Search NASA⌕ Search

Engineering topics

Kintner, Jonas

Publications and source records attributed to Kintner, Jonas.

Automatic Waveform Quality Control for Surface Waves Using Machine Learning

Surface-wave seismograms are widely used by researchers to study Earth’s interior and earthquakes. To extract information reliably and robustly from a suite of surface waveforms, the signals require quality control screening to reduce artifacts from signal complexity and noise. This process has usually been completed by human experts labeling each waveform visually, which is time consuming and tedious for large data sets. We explore automated approaches to improve the efficiency of waveform quality control processing by investigating logistic regression, support vector machines, K-nearest neighbors, random forests (RF), and artificial neural networks (ANN) algorithms. To speed up signal quality assessment, we trained these five machine learning (ML) methods using nearly 400,000 human-labeled waveforms. The ANN and RF models outperformed other algorithms and achieved a test accuracy of 92%. We evaluated these two best-performing models using seismic events from geographic regions not used for training. The results show that the two trained models agree with labels from human analysts but required only 0.4% of the time. Although the original (human) quality assignments assessed general waveform signal-to-noise, the ANN or RF labels can help facilitate detailed waveform analysis. Our investigations demonstrate the capability of the automated processing using these two ML models to reduce outliers in surface-wave-related measurements without human quality control screening.

58 GEOSCIENCES↗

Relocation of Light and Moderate-Magnitude ( M 4-6) Seismicity Along the Central Mid-Atlantic

In this study, we estimate precise relative epicentroid locations of earthquakes along the Mid-Atlantic Ridge (MAR) between 12°N and 19°N using intermediate-period Rayleigh and Love wave cross-correlation measurements. We relocate 152 earthquakes with magnitudes ranging between 4.0 and 6.0 and when possible use P waveform modeling to estimate centroid depths. We compare surface-wave locations to locations in the U.S. Geological Survey, Global Centroid Moment Tensor, and Pacific Marine Environmental Laboratory hydroacoustic catalogs (Smith et al., 2002, https://doi.org/10.1029/2001GL013912) and demonstrate that surface-wave based relative location techniques provide significant improvement on precision. Our analyses shift locations to more tectonically consistent alignments that correlate strongly with bathymetric features. Relocations along the Fifteen-Twenty and Marathon Fracture Zones (FZs) move to within a few kilometers of the transform bathymetry. Events along ridge segments scatter over a distance of about 10–20 km perpendicular to the local plate boundary and most ridge-segment seismicity clusters spatially. North of the Fifteen-Twenty FZ, ridge events along several segments locate preferentially in the North American Plate, suggesting recent asymmetric seismic strain release and perhaps asymmetric spreading. Along the ridge between 13°N and 14°N, moderate-magnitude seismic strain release concentrates in an area with two active oceanic core complexes (OCC) along the western rift boundary, but the largest events along this segment appear to occur along the eastern boundary rift valley faults and represent deformation opposite the OCC detachment surface processes. Finally, precise moderate-magnitude earthquake epicentroid locations for events that occurred over the last few decades complement the results of temporary seismometer deployments that have imaged microseismic activity along the MAR.

58 GEOSCIENCES↗