Jessie L. Bonner

Development of a Time-Domain, Variable-Period Surface-Wave Magnitude Measurement Procedure for Application at Regional and Teleseismic Distances, Part II: Application and Ms–mb Performance

The Russell surface-wave magnitude formula, developed in Part I of this two-part article, and the M_s(VMAX) measurement technique, discussed in this article, provide a new method for estimating variable-period surface-wave magnitudes at regional and teleseismic distances. The M_s(VMAX) measurement method consists of applying Butterworth bandpass filters to data at center periods between 8 and 25 sec. The filters are designed to help remove the effects of nondispersed Airy phases at regional and teleseismic distances. We search for the maximum amplitude in all of the variable-period bands and then use the Russell formula to calculate a surface-wave magnitude. In this companion article, we demonstrate the capabilities of the method by using applications to three different datasets. The first application utilizes a dataset that consists of large earthquakes in the Mediterranean region. The results indicate that the M_s(VMAX) technique provides regional and teleseismic surface-wave magnitude estimates that are in general agreement except for a small distance dependence of −0.002 magnitude units per degree. We also find that the M_s(VMAX) estimates are less than 0.1 magnitude unit different than those from other formulas applied at teleseismic distances such as Rezapour and Pearce (1998) and Vanĕk et al. (1962). In the second and third applications of the method, we demonstrate that measurements of M_s(VMAX) versus m_b provide adequate separation of the explosion and earthquake populations at the Nevada and Lop Nor Test Sites. At the Nevada Test Site, our technique resulted in the misclassification of two earthquakes in the explosion population. We also determined that the new technique reduces the scatter in the magnitude estimates by 25% when compared with our previous studies using a calibrated regional magnitude formula. For the Lop Nor Test Site, we had no misclassified explosions or earthquakes; however, the data were less comprehensive. A preliminary analysis of Eurasian earthquake and explosion data suggest that similar slopes are obtained for observed M_s(VMAX) versus m_b data with m_b <5. Thus the data are not converging at lower magnitudes. These results suggest that the discrimination of explosions from earthquakes can be achieved at lower magnitudes using the Russell (2006) formula and the M_s(VMAX) measurement technique.

The Surface Wave Magnitude for the 9 October 2006 North Korean Nuclear Explosion

Surface waves were generated by the North Korean nuclear explosion of 9 October 2006 and recorded at epicentral distances up to 34 degrees, from which we estimated a surface wave magnitude (M{sub s}) of 2.94 with an interstation standard deviation of 0.17 magnitude units. The International Data Centre estimated a body wave magnitude (m{sub b}) of 4.1. This is the only explosion we have analyzed that was not easily screened as an explosion based on the differences between the M{sub s} and m{sub b} estimates. Additionally, this M{sub s} predicts a yield, based on empirical M{sub s}/Yield relationships, that is almost an order of magnitude larger then the 0.5 to 1 kiloton reported for this explosion. We investigate how emplacement medium effects on surface wave moment and magnitude may have contributed to the yield discrepancy.

Application of a Cepstral F Statistic for Improved Depth Estimation

The depth of a seismic event can be determined using peaks corresponding to depth phase-delay times found in the cepstrum of the seismic signal. Until now, however, there has been no method available to determine the significance of various peaks in the cepstrum. We have formulated a cepstral F statistic by using a classic approach to detect a signal in a number of stationarily correlated time series. The method attaches a statistical significance level to peaks in the cepstrum of seismic data caused by echoes in the signal. The method is particularly well suited for detecting depth phase echoes ( pP and sP ) recorded on regional seismic arrays. Detections determined from the peaks in the cepstral F statistic are then stacked as a function of pP-P and sP-P travel times predicted from the IASPEI global 1-D model (Kennett and Engdahl, 1991), using a modification of the method of depth-phase beamforming (Woodgold, 1998; Murphy et al. , 1999). Tests on synthetic data show the method is most successful when the P -wave arrival has a signal-to-noise ratio (SNR) greater than 4-7 and the depth phase exhibits an SNR greater than ∼2. Tests on complex regional data suggest that the depth-phase SNR must be between 4 and 8 for consistent successful application of the method. We tested the method by using events from the Hindu-Kush region of Afghanistan with well-determined depths as recorded on arrays at teleseismic distances. We determined the correct depths for 14 of the 19 events in this dataset (74% success rate), and we determined incorrect depths for two events, for a false-alarm rate of 10.5%. The difference between the depths calculated by the cepstral method and the prototype International Data Center (pIDC) decrease with increasing magnitude. For events with m b > 3.8, the average difference between the cepstral and pIDC depths was 7.9 km, with no false alarms. To test the operational capabilities of this method as a tool for data center use, we first analyzed 32 events that occurred on 12 February 2000 as located by the pIDC. Additionally, we analyzed 29 events located in Central and South America between 05 May and 15 June 2000 with event characteristics published by the International Data Center (IDC). Most of these events were also located by the National Earthquake Information Center. Our method determined statistically significant depths for 40 of these 61 events, with 11 having a low SNR at three or more recording arrays, while another 10 were either too shallow for analysis or did not exhibit depth phases. Only seven of the 61 events had depth phases listed in the Reviewed Event Bulletin produced by the pIDC and IDC. For these events, the average difference between the reported and cepstral F -statistic depths was 4.0 km. The pIDC/IDC fixed the depth of 26 of 61 events to 0.0. The cepstral F -statistic method determined depths ranging from 16 to 92 km for 15 of the events that had been constrained as surface foci. Overall, we believe the method would be most valuable when used by analysts to detect possible depth phases that could then verify or improve network-calculated focal depths. Manuscript received 25 February 2001.

Evaluation of Short-Period, Near-Regional Ms Scales for the Nevada Test Site

Surface wave magnitude (M_s) estimation for small events recorded at near-regional distances will often require a magnitude scale designed for Rayleigh waves with periods less than 10 sec. We have examined the performance of applying two previously published M_s scales on 7-sec Rayleigh waves recorded at distances less than 500 km. First, we modified the Marshall and Basham (1972) M_s scale, originally defined for periods greater than 10 sec, to estimate surface wave magnitudes for short-period Rayleigh waves from earthquakes and explosions on or near the Nevada Test Site. We refer to this modification as ^(M+B) M_s(7), and we have used short-period, high-quality dispersion curves to determine empirical path corrections for the 7-sec Rayleigh waves. We have also examined the performance of the Rezapour and Pearce (1998) formula, developed using theoretical distance corrections and surface wave observations with periods greater than 10 sec, for 7-sec Rayleigh waves ^(R+P) (M_S(7)) as recorded from the same dataset. The results demonstrate that both formulas can be used to estimate M_s for nuclear explosions and earthquakes over a wider magnitude distribution than is possible using conventional techniques developed for 20-sec Rayleigh waves. These M_s(7) values scale consistently with other Ms studies at regional and teleseismic distances with the variance described by a constant offset; however, the offset for the ^(M+B) M_s(7) estimates is over one magnitude unit nearer the teleseismic values than the ^(R+P) M_s(7) estimates. Using our technique, it is possible to employ a near-regional single-station or sparse network to estimate surface wave magnitudes, thus allowing quantification of the size of both small earthquakes and explosions. Finally, we used a jackknife technique to determine the false-alarm rates for the ^(M+B) M_s(7)-m_b discriminant for this region and found that the probability of misclassifying an earthquake as an explosion is 10%, while the probability of classifying an explosion as an earthquake was determined to be 1.2%. The misclassification probabilities are slightly higher for the ^(R+P) M_s(7) estimates. Our future research will be aimed at examining the transportability of these methods.

Characteristics of Chemical Explosive Sources from Time-Dependent Moment Tensors

Using a frequency-domain linear inversion technique and near-source broadband data, we inverted for the time-dependent source moment tensors of eight chemical explosions detonated in an open-pit coal mine during the Source Phenomenology Experiments (SPE) conducted by a consortium of U.S. research institutions to investigate a suite of explosive-source related problems. The moment tensors of the explosions from the inversion are dominated by their isotropic components regardless of variations between explosions in source size, confinement condition, and whether the explosion was on a bench and collapsed the vertical face of the bench. The percentage of isotropic moment-tensor component ranges from 96% to 98% for largest part of the source-time histories. Source-configuration variations result in differences that are most apparent in long-period moment-tensor spectra reflecting possible secondary source effects such as cylindrical source shape, spall, and compensated linear vector dipole (CLVD). Unconfined explosions show more oscillatory diagonal moment-tensor component time histories than confined and partially confined explosions possibly due to stronger free-surface effects such as material cast. Compared with pit explosions, deviatoric components of moment tensors of the two bench explosions are of higher amplitudes. There is a discernible long-period (<5 Hz) signal on one of the off-diagonal components, which could be related to the presence of the bench face in the source region and the horizontal material cast by the explosions. Although off-diagonal moment-tensor components comprise a small portion of the moment tensor, they are capable of generating a disproportionally large amount of shear waves.

Azimuthal Variation of Short-Period Rayleigh Waves from Cast Blasts in Northern Arizona

We have examined the generation and propagation of short-period (SP) Rayleigh waves ( Rg ) from two cast blasts of similar yield (∼726,000 kg of explosives), delay sequence, and near-source geology at a coal mine in northern Arizona. The blasts were oriented approximately perpendicular to each other, allowing for radiation pattern studies at regional distances. Near-source seismic ( P and Lg phases, the observation of these Rg radiation patterns at regional distances could act as a cast blast discriminant in regions of monitoring concern.

Determination of Love- and Rayleigh-Wave Magnitudes for Earthquakes and Explosions

Since the 1960s, comparing a Rayleigh-wave magnitude M s to the body-wave magnitude m b ( M s: m b) has been a robust tool for the discrimination of earthquakes and explosions. In this article, we apply a Rayleigh-wave formula as is to Love waves and examine the possibilities for discrimination using only surface-wave magnitudes ( M s: M s). To calculate the magnitudes, we apply the time-domain magnitude technique called M s(VMAX), developed by Russell (2006), to Rayleigh and Love waves from explosions and earthquakes. Our results indicate that, for the majority of the earthquakes studied (>75%), the M s(VMAX) obtained from Love waves is greater than that estimate from Rayleigh waves. Conversely, 79 of 82 nuclear explosions analyzed (96%) had network-averaged M s(VMAX)-Rayleigh equal to or greater than the M s(VMAX)-Love. We used logistic regression to examine an ![Graphic][1] discriminant. Cross-validation analysis of the new discriminant correctly identifies 57 of 82 explosions and 246 of 264 earthquakes, while misidentifying 22 explosions as earthquakes and 11 earthquakes as explosions. Further comparative research is planned for ![Graphic][2] versus M s: m b using common data. We fully expect that ![Graphic][3] will contribute significantly to multivariate event identification. [1]: /embed/inline-graphic-1.gif [2]: /embed/inline-graphic-2.gif [3]: /embed/inline-graphic-3.gif

Development of a Velocity Model for Black Mesa, Arizona, and the Southern Colorado Plateau from Multiple Data Sets

During the summer of 2003, a consortium of scientific institutions conducted a set of seismic experiments on Black Mesa, Arizona, in the southern Colorado Plateau to determine the velocity structure and crustal thickness below the mesa. We detonated a series of explosions, which were recorded by 130 near-source, vertical-component sensors and 25 broadband seismometers. The broadband stations, deployed in a linear array to the south of the mesa, also recorded earthquakes at regional and teleseismic distances during the duration of the deployment. Prior to the explosion series, we conducted a shallow refraction study at the site by using 20 three-component geophones. We utilized the multiple data sets recorded by the experiment to develop a new velocity model for Black Mesa. We analyzed the shallow refraction and explosion data to generate P - and S -velocity models for the upper crust. The P velocities ( α ) were determined by examining first-arrival times as a function of distance from the source. In addition, we performed ground-roll and surface-wave studies to develop a shear-wave velocity ( β ) structure for the upper crust. The results indicate an upper-crustal velocity structure that consists of very slow velocity sediments ( α β α and β , respectively. For the deeper-crustal and upper-mantle velocity, we used refraction data from broadband stations and inverted surface-wave dispersion data. We were not able to confidently determine the depth of the Moho, as our results indicate it to be between 37 and 46 km. The velocity model developed for Black Mesa was validated using comparisons between observed and synthetic waveforms and small ground-truth explosion location studies.

Effect of Fractures on Seismic Amplitudes from Explosions

Improved understanding of the seismic radiation generated by explosions in low coupling (damaged/fractured) media is extremely important for nuclear monitoring, as source coupling affects both detection and yield estimation. Some empirical evidence for seismic amplitude reductions have been noted for nuclear and chemical explosions detonated in fractured media (e.g., Sokolova, 2008). In order to define the physical mechanism responsible for the amplitude reduction and quantify the degree of the amplitude reduction in fractured rocks, we conducted Phase I of a multi‐phase explosion experiment in central New Hampshire. The experiment involved conducting explosions of various yields, including a 46.3‐kg explosion in the damage/fracture zone of a 231.8‐kg explosion and a 46.8‐kg shot in nearby undamaged rock. Our analysis confirms a seismic amplitude reduction in damaged rock by a factor of 2–3. The amplitude differences are frequency dependent, with the explosion in the undamaged rock having a higher corner frequency than the explosion in the damaged zone. The overshoot parameter for the virgin/undamaged rock shots is higher than that for the damaged rock shot. We found that the corner frequency correlates with the overshoot parameter, and only weakly correlates with the yield. Additional experiments will be conducted in the near future to further quantify seismic‐wave characteristics as a function of the depth of burial, type of explosives, and other factors. Online Material: Movies of explosions.

SOURCE PHENOMENOLOGY EXPERIMENTS IN ARIZONA

The Arizona Source Phenomenology Experiments (SPE) have resulted in an important dataset for the nuclear monitoring community. The 19 dedicated single-fired explosions and multiple delay-fired mining explosions were recorded by one of the most densely instrumented accelerometer and seismometer arrays ever fielded, and the data have already proven useful in quantifying confinement and excitation effects for the sources. It is very interesting to note that we have observed differences in the phenomenology of these two series of explosions resulting from the differences between the relatively slow (limestone) and fast (granodiorite) media. We observed differences at the two SPE sites in the way the rock failed during the explosions, how the S-waves were generated, and the amplitude behavior as a function of confinement. Our consortiums goal is to use the synergy of the multiple datasets collected during this experiment to unravel the phenomenological differences between the two emplacement media. The data suggest that the main difference between single-fired chemical and delay-fired mining explosion seismograms at regional distances is the increased surface wave energy for the latter source type. The effect of the delay-firing is to decrease the high-frequency P-wave amplitudes while increasing the surface wave energy because of the longer source duration and spall components. The results suggest that the single-fired explosions are surrogates for nuclear explosions in higher frequency bands (e.g., 6-8 Hz Pg/Lg discriminants). We have shown that the SPE shots, together with the mining explosions, are efficient sources of S-wave energy, and our next research stage is to postulate the possible sources contributing to the shear-wave energy.