Teleseismic Detection Research Articles

Constraints on Terminus Dynamics at Greenland Glaciers From Small Glacial Earthquakes

Many large calving events at Greenland's marine‐terminating glaciers generate globally detectable glacial earthquakes. We perform a cross‐correlation analysis using regional seismic data to identify events below the teleseismic detection threshold, focusing on the 24 hr surrounding known glacial earthquakes at Greenland's three largest glaciers. We detect additional seismic events in the minutes prior to more than half of the glacial earthquakes we study and following one third of them. Waveform modeling shows source mechanisms like those of previously known glacial earthquakes, a result consistent with available imagery. The seismic events thus do not represent a failure of the high subaerial ice cliff like that expected to trigger large‐scale calving and a marine ice‐cliff instability but, rather, rotational, buoyancy‐driven calving events, likely of the full glacier thickness. A limited investigation of the prevalence of smaller seismic events at times outside glacial‐earthquake windows identifies several additional events. However, we find that calving at the three glaciers we study—Jakobshavn Isbræ, Helheim Glacier, and Kangerdlugssuaq Glacier—often occurs as sequences of discrete buoyancy‐driven events in which multiple icebergs ranging in size over as much as three orders of magnitude are all lost within ∼30 min. We demonstrate a correlation between glacial‐earthquake magnitude and iceberg size for events with well‐constrained iceberg‐area estimates. Our results suggest that at least 10–30% more dynamic mass loss occurs through buoyancy‐driven calving at Greenland's glaciers than previously appreciated.

Improving Regional and Teleseismic Detection for Single-Trace Waveforms Using a Deep Temporal Convolutional Neural Network Trained with an Array-Beam Catalog.

The detection of seismic events at regional and teleseismic distances is critical to Nuclear Treaty Monitoring. Traditionally, detecting regional and teleseismic events has required the use of an expensive multi-instrument seismic array; however in this work, we present DeepPick, a novel seismic detection algorithm capable of array-like detection performance from a single-trace. We achieve this performance through three novel steps: First, a high-fidelity dataset is constructed by pairing array-beam catalog arrival-times with single-trace waveforms from the reference instrument of the array. Second, an idealized characteristic function is created, with exponential peaks aligned to the cataloged arrival times. Third, a deep temporal convolutional neural network is employed to learn the complex non-linear filters required to transform the single-trace waveforms into corresponding idealized characteristic functions. The training data consists of all arrivals in the International Seismological Centre Database for seven seismic arrays over a five year window from 1 January 2010 to 1 January 2015, yielding a total training set of 608,362 detections. The test set consists of the same seven arrays over a one year window from 1 January 2015 to 1 January 2016. We report our results by training the algorithm on six of the arrays and testing it on the seventh, so as to demonstrate the generalization and transportability of the technique to new stations. Detection performance against this test set is outstanding, yielding significant improvements in recall over existing techniques. Fixing a type-I error rate of 0.001, the algorithm achieves an overall recall (true positive rate) of 56% against the 141,095 array-beam arrivals in the test set, yielding 78,802 correct detections. This is more than twice the 37,572 detections made by an STA/LTA detector over the same period, and represents a 35% improvement over the 58,515 detections made by a state-of-the-art kurtosis-based detector. Furthermore, DeepPick provides at least a 4 dB improvement in detector sensitivity across the board, and is more computationally efficient, with run-times an order of magnitude faster than either of the other techniques tested. These results demonstrate the potential of our algorithm to significantly enhance the effectiveness of the global treaty monitoring network.

Spiral‐Arm Seismic Arrays

Abstract Seismic arrays have many uses for signal enhancement, from surface‐wave characterization of the near surface to teleseismic detection in the context of monitoring nuclear tests. Many variants of the geometrical configuration of stations have been used with the objective of maximizing potential resolution of the incoming wavefronts direction of arrival. A versatile class of array configurations, with good resolution properties, can be constructed with multiple spiral arms. The array response is comparable with the same number of full circles, but with far fewer stations and is robust to minor position changes in emplacement. The desirable properties of the spiral‐arm arrays are illustrated for a permanent array in the Precambrian Pilbara craton in northwestern Australia and for a temporary array on ancient sediments in southern Queensland, Australia. In each case, the practical array response is very good and matches the theoretical expectations. The spiral‐arm configuration allows the deployment of relatively large aperture arrays with a limited number of stations, which is advantageous in a broad range of seismic applications, including near‐surface characterization. Online Material: Figures illustrating the relation between spiral‐arm and multiring circular arrays.

A quantitative evaluation of the annual variation in teleseismic detection capability at Syowa Station, Antarctica

In this study, we evaluate the annual variation in teleseismic detection capability at Syowa Station (69.0°S, 39.6°E) located in East Antarctica, a variation that has been noted in previous studies. For the quantitative evaluation of the annual variation, we introduced a statistical model of a magnitude–frequency distribution of earthquakes covering the entire magnitude range. The annual variation in the model parameter that quantifies the detection capability was then estimated by using Bayesian analysis. In the estimation, we incorporated the annual variation in air temperature at the station and succeeded in clarifying the significant effect that the variation in temperature has on the teleseismic detection capability.

Teleseismic Detection of a Slow Precursor to the Great 1989 Macquarie Ridge Earthquake

Low-frequency spectra for the 1989 Macquarie Ridge earthquake (magnitude 8.2) show an amplitude increase and a phase-delay decrease below 6 millihertz that require a short-term slow precursor. This earthquake can be modeled as a compound event in which a fast-rupturing, ordinary earthquake was initiated by an episode of slow, smooth deformation that began more than 100 seconds before the main shock. The moment released in the slow precursor was large, about 3 x 10(20) newton-meters, equivalent to an event of magnitude 7.6. The data are consistent with the precursor being generated in a region of the oceanic upper mantle below the main rupture.

Earthquake swarms on the Mid‐Atlantic Ridge: Products of magmatism or extensional tectonics?

By analogy with the microearthquake swarms which accompany terrestrial volcanic eruptions, teleseismically observed earthquake swarms along mid‐ocean ridges are commonly thought to be indicators of volcanism, particularly dike emplacement. We test this assumption by investigating the spatial and temporal patterns and other characteristics of earthquakes in 34 swarms on the Mid‐Atlantic Ridge. Resolution of the relative locations of swarm events is improved by the use of a multiple‐event relocation technique. Most swarms extend 5 to 15 km across the ridge axis and 5 to 20 km along it. None of the swarms displays the large length‐to‐width ratio and temporal migration of epicenters along axis which are characteristic of earthquake swarms in the well‐studied rift zones of Hawaii and Iceland. The relative locations, magnitudes, and source mechanisms of swarm events are most easily explained if teleseismically observable swarms are expressions of extension across a substantial portion of the median valley and perhaps into the rift mountains. The possibility that some of these swarms include events directly associated with magmatic intrusion cannot be excluded, however. Like their terrestrial counterparts, purely volcanic earthquakes on the mid‐ocean ridges probably occur dominantly at magnitudes below the teleseismic detection threshold of global seismic networks (mb < 4.5). The largest normal‐faulting earthquakes on mid‐ocean ridges and many of the teleseismically observed swarms likely occur on the faults marking the inner rift wall, where the crust first experiences significant faulting and is broken into numerous closely‐spaced normal faults. We speculate that the high degree of order in the dimensions, spacing, throws, and mechanical properties of these faults, combined with stress concentrations arising naturally from the mechanical behavior of the lithosphere at rifted, slow‐spreading ridges, may account for the common occurrence of earthquake swarms.

Study of magnitudes, seismicity, and earthquake detectability using a global network

Abstract Based on 10 yr of observations reported in the International Seismological Centre bulletins (1971-1980), m b magnitudes for about 70,000 earthquakes have been recomputed using a maximum-likelihood estimation technique. Reportings from a network of 115 globally distributed stations were used in these calculations. Comparison to conventional m b estimates show that the network magnitude bias problem is quite significant at low and intermediate magnitudes. Recurrence statistics based on the revised m b estimates give b values consistent with those obtained from LASA and NORSAR data, and indicate that the number of earthquakes worldwide of m b ≧ 4.0 averages about 7,500 annually. This number is much lower than previous estimates. The teleseismic detection capability of the network (requiring at least four detecting stations) has been evaluated based on recurrence statistics. The estimated 90 per cent incremental m b threshold ranges from 3.9 to 4.5 in the northern hemisphere and from 4.2 to 4.8 in the southern hemisphere. This is consistent with results obtained by the “Networth” approach.

Teleseismic detection in the Aleutian Island Arc

Recently it has become apparent that teleseismic detection has decreased substantially in many regions of the world. The major decrease was related to the closure of the VELA arrays in the United States during the late 1960's. This detection decrease has been recognized in South and Central America, Mexico, the Kuriles, the Caribbean, Tonga, and the New Hebrides. In this paper the effect of the closure of these arrays on the reporting of events in the Aleutian Island Arc is examined. In the Aleutians, the detection history is complicated by the short‐term installation of a local network on and near Amchitka Island during the early 1970's. The temporal coincidence of the installation of this network and the closure of the VELA arrays delayed the detection decrease in the central Aleutians until the Amchitka network was closed in early 1973. Reporting in the eastern Aleutians was unaffected by the installation of the Amchitka network. In that region the detection decreased between 1968 and 1970, the time of the closure of the VELA arrays. New techniques have been developed which make it possible to determine the effect of station installation or closure on the reporting in some regions. These techniques rely on plots which show the distribution of an observed seismicity rate change in the magnitude domain. These plots make it possible to recognize probable detection changes and to determine quantitatively magnitude cutoffs which avoid these changes. The magnitude level at which these cutoffs are made is termed the minimum magnitude of homogeneity (mmin h). The reporting of events with mb≤4.6 in the Aleutians decreased substantially during the mid‐1970's, so mmin h in this region is 4.7. This is different from the magnitude of completeness (mmin c) which is mb = 5.0±0.1. If one is interested in examining seismicity rates for changes which may be precursors to earthquakes, then awareness of detection‐related changes and magnitude cutoffs which avoid these changes are essential for meaningful results.

Teleseismic Detection between and below the Microseism Peaks

In this communication teleseismic signals and background noise will be examined in the frequency band 0·02 to 0·1 Hz (10 to 50 s) extending from the noise valley near 25 s to the noise minimum or crevice near 15 s between the microseism peaks. Power spectra and digital band pass filtering of the background noise and a small event will be given by way of illustration. If a low noise broad band accelerometer and direct digital recording with large dynamic range are not both used, long period (25 to 100 s) and short period (10 to 15 s) teleseismic signals cannot be obtained simultaneously without degradation of the data. Typical signal to ambient noise levels for earthquakes obtainable in these two frequency bands will also be presented.

Teleseismic Detection with Wide Band Vertical and Horizontal Accelerometers

THE observation of distant seismic events using a wide-band vertical accelerometer has recently been reported1. This is a preliminary report of the observation of teleseisms with both the vertical accelerometer and a new wide-band horizontal accelerometer at a surface site. Previous surface measurements of teleseisms with horizontal seismometers have been complicated by high levels of instrument and ground noise2. With present instrumentation and the aid of simple digital filtering, it is possible to obtain signal-to-noise ratios for a horizontal instrument at the surface comparable to those for the vertical instrument. In fact, if Love waves are produced, the signal-to-noise ratio for the surface horizontal accelerometer often exceeds that for the vertical.