bayesian algorithm to find an ensemble of RF solutions, which provides a quantitative way of deciding which features are well resolved and warrant geological interpretation. An application in the northern Los Angeles basin demonstrates the ability of our method to produce reliable and easy‐to‐interpret RF images. The use of dense seismic networks and the state‐of‐the‐art Bayesian array‐based CRF method can provide a robust approach for subsurface structure imaging.
more »
« less
An array-based receiver function deconvolution method: methodology and application
SUMMARY Receiver functions (RFs) estimated on dense arrays have been widely used for the study of Earth structures across multiple scales. However, due to the ill-posedness of deconvolution, RF estimation faces challenges such as non-uniqueness and data overfitting. In this paper, we present an array-based RF deconvolution method in the context of emerging dense arrays. We propose to exploit the wavefield coherency along a dense array by joint inversions of waveforms from multiple events and stations for RFs with a minimum number of phases required by data. The new method can effectively reduce the instability of deconvolution and help retrieve RFs with higher fidelity. We test the algorithm on synthetic waveforms and show that it produces RFs with higher interpretability than those by the conventional RF estimation practice. Then we apply the method to real data from the 2016 Incorporated Research Institutions for Seismology (IRIS) community wavefield experiment in Oklahoma and are able to generate high-resolution RF profiles with only three teleseismic earthquakes recorded by the temporary deployment. This new method should help enhance RF images derived from short-term high-density seismic profiles.
more »
« less
- Award ID(s):
- 1722879
- PAR ID:
- 10198332
- Date Published:
- Journal Name:
- Geophysical Journal International
- Volume:
- 222
- Issue:
- 1
- ISSN:
- 0956-540X
- Page Range / eLocation ID:
- 1 to 14
- Format(s):
- Medium: X
- Sponsoring Org:
- National Science Foundation
More Like this
-
-
ABSTRACT The receiver function (RF) is a widely used crustal imaging technique. In principle, it assumes relatively noise-free traces that can be used to target receiver-side structures following source deconvolution. In practice, however, mode conversions and reflections may be severely degraded by noisy conditions, hampering robust estimation of crustal parameters. In this study, we use a sparsity-promoting Radon transform to decompose the observed RF traces into their wavefield contributions, that is, direct conversions, multiples, and incoherent noise. By applying a crustal mask on the Radon-transformed RF, we obtain noise-free RF traces with only Moho conversions and reflections. We demonstrate, using a synthetic experiment and a real-data example from the Sierra Nevada, that our approach can effectively denoise the RFs and extract the underlying Moho signals. This greatly improves the robustness of crustal structure recovery as exemplified by subsequent H−κ stacking. We further demonstrate, using a station sitting on loose sediments in the Upper Mississippi embayment, that a combination of our approach and frequency-domain filtering can significantly improve crustal imaging in reverberant settings. In the presence of complex crustal structures, for example, dipping Moho, intracrustal layers, and crustal anisotropy, we recommend caution when applying our proposed approach due to the difficulty of interpreting a possibly more complicated Radon image. We expect that our technique will enable high-resolution crustal imaging and inspire more applications of Radon transforms in seismic signal processing.more » « less
-
Abstract Fault zone complexities contain important information about factors controlling earthquake dynamic rupture. High‐resolution fault zone imaging requires high‐quality data from dense arrays and new seismic imaging techniques that can utilize large portions of recorded waveforms. Recently, the emerging Distributed Acoustic Sensing (DAS) technique has enabled near‐surface imaging by utilizing existing telecommunication infrastructure and anthropogenic noise sources. With dense sensors at several meters' spacing, the unaliased wavefield can provide unprecedented details for fault zones. In this work, we use a DAS array converted from a 10‐km underground fiber‐optic cable across Ridgecrest City, California. We report clear acausal and coda signals in ambient noise cross‐correlations caused by surface‐to‐surface wave scattering. We use these scattering‐related waves to locate and characterize potential faults. The mapped fault locations are generally consistent with those in the United States Geological Survey Quaternary Fault database of the United States but are more accurate than the extrapolated ones. We also use waveform modeling to infer that a 35 m wide, 90 m deep fault with 30% velocity reduction can best fit the observed scattered coda waves for one of the identified fault zones. These findings demonstrate the potential of DAS for passive imaging of fine‐scale faults in an urban environment.more » « less
-
Both large and small earthquakes rupture in complex ways. However, microearthquakes are often simplified as point sources and their rupture properties are challenging to resolve. We leverage seismic wavefields recorded by a dense array in Oklahoma to image microearthquake rupture processes. We construct machine-learning enabled catalogs and identify four spatially disconnected seismic clusters. These clusters likely delineate near-vertical strike-slip faults. We develop a new approach to use the maximum absolute SH-wave amplitude distributions (S-wave wavefields) to compare microearthquake rupture processes. We focus on one cluster with earthquakes located beneath the dense array and have a local magnitude range of -1.3 to 2.3. The S-wave wavefields of single earthquakes are generally coherent but differ slightly between the low-frequency (<12 Hz) and high-frequency (>12 Hz) bands. The S-wave wavefields are coherent between different earthquakes at low frequencies with average correlation coefficients greater than 0.95. However, the wavefield coherence decreases with increasing frequency for different earthquakes. This reduced coherence is likely due to the rupture differences among individual earthquakes. Our results suggest that earthquake slip of the microearthquakes dominates the radiated S-wave wavefields at higher frequencies. Our method suggests a new direction in resolving small earthquake source attributes using dense seismic arrays without assuming a rupture model.more » « less
-
We constrained sedimentary basin structure using a nodal seismic array consisting of ten dense lines that overlie multiple basins in the northern Los Angeles area. The dense array consists of 758 seismic nodes, spaced ~250–300 m apart along linear transects, that recorded ground motions for 30–35 days. We applied the receiver function (RF) technique to 16 teleseismic events to investigate basin structure. Primary basin-converted phases were identified in the RFs. A shear wave velocity model produced in a separate study using the same dataset was incorporated to convert the basin time arrivals to depth. The deepest part of the San Bernardino basin was identified near the Loma Linda fault at a depth of 2.4 km. Basin depths identified at pierce points for separate events reveal lateral changes in basin depth across distances of ~2–3 km near individual stations. A significant change in basin depth was identified within a small distance of ~4 km near the San Jacinto fault. The San Gabriel basin exhibited the largest basin depths of all three basins, with a maximum depth of 4.2 km. The high lateral resolution from the dense array helped to reveal more continuous structures and reduce uncertainties in the RFs interpretation. We discovered a more complex basin structure than previously identified. Our findings show that the basins’ core areas are not the deepest, and significant changes in basin depth were observed near some faults, including the San Jacinto fault, Fontana fault, Red Hill fault and Indian Hill fault.more » « less
- Free Publicly Accessible Full Text
-
Accepted Manuscript1.0
- Journal Article:
- https://doi.org/10.1093/gji/ggaa113
