Search NASA⌕ Search
Home/Archive/DOE OSTI

DOE OSTI · 1981026

A Data-Driven Framework for Automated Detection of Aircraft-Generated Signals in Seismic Array Data Using Machine Learning

Zhang, Xinxiang·Arrowsmith, Stephen·Tsongas, Sotirios·Hayward, Chris·Meng, Haoran·Ben-Zion, Yehuda
View original source ↗Export metadata ↓

Abstract

Abstract Ground motions associated with aircraft overflights can cover a significant portion of the seismic data collected by shallowly emplaced seismometers, such as new nodal and Distributed Acoustic Sensing systems. This article describes the first published framework for automated detection of aircraft on single channel and multichannel seismic data. The seismic data are converted to spectrograms in a sliding time window and classified as aircraft or nonaircraft in each window using a deep convolutional neural network trained with analyst-labeled data. A majority voting scheme is used to convert the output from the sequence of sliding time windows onto a decision time sequence for each channel and to combine the binary classifications on the decision time sequences across multiple channels. Precision, recall, and F-score are used to quantify the detection performance of the algorithm on nodal data using fourfold time-series cross validation. By applying our framework to data from the Sage Brush Flats nodal array in Southern California, we provide a benchmark performance and demonstrate the advantage of using an array of sensors.

Explore related subjects

Geochemistry & Geophysics

Keep this discovery

Explore connections, maps & timelines

BibTeXRIS

Zhang, Xinxiang, Arrowsmith, Stephen, Tsongas, Sotirios, Hayward, Chris, Meng, Haoran, Ben-Zion, Yehuda. 2021-11-03. A Data-Driven Framework for Automated Detection of Aircraft-Generated Signals in Seismic Array Data Using Machine Learning. https://doi.org/10.1785/0220210198

Cite the original work for its findings. Save a collection to share your selection of sources.

KEEP EXPLORING

Related reports

DOE OSTI2024

Formation of a Large Cold Groundwater Mantle Helium Anomaly and High Temperature Geothermal Resources in Response to Bimodal Magmatism Near Roosevelt Hot Springs and Utah FORGE, Milford Valley, Southwest Utah

Abstract A large mantle helium anomaly and separate domains of high heat flow are the predominant manifestations of bimodal magmatic activity in the Milford valley. The mantle helium anomaly (1.9–2.6 R/Ra) covers 270 km 2 and is subdivided into two separated domains: a cold shallow groundwater regime and high temperature hydrothermal activity. The zone of anomalous heat flow covers >100 km 2 and is also subdivided into two adjacent domains, comprising hydrothermal activity at Roosevelt Hot Springs (RHS) (3–7 W/m 2 ) and conductive heat flow (100–180 mW/m 2 ). While the transfer of heat and mantle helium at RHS are coupled, heat and helium transfer are decoupled in the adjacent cold groundwater regime to the west. Both the mantle helium and geothermal anomalies are attributed to recent mafic‐felsic magmatic intrusions of >400 km 3 , however, the absence of volcanic eruptions <500,000 years indicates magmas stall before rising to shallow crustal level <10 km depth. Deep level magmatism produces a felsic composition melt, which is inferred to be responsible for the widespread and near uniform range of diluted mantle helium values. A thick and impermeable mass of crystalline granitic basement rock at the mid‐crustal level divides the ascent of mantle helium into separate flow paths. It may also impede the rise of buoyant magma trapping thermal energy that facilitates partial melting, slow cooling, and development of a thick thermal aureole. Partitioning of convective and conductive thermal regimes and independent flow paths supplying deeply derived helium characterize the development of a large long‐lived magma‐related geothermal system.

Geochemistry & Geophysics↗
DOE OSTI2024

Geodetic Evidence for Distributed Shear Below the Brittle Crust of the Walker Lane, Western United States

Abstract Models of active deformation of the Earth's crust are predominantly represented with dislocations having a downdip continuation into the lower crust, where the fault slips continuously. This model predicts surface strain accumulation concentrated near the fault during the interseismic period. In an alternative model, faults do not extend beneath the elastic portion of the crust and are accompanied by a wide zone of distributed shear underneath, predicting a more constant strain rate lacking concentrations at the faults. We use high‐precision GPS data collected across the northern and central Walker Lane, USA— a region of complex faulting near the western edge of the Basin and Range Province to evaluate which model is appropriate. Despite the existence of dense continuous and semi‐continuous geodetic networks that have been surveyed for ∼20 years, the horizontal velocities reveal no evidence of localized strain accumulation across the fault surface expressions. Instead, deformation within the Walker Lane is uniformly linear, suggesting that the surface deformation reflects distributed shear within the ductile crust rather than focused deformation at faults. This suggests no downdip extension of the faults below the seismogenic layer. The shear zone is 172 ± 6 km wide in the northernmost Walker Lane narrowing to 116 ± 4 km in the central Walker Lane. The total velocity budget across the shear zone is 7.2 ± 0.1 mm/yr in the north, increasing to 10.1 ± 0.1 mm/yr in the central Walker Lane. We conclude that assuming the presence of lower crustal dislocations when estimating geodetic faults slip rates may be inappropriate.

Geochemistry & Geophysics↗