Detection of records of weak local earthquakes using neural networks

Manual processing of large volumes of continuous observations produced by local seismic networks takes a lot of time, therefore, to solve this problem, automatic algorithms for detecting seismic events are used. Deterministic methods for solving the problem of detection, which do an excellent job of detecting intensive earthquakes, face critical problems when detecting weak seismic events (earthquakes). They are based on principles based on the calculation of energy, which causes multiple errors in detection: weak seismic events may not be detected, and high-amplitude noise may be mistakenly detected as an event. In our work, we propose a detection method capable of surpassing deterministic methods in detecting events on seismograms, successfully detecting a similar or more events with fewer false detections.

SEISMICITY of NORTHERN EURASIA in 2015

The review of the Northern Eurasia seismicity for 2015 includes a description of seismic networks, the results of analysis of the seismic regime and individual noticeable earthquakes in 16 regions of Russia and neighbouring countries. Seismic monitoring was carried out by the networks of seismic station of Russia, Azerbaijan, Armenia, Belarus, Kazakhstan, Kyrgyzstan, Latvia, Moldova, Turkmenistan, Tajikistan, Uzbekistan, Ukraine, including 599 digital, 7 analogue stations and eight seismic groups. In 2015, these networks registered about 27 thousand tectonic earthquakes, over 6 thousand volcanic earthquakes, 599 explosions, 23 mountain-tectonic shocks and induced earthquakes. Focal mechanisms of 592 earthquakes were determined, the information on manifestations of 449 perceptible earthquakes was collected. 26 shocks were felt in settlements of Northern Eurasia with an intensity Ii≥5. According to estimates of the annual number and released seismic energy in 2015 in comparison with the long-term characteristics of the seismic regime, the seismic process in most regions of Northern Eurasia proceeded in the “background” regime. An exception is Tajikistan and adjacent territories, where two strong earthquakes occurred – the Hindu Kush earthquake on October 26 with Mw=7.5, h=230 km in northern Afghanistan, near the border with Tajikistan, and the Sarez earthquake on December 7 with Mw=7.2, Ms=7.6, h=20 km in Tajikistan. Both earthquakes were accompanied by numerous aftershocks and were felt in Tajikistan with intensities Imax=7 and Imax=7–8 respectively, on the MSK-64 scale. Notable event on the territory of Northern Eurasia in 2015 is the emergence of the Muyakan sequence of earthquakes, the largest for the period of instrumental observations in the region "Baikal and Transbaikalia", as a result of which the number of recorded earthquakes in the region quadrupled concerning 2014. The other interesting fact is occurrence of tangible earthquakes in the regions, traditionally considered weakly seismic – near the Semipalatinsk test area in Eastern Kazakhstan (Chingiz earthquake on January 20, Ms=4.1, I0=5–6), in the Middle Urals (Middle Ural earthquake on October 18 with ML=4.7, I0=6) and in the southwest of East -European platform (Poltava earthquake on February 3 with KR=10.7, I0=6).

Benchmarking Automated Rayleigh-Wave Arrival Angle Measurements for USArray Seismograms

Surface-wave arrival angles are an important secondary set of observables to constrain Earth’s 3D structure. These data have also been used to refine information on the alignments of horizontal seismometer components with the geographic coordinate system. In the past, particle motion has been inspected and analyzed on single three-component seismograms, one at a time. But the advent of large, dense seismic networks has made this approach tedious and impractical. Automated toolboxes are now routinely used for datasets in which station operators cannot determine the orientation of a seismometer upon deployment, such as conventional free-fall ocean bottom seismometers. In a previous paper, we demonstrated that our automated Python-based toolbox Doran–Laske-Orientation-Python compares favorably with traditional approaches to determine instrument orientations. But an open question has been whether the technique also provides individual high-quality measurements for an internally consistent dataset to be used for structural imaging. For this feasibility study, we compared long-period Rayleigh-wave arrival angles at frequencies between 10 and 25 mHz for 10 earthquakes during the first half of 2009 that were recorded at the USArray Transportable Array—a component of the EarthScope program. After vigorous data vetting, we obtained a high-quality dataset that compares favorably with an arrival angle database compiled using our traditional interactive screen approach, particularly at frequencies 20 mHz and above. On the other hand, the presence of strong Love waves may hamper the automated measurement process as currently implemented.

Modeling Seismic Network Detection Thresholds Using Production Picking Algorithms

Estimating the detection threshold of a seismic network (the minimum magnitude earthquake that can be reliably located) is a critical part of network design and can drive network maintenance efforts. The ability of a station to detect an earthquake is often estimated by assuming the spectral amplitude for an earthquake of a given size, assuming an attenuation relationship, and comparing the predicted amplitude with the average station background noise level. This approach has significant uncertainty because of unknown regional attenuation and complications in computing small event power spectra, and it fails to account for the specific capabilities of the automatic seismic phase picker used in monitoring. We develop a data-driven approach to determine network detection thresholds using a multiband phase picking algorithm that is currently in use at the U.S. Geological Survey National Earthquake Information Center. We apply this picking algorithm to cataloged earthquakes to determine an empirical relationship of the observability of earthquakes as a function of magnitude and distance. Using this relationship, we produce maps of detection threshold using station spatial configuration and station noise levels. We show that quiet, well-sited stations significantly increase the detection capabilities of a network compared with a network composed of many noisy stations. Because our method is data driven, it has two distinct advantages: (1) it is less dependent on theoretical assumptions of source spectra and models of regional attenuation, and (2) it can easily be applied to any seismic network. This tool allows for an objective approach to the management of stations in regional seismic networks.

Microseismic network sensitivity in case of no seismic activity

Underground human activities, such as mining, shale gas and oil exploitation, waste-water disposal or geo-thermal plants, can cause earthquakes. These industry projects need to be monitored by local seismic networks in order to contain the risk. An ideal seismic network should have a triangulated grid, with spacing equal to the depth of the industrial activity with no associated industry noise. In many cases, stations are placed near noisy roads, factories or in a private garden, none of which are located at optimal nodes and which thus introduce great variations in the nose level. In this article, we present a work-flow to determine the sensitivity of any local network, even if there is no local event recorded. In other words: how small are the earthquakes that such seismic networks detect? This knowledge can be used as an argument for claiming an area to be seismically silent-inactive down to a certain magnitude or for evaluating the effect of an additional seismic station.A brief theory and work-flow description is followed by two real-case demonstrations from Czech Republic, Europe: first, a proof-test on a well- studied seismically active area of West Bohemia / Vogtland and second, an application to an uprising geothermal project in Litoměřice, where no seismic activity was detected in years of monitoring.

Persistent, “Mysterious” Seismoacoustic Signals Reported in Oklahoma State during 2019

ABSTRACT We report on the source of seismoacoustic pulses that were observed across the state of Oklahoma (OK) during summer of 2019, and the subject of national media coverage and speculation. Seismic network data collected across four U.S. states and interviews with witnesses to the pulse’s effect on residential structures demonstrate that they were triggered by routine ammunition disposal operations conducted by McAlester Army Ammunition Plant (McAAP). During these operations, conventional explosives destroy obsolete munitions stored in pits through a controlled sequence of electronically timed shots that occur over tens of minutes. Despite noise-abatement efforts that reduce coupling of acoustic energy with air, some lower frequency, subaudible (infrasonic) sound radiates from these shots as discrete pulses. We use nine months of blast log documents, seismic network records, analyst picks, and physical modeling to demonstrate that seismic stations as far as 640 km from McAAP sample these pulses, which record seasonal patterns in stratospheric and tropospheric winds, as well as the dynamic formation of waveguides and shadow zones. Digital short-term average to long-term average detectors that we augment with dynamic thresholds and time-binning operations identify these pulses with a fair probability, when compared with visual observations. Our analyses thereby provide estimates of observation rates for both partial and full sequences of these pulses, as well as single shots. We suggest that disposal operations can exploit existing, composite seismic networks to predict where residents are likely to witness blasting. Crucially, our data also show that dense seismic networks can record multiscale atmospheric processes in the absence of infrasound arrays.

Seismic Phase Association Based on the Maximum Likelihood Method

Phase association is a process that links seismic phases triggered at the stations of a seismic network to declare the occurrence of earthquakes. During phase association, a set of phases from different stations is examined to determine the common origin of phases within a specific region, predominantly on the basis of a grid search and the sum of observations. The association of seismic phases in local earthquake monitoring systems or earthquake early warning systems is often disturbed not only by transient noises, but also by large regional or teleseismic events. To mitigate this disturbance, we developed a seismic phase association method, binder_max, which uses the maximum likelihood method to associate seismic phases. The method is based on the framework of binder_ew, the phase associator of Earthworm, but it uses a likelihood distribution of the arrival information instead of stacking arrival information. Applying binder_max to data from seismic networks of South Korea and Ohio, United States, we found a significant improvement in the robustness of the method against large regional or teleseismic events compared to binder_ew. Our results indicate that binder_max can associate seismic phases of local earthquakes to the same degree as binder_ew as well as can avoid many of the false associations that have limited binder_ew.