Station Corrections Research Articles

Revised Velocity Structure of Western Montana

Using a one-dimensional (1D) layered earth approach, we determined a new crust and upper-mantle velocity model for western Montana to improve earthquake hypocenter locations. P -wave arrival times recorded on 280 stations from 1432 well-recorded earthquakes provided input for a sparse damped least-squares sensitivity matrix. We solved for 1D velocity structure, refractor depths, station corrections, and hypocentral positions by minimizing travel-time residuals. The new model has three layers with P -wave velocities of 5.70, 6.12, and 6.53 km/sec, with corresponding interface depths of 7.0, 19.8, and 39.7 km below the surface. The upper-mantle velocity is 8.00 km/sec. We determined station corrections and show that the Yellowstone caldera is the only regional geologic feature not adequately modeled by our new 1D velocity model. Use of the new model and station corrections reduced hypocenter location uncertainties and travel-time residuals, implying improved hypocenter locations for western Montana earthquakes. Online material : Table of seismographic stations.

Urgent aftershock observation of the 2004 off the Kii Peninsula earthquake using ocean bottom seismometers

Abstract The 2004 off the Kii Peninsula earthquake occurred on September 5, 2004. Knowing the precise aftershock distribution is important for understanding the mechanism of this earthquake. However, the hypocenter of the main shock was located more than 100 km offshore from the nearest station of the land observation network. In the three days after the main shock, we started ocean bottom seismometer (OBS) observation in order to determine the precise distribution of the aftershocks. We assumed a seismic velocity structure for the hypocenter calculation, based on the results of previous seismic refraction study. The station corrections were incorporated to locate the hypocenter precisely. The hypocenters located within an area covered by five OBSs show relatively small errors. It is found that the OBS-located hypocenters are located about 5.5 km east-southeast from those by JMA and the depth range of the aftershocks is about 5–25 km just beneath the Nankai trough axis. The aftershock hypocenters can be grouped into two clusters at different depths of about 10 km and about 20 km. It is inferred that the main shock also has a depth of 5–25 km. Since this extent of the main shock was larger than one of the oceanic crust of the Philippine Sea plate, the fault plane of the main shock extended at the upper most mantle of the Philippine Sea plate. Although we cannot assign the actual fault plane of the main shock form our observation results, it is clarified that intra-plate earthquakes occurred near the trench region. Our OBS result supports that the main shock was the earthquake not at the plate boundary but within the bending Philippine Sea plate near the trough axis.

Local and Duration Magnitudes in Northwestern Italy, and Seismic Moment Versus Magnitude Relationships

In the present work, we develop some local magnitude scales for northwestern Italy based on vertical short-period records. This study is motivated by the possibility of applying the computed scales to an instrumental catalog of more than 25,000 local earthquakes, as this region has been continuously monitored by 12 short-period vertical-component (1c) stations since the mid-1980s. Furthermore, a digital network of three-component (3c) broadband or 5 second sensors has monitored northwestern Italy since 1996. Today, a significant number of earthquakes have been simultaneously recorded by both networks, allowing the calibration of the 1c local scale by using magnitudes computed according to a scale derived for the 3c digital network. Moreover, because station Sant’ Anna di Valdieri houses both a 3c (code stv2) and 1c (code stv) sensors, the magnitude scales for the two networks can be developed using the same reference station. The magnitude scale M L = log A + log( R /100) + 0.0054( R − 100) + 3 − S is derived for the 3c digital network with the requirement that the correction S of station stv2 is zero. This scale is based on 10,057 maximum amplitudes (2822 earthquakes) computed from horizontal synthesized Wood-Anderson seismograms, in the hypocentral distance 10 to 310 km and in the range 0 ≤ M L ≤ 5. With respect to an carlier magnitude scale derived for the 3c network constraining the sum of all the station corrections to zero, the magnitudes predicted by the previous equations show an average bias of (−0.2 ± 0.1), which can be ascribed to the different constraint applied to the station corrections. The magnitudes predicted by the scale for the 3c network are used to calibrate magnitude scales based on either total duration or maximum amplitude from synthesized Wood-Anderson seismograms computed for each short-period vertical recording. The magnitude scale obtained considering maximum amplitudes from vertical short-period recordings is M L = log A + log( R /100) + 0.0041 ( R − 100) + 3 − S ′. The reliability of the obtained magnitude scales is assessed using 827 earthquakes different from those we considered in the regression analysis. Finally, the following seismic moment versus local magnitude relations are valid in the western Alps in the range 0 where M L3C is the local magnitude computed starting from the horizontal component of broadband (flat frequency response, from 0.033 to 50 Hz) or semibroadband (flat frequency response, from 0.2 to 40 Hz) sensors and M L1C is the magnitude computed starting from the vertical short-period recordings. [1]: /embed/graphic-1.gif

A Study of the Local Magnitude Scale in the Basin of Mexico: Mutually Consistent Estimates of log A0 and Ground-Motion Scaling

A study of the local magnitude scale in the Basin of Mexico is presented using about 1810 synthetic Wood-Anderson peak amplitudes at 13 stations from the Valley of Mexico seismic network (rsvm). Data from all available rsvm earthquakes located in central Mexico were used, and the events within the Basin of Mexico were relocated. Parametric and nonparametric expressions for the log A ( r ) curve were determined and corrected to Richter’s reference level, which was obtained after the linear separation of source, site, and distance terms. The parametric attenuation function is best expressed as log A ( r ) = −0.48 log ( r /100) − 0.0018 ( r − 100) − 3. The results of the attenuation curve are consistent with ground-motion scaling studies (geometrical spreading, Qs , and stochastic simulations): the Basin of Mexico is characterized by a complex attenuation model at short distances, but the general trend is that of a highly attenuative region. We observe that the coda magnitude has been underestimating the excitation levels of local events in the Basin of Mexico and may require station corrections to improve its accuracy. Our study will provide local magnitudes to the seismicity catalog in the Basin of Mexico, thus contributing to more successful earthquake hazard analyses in the most important and heavily populated region of Mexico.

Local Magnitude in Northeastern Italy

We consider a large data set of waveforms (5053 recordings for the horizontal components and 9933 for the vertical) to calibrate a local magnitude scale valid for northeastern Italy in the hypocentral distance range of 10–250 km. The data refer to 1096 events occurring in the period January 1995 to December 2002 in the area between 44.4° and 47.6° N and 10.1° and 15.7° E. For each available signal, we simulate the corresponding Wood-Anderson seismogram. Using the geometrical mean of the two horizontal peaks, we simultaneously estimate the distance attenuation, the station correction terms, and the magnitude of the events. Estimations are obtained with the standard parametric approach accounting for the geometrical spreading and the anelastic attenuation, and by using two different methods of nonparametric regression, based on a stepwise-linear approximation and on kernel functions, respectively. The nonparametric solutions agree and show stronger attenuation in the first 70 km than what was found in central California. Furthermore, a noticeable discontinuity occurs around 100 km, because later phase arrivals. Such discontinuity is not fitted well by the parametric curve. We investigate the effect of magnitude-dependent attenuation on magnitude estimation. We also account for the bias induced by data truncation (i.e., data loss due to the limited dynamics of the instruments). The analysis shows that at short distances the standard, magnitude-independent attenuation model underestimates high magnitudes (e.g., by 0.49 units at 10 km for magnitude 5.5) and overestimates low magnitudes (e.g., by 0.17 units at 10 km for magnitude 1.5, 0.34 units if we extrapolate the relation down to magnitude 0.5). Finally, various factors suggest one should adopt the solution proposed in the literature to redefine the M L scale such that M L 3 corresponds to 10 mm of motion on a Wood-Anderson instrument at 17 km hypocentral distance.

Jabuka island (Central Adriatic Sea) earthquakes of 2003

We present analyses of one of the strongest earthquake sequences ever recorded within the Adriatic microplate, which occurred near the Jabuka island in the very centre of the Adriatic Sea. The mainshock (29 March 2003, 17:42, M L=5.5) was preceded by over 150 foreshocks, and followed by many aftershocks, over 4600 of which were recorded on the closest station HVAR (about 90 km to the east). As the epicentre was in the open sea and due to the absence of nearby stations, we were able to confidently locate only 597 events. Hypocentral locations were computed by a grid-search algorithm after seven iterations of refining hypocentres and adjusting station corrections. Epicentres lie in a well-defined area of about 300 km 2, just to the W and NW of the Jabuka island. The vertical cross-sections reveal that hypocentres dip to the NE, closely matching faults from the Jabuka-Andrija fault system, as identified on the available reflection profiles in the area. The fault-plane solution of the main shock based on the first-motion polarity readings agrees well with the CMT solutions and indicates faulting caused by a S–N directed tectonic pressure, on a reverse, dip-slip fault. This is in very good agreement with the seismotectonic framework of the area. These earthquakes are important as they identify the Jabuka-Andrija fault system as an active one, which can significantly influence seismic hazard on the islands in the central Adriatic archipelago and on the Croatian coast between Zadar and Split. Along with several other sequences which occurred in the last two decades, they force us to change our notion of Adria as nearly aseismic, compact and rigid block. In fact, it turns out that recent seismicity of the Central Adriatic Sea is comparable to the seismicity of several well known earthquake-prone areas in the circum-Adriatic region.

Spatial Variations of Local Magnitude in Earthquakes Recorded in Northeastern Spain

Since 1984, a seismic surveillance of the Baix Ebre Region (NE of the Iberian Peninsula) and the La Cerdanya Valley (Eastern Pyrenees) has been performed by the Institut d’Estudis Catalans and the Universitat de Barcelona, by means of four digital stations. An investigation of the apparent magnitude has been carried out to explain the observed anomalous differences between the amplitude of the recorded events and their associated magnitude, as given by the different agencies. Calculating the local magnitude with independent attenuation parameters and comparing the obtained results using the different stations, we obtain differences up to 1.0. These differences are interpreted in terms of very local effects and point out the uselessness of some stations to assign magnitudes of local events. Results emphasize that the effectiveness of the trigger algorithms in digital stations rarely depends on the apparent (local) magnitude. As expected, the magnitude sensitivity is related to the values of the station correction terms, meaning that extra information is required for an optimum surveillance of a local area. A 2-D analysis of the distribution of the differences between the computed and reported magnitude shows no significant azimuthal dependence. This indicates that the obtention of a local magnitude scale with an appropriate station correction term related to epicentral distance, results in an important factor to provide magnitude to local events. The obtained results corroborate previous works done at the same area and emphasis is placed on further analysis directed to obtain a better characterization of the local site influence.

Precise hypocentre relocation of microearthquakes in a high-temperature geothermal field: the Torfajökull central volcano, Iceland

The Torfaj¨okull volcanic system is one of approximately 30 active volcanoes comprising the neovolcanic zones of Iceland. The central volcano of the system is the largest silicic centre in Iceland with a caldera of approximately 12 km diameter. Its high-temperature geothermal system is one of the most powerful in Iceland. Torfaj¨okull is a source of persistent seismicity, where both high- and low-frequency earthquakes occur. To study this microseismicity in detail, a temporary array of 20 broad-band seismic stations was deployed between 2002 June and November. These temporary stations were embedded in the permanent South Iceland Lowland (SIL) network and data from nine adjacent SIL stationswere included in this study.Aminimum one-dimensional (1-D) velocity model with station corrections was computed for earthquake relocation by inverting manually picked P- and S-wave arrival times from events occurring in the Torfaj¨okull volcanic centre and its surroundings. High-frequency earthquakes from the Torfaj¨okull volcanic centre were then relocated calculating a non-linear, probabilistic solution to the earthquake location problem. Subsequently, we correlated the waveforms of these 121 events (∼2000 observations) to define linked events, calculated the relative traveltime difference between event pairs and solved for the hypocentral separation between these events. The resulting high-resolution pattern shows a tighter clustering in epicentre and focal depth when compared with original locations. Earthquakes are mainly located beneath the caldera with hypocentres between 1 and 6 km depth and lie almost exclusively within the geothermal system. A sharp cut-off in seismicity at 3 km suggests either that there is a marked temperature increase or that this is a structural boundary. No seismic activity was observed in the fissure swarms to the northeast (NE) and southwest (SW) of the volcanic centre.

Seismicity of the northern part of the Itoigawa-Shizuoka Tectonic Line

The relationship between recent seismicity and the active fault system was investigated using the precise hypocentral distribution by taking into consideration the heterogeneity of the velocity structure around the northern part of the Itoigawa-Shizuoka Tectonic Line (ISTL), the western boundary of the northern Fossa Magna basin. Since a velocity structure suitable for the ISTL is necessary for precise hypocentral determination, deep seismic reflection and refraction profiling were undertaken in 2002 across the northern part of the ISTL and a reliable velocity model was obtained. The hypocenters were located using the station corrections derived from the velocity model. The obtained hypocenters were found to be distributed in a straight line in an almost north-tosouth direction. The hypocenters, whose lower limit was approximately 15 km, were distributed around a vertical plane. The focal mechanism solutions were primarily strike-slip and the direction of their P-axis was NW-SE to WNW-ESE. The fault planes expected from these mechanisms are consistent with the distribution of the hypocenters. However, they are not consistent with the thrust fault planes of previous large earthquakes, which were inferred from the near-surface feature of the active fault system. This suggests that the recent seismicity cannot be attributed to aftershocks of the previous large earthquakes and is not induced by the movement of the active thrust fault.

Fine‐scale seismic structure of the shallow volcanic crust on the East Pacific Rise at 9°50′N

We use a combination of body wave and interface wave observations from an on‐bottom seismic refraction survey to constrain the fine‐scale seismic structure of the upper crust in a ∼3 × 3 km field area centered on the East Pacific Rise at 9°50′N. We detonated 18 explosive shots (18 sources) in a circular pattern (1.5 km radius) on the rise axis and recorded seismic arrivals with eight ocean bottom seismometers (eight receivers). We observed 30–40 Hz compressional body waves from all shots (144 P waves) and 1–3 Hz Stoneley (interface) waves on a subset of source‐receiver pairs (58 interface waves). Using a station correction inversion, we find that roughly half of the variance in the P wave first‐arrival times results from lateral variations in the thickness of the surface low‐velocity layer (SLVL), a layer of extremely porous lava and basalt breccia with an average P wave velocity of 2.2 km s−1. The SLVL thickness increases from <20 m along the axial summit trough (AST) to ∼120 m at near‐axis lava depocenters, which are not symmetric about the rise axis. Depocenters are located ∼0.5 km to the west and ∼1.5 km to the east of the rise axis. Tomographic inversion of the Stoneley wave first arrivals reveals that shear velocities in the SLVL covary with the layer thickness, exhibiting a similar asymmetric pattern, with shear velocities increasing from ∼320 m s−1 near the AST to ∼520 m s−1 at the near‐axis depocenters. Our analysis demonstrates that the seismic characteristics of the extrusive layer near the rise axis are related primarily to volcanic features and processes. The thickness and velocity of the SLVL are low on the axis and within channel networks that deliver lava flows away from the axis and then increase rapidly at the distal ends of the channels where the lavas are deposited. We find that azimuthal anisotropy exerts only a weak influence on our P wave first‐arrival times, which we model as weak (4%) seismic azimuthal anisotropy in the upper dikes with a fast axis oriented N23°–32°W. We find no evidence for seismic azimuthal anisotropy in the extrusive layer.

Minimum 1D Velocity Model from Local Earthquake Data in the Provence Region, South-Eastern France

We computed a one-dimensional (1D) velocity model and station corrections refering to the Provence region (South-eastern France) by inverting P-wave arrival times recorded on an eight-station local seismic network. Using this velocity model and the program HYPOELLIPSE (Lahr, 1989), we relocated a set of 108 local events. The quality comparison between previous earthquake locations and new relocated shows a good improvement.

Hypocentre determination offshore of eastern Taiwan using the Maximum Intersection method

SUMMARY The maximum intersection (MAXI) method, which derives from the master station method (MSM), determines within a 3-D velocity model the absolute hypocentral location based on observed arrival times. First, the spatial node that better satisfies the arrival time differences computed at all station pairs, plus or minus an error tolerance value (in seconds), is defined as the preliminary hypocentral solution (PRED). Second, because PRED depends neither on the estimate of origin time nor on the residual root mean square (rms), residual outliers are objectively detected and cleaned out from the original data set without any iterative process or weighting. Third, a statistical minimization (residual rms) is conducted in a small domain around the PRED node, which results in a unique FINAL solution. The MAXI method is applied to the determination of earthquake hypocentres (with the proper station correction terms) in the southernmost extremity of the Ryukyu subduction zone, where several dense seismic clusters occur near the seismogenic plate interface. The location of earthquakes, recorded at both the Taiwanese and Japanese networks, is obtained for about a thousand events (between 1992 and 1997). The process uses a detailed 3-D velocity model based on multiple geophysical data sources obtained in the junction area between subduction and collision (east of Taiwan). The earthquake clustering and the significant drop in residual statistics (1.20, 0.80 and 0.35 s, for Taiwanese catalogue, MSM and MAXIM solutions respectively) indicate the accuracy of the method, which can be used to routinely determine absolute hypocentre location based on observed arrival times.

Simultaneous inversion for local earthquake hypocentres, station corrections and 1-D velocity model of the Jamaican crust

A new 1-D velocity model for Jamaica has been obtained from the simultaneous inversion of P- and S-wave travel times of well-solved local earthquakes. The crust is modelled as four flat uniform layers with compressional velocities of 5.15 km/s to 4.5 km depth, 6.30 km/s to 9.5 km, 6.60 km/s to 22.0 km and 7.11 km/s to 30.0 km. The upper mantle P-velocity is 7.81 km/s, which is consistent with upper mantle velocities found elsewhere in the Caribbean Basin. The new model reduces significantly the root mean square misfit of the selected data from 0.128 to 0.112 when compared to the best model previously used. In addition, the model solves events to more acceptable focal depths of 10 to 25 km. Station corrections of −0.20 to +0.06 for P and −0.75 to 0.13 for S-waves relative to the reference station were obtained. The new model is expected to improve Jamaican earthquake locations particularly in obtaining more reliable focal depths, which in turn will enhance the interpretation and understanding of local seismicity and neotectonics.

Intra-plate seismicity in the subducting Philippine Sea Plate, southwest Japan: magnitude–depth correlations

Recent studies of intra-plate seismicity have focused on detailed earthquake locations and potential hypotheses for intermediate depth events. Three main competing hypotheses for intra-plate events are: (1) transformational faulting; (2) shear melt instabilities; and (3) dehydration reactions. Recent work, using precise hypocenters and thermal-petrological models, indicates that the spatial distribution of intra-plate events coincides with dehydration reactions. These studies have mainly focused on the location of events and the proposed reactions with little attention given to the magnitude distribution of events. Presented here is a magnitude–depth correlation associated with intra-plate events for southwest Japan. Earthquake hypocenters were obtained from relocating a data set consisting of Japan Meteorological Agency (JMA) seismicity data and Japan University Network Earthquake Catalog (JUNEC). Data from 1985 to 1996 were combined and relocated using a standard location routine (HypoInverse) that allows regional variations in velocity models and station corrections. HypoInverse results show a well defined plane of seismicity associated with the subducting Philippine Sea Plate. The data were further relocated using a double-difference relocation algorithm (HypoDD). HypoDD results show a magnitude–depth correlation for Kii peninsula and northern Kyushu. Histogram plots indicate that events near the plate interface were generally magnitude 3 or less, whereas most large events were located within the lower crust or possibly the subducting oceanic mantle. This observation also appears to hold true for intra-plate events beneath Tohoku (Japan trench) and Hokkaido (Kuriles), but is not clearly observed beneath the Tokai region. A dehydration reaction model can well explain the magnitude–depth correlation.

Relocation of the 2001 Earthquake Sequence in Aegion, Greece

The western part of the Corinth Gulf attracts attention because of its seismically active complex fault system and considerable seismic hazard. Close to the city of Aegion, damaged by the ML 6.2 earthquake of 1995, a sequence of small earthquakes occurred from February to May 2001. The sequence, comprising 171 events of ML 1.8 to 4.7, was recorded by a short-period network of the University of Patras, PATNET. As most stations have single component-recording, the S-wave arrival time readings were scarce. A subset of 139 events was recorded by at least 5 stations, and in this study we limit ourselves just to that sub-set. A preliminary location is performed by a standard linearized kinematic approach, with several starting depths and crustal models. Then the mainshock is re-located, and finally it is used as a master event to locate the remaining events. The mainshock relocation is performed by a systematic 3D grid search, and the trade-off between depth and origin time is eliminated by a special procedure, the so-called station difference (SD) method. In the SD method, instead of inverting arrival times directly, their intra-station differences are employed. The station corrections, determined from the master event, are also used. As a result, the sub-set is imaged as a relatively tight cluster, occupying space of about 5 by 5 km horizontally and 10 km vertically, with the mainshock inside (at a depth of 7 km). The results should be interpreted with caution, mainly as regards the „absolute“ depth position of the cluster. A more accurate location would require a local network with both P and S readings.

Site Response from Incident Pnl Waves

We developed a new method of determining site response and amplification for use in hazard analysis and station corrections. The method employs the conversion of P to S energy beneath a soft-rock station, which results in complex receiver functions that are frequency and amplitude dependent. At low frequencies ( 0.5 Hz) can be normalized to these low-frequency levels to quantify the amount of high-frequency amplification. Our results agree with previous studies of the Los Angeles Basin and provide the means of calibrating station responses at high frequencies.

A SOURCE PARAMETERS STUDY OF THE AFTERSHOCK SEQUENCE OF THE KOZANI- GREVENA 1995 EARTHQUAKE BASED ON ACCELERATION RECORDS

Strong motion recordings of the May 13, 1995 Mw=6.6, earthquake sequence that occurred in the Kozani-Grevena region (Western Macedonia, Greece) have been analyzed for the determination of their source parameters. The data set for this study comes from a temporarily deployed accelerograph network and the source parameters using the shear-wave displacement spectra have been estimated. For this estimation the spectral records have been corrected for the site effects and for the propagation path (geometrical spreading and anelastic attenuation). The magnitude of each event was also re-calculated by estimating appropriate station corrections. The derived relationships arelogMo =(1.43 ±0.09) M, + (16.92 ± 0.29), 2.0 < ML< 5.0 (1)logfc = (-0.56± 0.08) · ML + (2.52 + 0.29), 2.0 < ML< 5.0 (2)logM0 = (-2.20 + 0.08) · log fc + (23.16 ± 0.84), 0.6 < fc < 10.0 (3)The near-surface attenuation parameter κ0 has also been determined for the strong motion stations sites. These values of κ0 are in good agreement with those of Margaris and Boore (1998) for the geological formation on which each station was positioned. The obtained source parameters are in good agreement with those from previous studies for the Aegean region.

Location Robustness of Earthquakes in Sichuan Province, China

The Sichuan Province in China is located in the middle section of a large north-south seismicity belt. The Xianshuihe Fault striking northwest-southeast, the Longmenshan east fault striking north-south, and the Anninghe Fault striking north-south together form a “Y” shape. Sichuan is located in the east margin of the Qinghai-Tibet crustal block. More than 8,000 earthquakes with magnitude 0.0 occur in Sichuan every year, and 2,000 of these can be located (Han and Lu, 2001). Since 26 B.C., 240 strong earthquakes have occurred in Sichuan with magnitude ML ≥ 5.0. Fifty of these events had ML 6.0-6.9, and 19 of them had ML 7.0-7.9. These events are distributed mainly along several faults and form seismic zones such as the Xianshuihe, the Songpan-Longmenshan, the Anninghe-Zemuhe, the Jinshajiang, and the Mingshan-Mabian-Zhaotong seismic zones (Han and Lu, 2001). A map of earthquake epicenters in Sichuan Province for events with magnitude greater than 2.5 during the period 1970 through 30 April 2003 is shown in Figure 1. In general, seismic epicenters in Sichuan Province show a diffuse spatial pattern rather than sharp linear alignment. The complex tectonic structure of the region, with crustal thickness varying between about 40 and 65 km (Ren et al. , 1987; Beghoul et al. , 1993; Vincent, 2000) and great sediment thickness over the Sichuan basin, creates a three-dimensional seismic velocity structure that affects the calculated locations. In this study we selected from a group of 1D velocity models the one that best fit the travel times of three ground-truth (GT) events from Sichuan, China. Based on this 1D velocity model, a set of station corrections for 106 stations in Sichuan area was generated. For the purpose of locating earthquakes quickly and within the desired accuracy, we modified the software programs Hypoinverse (Hypo) …

Local attenuation in western Nagano, central Japan, estimated from seismograms recorded in three boreholes

[1] Seismic attenuation parameters three borehole stations in Western Nagano, Central Japan. Based on the difference in attenuation between combinations of two boreholes, the local attenuation (QP−1) is 1.23 × 10−3 s in the frequency range of 80 to 250 Hz. Obtained QP−1 is consistent with the trend estimated by grid search with the assumption of the ω2 source model. Even though the boreholes are located on the fracture area of the 1984 Nagano-Ken-Seibu Earthquake, the estimated QP−1 is low. The result is not completely consistent with a previous study, and a result of QS−1 obtained from the twofold spectral ratio, but the difference can be interpreted as the difference of frequency ranges. The differences of station corrections of attenuation between boreholes are much smaller than the difference between borehole and surface, and the result shows the advantage of borehole seismometers.

Empirical path and station corrections for surface-wave magnitude,Ms, using a global network

SUMMARY One of the most commonly used methods of estimating the size of shallow earthquakes is the surface-wave magnitude scale, Ms. It has been known since the inception of the scale that the Ms of a seismic disturbance recorded at different stations can vary due to path propagation and station terms, but it has been difficult to quantify these variations on a global basis because, until recently, there has been no attempt to record Ms in a uniform way on a worldwide network. Here we use Ms measurements within the Reviewed Event Bulletin (REB) of the International Data Centre (IDC) which is produced to help monitor compliance with the Comprehensive Nuclear Test Ban Treaty (CTBT). The IDC collects waveforms from the as yet incomplete seismological network of the International Monitoring System (IMS) and Ms is then measured using an automatic algorithm, providing a unique set of surface-wave observations. In this paper we show that the Ms values reported in the REB show distinct geographic variations. We present preliminary attempts to model these observations in terms of a set of station corrections and a 2-D model which can be used to predict path corrections for Ms. The model shows a striking correlation with the known tectonic regions of the Earth, suggesting that this type of data set may provide a valuable tool for investigating shallow Earth structure. After modelling, the residuals show some systematic patterns which cannot be explained using the model basis we have used. These residuals are presumably due either to source radiation patterns or to complicated propagation effects not allowed for in the model, such as refraction at ocean-continent boundaries. From the CTBT monitoring viewpoint, Ms station and path corrections are valuable because they should improve the effectiveness of event screening using mb: Ms. However, we recommend that implementation of such corrections should wait until the full seismological network of the IMS is in place, and until a fuller understanding of path effects on Ms is available. In particular, it is vital that path propagation effects do not distort station corrections, and that earthquake radiation patterns should not influence corrections which may be applied to surface-waves from suspected explosions.