Seismic quiescence precursory to a past and a future Kurile island earthquake

A systematic search was made for seismicity rate changes in the segment of the Kurile island arc from 45°N to 53°N by studying the cumulative seismicity of shallow (h100 km) earthquakes within 11 overlapping volumes of radius 100 km for the time period 1960 through beginning of 1978. We found that in most parts of this island arc and most of the time the seismicity rate as obtained from the NOAA catalogue and not excluding any events is fairly constant except for increased seismicity in the mid 1960s in the southern portion due to the great 1963 mainshock there, and for seismicity quiescence during part of the time period studied within two well defined sections of the arc. The first of these is a volume of 100 km radius around a 1973 (M _s =7.3) mainshock within which the seismicity rate was demonstrated at the 99% confidence level to have been lower by 50% during 2100 days (5.75 years) before this mainshock. The second volume of seismic quiescence coincides with the 400 km long north Kuriles gap. In this gap the seismicity rate is shown (at the 99% confidence level) to be lower by 50% from 1967 to present (1978), in comparison with the rate within the gap befor 1967, as well as with the rate surrounding the gap. We propose that the anomalously low seismicity rate within the Kuriles gap is a precursor to a great earthquake, the occurrence time of which was estimated by the following preliminary relation between precursory quiescence time and source dimensionT=190L ^0.545. We predict that an earthquake with source length of 200–400 km (M>8) will occur along the north Kurile island arc between latitude 45.5°N and 49.2°N at a time between now and 1994.     

The 1986 Kermadec earthquake and its relation to plate segmentation

To evaluate the tectonic significance of the October 20, 1986 Kermadec earthquake (M _w =7.7), we performed a comprehensive analysis of source parameters using surface waves, body waves, and relocated aftershocks. Amplitude and phase spectra from up to 93 Rayleigh waves were inverted for centroid time, depth, and moment tensor in a two-step algorithm. In some of the inversions, the time function was parameterized to include information from the body-wave time function. The resulting source parameters were stable with respect to variations in the velocity and attenuation models assumed, the parameterization of the time function, and the set of Rayleigh waves included. The surface wave focal mechanism derived (=275°, =61°, =156°) is an oblique-compressional mechanism that is not easy to interpret in terms of subduction tectonics. A seismic moment of 4.5×10²⁰ N-m, a centroid depth of 45±5 km, and a centroid time of 13±3 s were obtained. Directivity was not resolvable from the surface waves. The short source duration is in significant contrast to many large earthquakes.We performed a simultaneous inversion ofP andSH body waves for focal mechanism and time function. The focal mechanism agreed roughly with the surface wave mechanism. Multiple focal mechanisms remain a possibility, but could not be resolved. The body waves indicate a short duration of slip (15 to 20 s), with secondary moment release 60s later. Seismically radiated energy was computed from the body-wave source spectrum. The stress drop computed from the seismic energy is about 30 bars. Sixty aftershocks that occurred within three months of the mainshock were relocated using the method of Joint Hypocentral Determination (JHD). Most of the aftershocks have underthrusting focal mechanisms and appear to represent triggered slip on the main thrust interface. The depth, relatively high stress drop, short duration of slip, and paucity of true aftershocks are consistent with intraplate faulting within the downgoing plate. Although it is not clear on which nodal plane slip occurred, several factors favor the roughly E-W trending plane. The event occurred near a major segmentation in the downgoing plate at depth, near a bend in the trench, and near a right-lateral offset of the volcanic are by 80 km along an E-W direction. Also, all events in the region from 1977 to 1991 with CMT focal mechanisms similar to that of the Mainshock occurred near the mainshock epicenter, rather than forming an elongate zone parallel to the trench as did the aftershock activity. We interpret this event as part of the process of segmentation or tearing of the subducting slab. This segmentation appears to be related to the subduction of the Louisville Ridge, which may act as an obstacle to subduction through its buoyancy.     

Rupture process of the 19 August 1992 Susamyr, Kyrgyzstan, earthquake

The Susamyr earthquake of August 19, 1992 in Kyrgyzstan is one of the largest events (M_s = 7.4, M_b = 6.8) of this century in this region of Central Asia. We used broadband and long period digital data from IRIS and GEOSCOPE networks to investigate the source parameters, and their space-time distribution by modeling both body and surface waves. The seismic moment (M₀ = 6.8 × 10¹⁹ N m) and the focal mechanism were determined from frequency-time analysis (FTAN) of the fundamental mode of long period surface waves (100–250 s). Then, the second order integral moments of the moment-rate release were estimated from the amplitude spectra of intermediate period surface waves(40–70 s). From these moments we determined a source duration of 11–13 s, major and minor axes of the source of 30 km and 10–22 km, respectively; and an instant centroid velocity of 1.2 km/s. Finally, we performed a waveform inversion of P and SH waves at periods from 5–60 s. We found a source duration of 18–20 s, longer than the integral estimate from surface wave amplitudes. All the other focal parameters inverted from body waves are similar to those obtained by surface waves ( = 87° ± 6°, = 49° ± 6°, = 105° ± 3°, h = 14 ± 2 km, and M₀ = 5.8 ± 0.7 × 10¹⁹ N m). The initial rupture of this shallow earthquake was located at the south-west border of Susamyr depression in the western part of northern Tien Shan. A finite source analysis along the strike suggests a westward propagation of the rupture. The main shock of this event was preceded 2 s earlier by small foreshock. The main event was almost immediately followed by a very strong series of aftershocks. Our surface and body wave inversion results agree with the general seismotectonic features of the region.     

Instantaneous photochemical rates in the global stratosphere

Starting with the average actual distribution of ozone (Dütsch [15]) and temperature in the stratosphere, we have calculated the solar intensity as a function of wavelength and the instantaneous rates (molecules cm^–3 sec^–1) for each Chapman reaction and for each of several reactions of the oxides of nitrogen. The calculation is similar to that ofBrewer andWilson [5]. These reaction rates were calculated independently in each volume element in spherical polar coordinates defined by R=1 km from zero to 50, =5° latitude, and ø=15° longitude (thus including day and night conditions). Calculations were made for two times: summer-winter (January 15) and spring-fall (March 22). As input data we take observed solar intensities (Ackerman [1]) and observed, critically evaluated. constants for elementary chemical and photochemical reactions; no adjustable parameters are employed. (These are not photochemical equilibrium calculations.) According to the Chapman model, the instantaneous, integrated, world-wide rate of formation of ozone from sunlight is about five times faster than the rate of ozone destruction, and locally (lower tropical stratosphere) the rate of ozone formation exceeds the rate of destruction by a factors as great as 1000. The global rates of increase of ozone are more than 50 times faster thanBrewer andWilson's [5] estimate of the average annual transfer rate of ozone to the troposphere. The rate constants of the Chapman reactions are believed to be well-enough known that it is highly improbable that these discrepancies are, due to erroneous rate constants. It is concluded that something else besides neutral oxygen species is very important in stratospheric ozone photochemistry. The inclusion of a uniform concentration of the oxides of nitrogen (NO_x as, NO and NO₂) averaging 6.6×10^–9 mole fraction gives a balance between global ozone formation and destruction rates. The inclusion of a uniform mole fraction of NO_x at 28×10^–9 also gives a global balance. These calculations support the hypethesis (Crutzen [10],Johnston [24]) that the oxides of nitrogen are the most important factor in the global, natural ozone balance. Several authors have recently evaluated the natural source strength of NO_x in the stratosphere; the projected fleets of supersonic transports would constitute an artificial source of NO_x about equal to the natural value, thus promising more or less to double an active natural stratospheric ingredient.     

Thermal Structure of the Ionian Slab

We propose a thermal model of the subducting Ionian microplate. The slab sinks in an isothermal mantle, and for the boundary conditions we take into account the relation between the maximum depth of seismicity and the thermal parameter L_th of the slab, which is a product of the age of the subducted lithosphere and the vertical component of the convergence rate. The surface heat-flux dataset of the Ionian Sea is reviewed, and a convective geotherm is calculated in its undeformed part for a surface heat flux of 42 mW m^–2, an adiabatic gradient of 0.6 mK m^–1, a mantle kinematic viscosity of 10¹⁷ m² s^–1 and an asthenosphere potential temperature of 1300°C. The calculated temperature-depth distribution compared to the mantle melting temperature indicates the decoupling limit between lithosphere and asthenosphere occurs at a depth of 105 km and a temperature of 1260°C. A 70–km thick mechanical boundary layer is found. By considering that the maximum depth of the seismic events within the slab is 600 km, a L_th of 4725 km is inferred. For a subduction rate equal to the spreading rate, the corresponding assimilation and cooling times of the microplate are about 7 and 90 Myr, respectively. The thermal model assumes that the mantle flow above the slab is parallel and equal to the subducting plate velocity of 6 cm yr^–1, and ignores the heat conduction down the slab dip. The critical temperature, above which the subduced lithosphere cannot sustain the stress necessary to produce seismicity, is determined from the thermal conditions governing the rheology of the plate. The minimum potential temperature at the depth of the deepest earthquake in the slab is 730°C.     

Partial breaking of a mature seismic gap: The 1987 earthquakes in New Britain

To better understand the mechanics of subduction and the process of breaking a mature seismic gap, we study seismic activity along the western New Britain subduction segment (147°E–151°E, 4°S–8°S) through earthquakes withm _b 5.0 in the outer-rise, the upper area of subducting slab and at intermediate depths to 250 km, from January 1964 to December 1990. The segment last broke fully in large earthquakes of December, 28, 1945 (M _s =7.9) and May 6, 1947 (M _s =7.7.), and its higher seismic potential has been recognized byMcCann et al., (1979). Recently the segment broke partially in two smaller events of February, 8, 1987 (M _s =7.4) and October 16, 1987 (M _s =7.4), leaving still unbroken areas.We observe from focal mechanisms that the outer-rise along the whole segment was under pronounced compression from the late 60's to at least October 1987 (with exception of the tensional earthquake of December 11, 1985), signifying the mature stage of the earthquake cycle. Simultaneously the slab at intermediate depths below 40 km was under tension before the earthquake of October 16, 1987. That event, with a smooth rupture lasting 32 sec, rupture velocity of 2.0 km/sec, extent of approximately 70 km and moment of 1.2×10²⁷ dyne-cm, did not change significantly the compressive state of stress in the outer-rise of that segment. The earthquake did not fill the gap completely and this segment is still capable of rupturing either in an earthquake which would fill the gap between the 1987 and 1971 events, or in a larger magnitude event (M _s =7.7–7.9), comparable to earthquakes observed in that segment in 1906, 1945 and 1947.     

Coseismic slip in the 1964 Prince William Sound earthquake: A new geodetic inversion

The 1964 Prince William Sound earthquake (March 28, 1964;M _w =9.2) caused crustal deformation over an area of approximately 140,000 km² in south central Alaska. In this study geodetic and geologic measurements of this surface deformation were inverted for the slip distribution on the 1964 rupture surface. Previous seismologic, geologic, and geodetic studies of this region were used to constrain the geometry of the fault surface. In the Kodiak Island region, 28 rectangular planes (50 by 50 km each) oriented 218°N, with a dip varying from 8^o nearest the Aleutian trench to 9^o below Kodiak Island, define the rupture surface. In the Prince William Sound region 39 planes with variable dimensions (40 by 50 km near the trench, 64 by 50 km inland) and orientation (218°N in the west and 270°N in the east) were used to approximate the complex faulting. Prior information was introduced to constrain offshore dip-slip values, the strike-slip component, and slip variation between adjacent planes. Our results suggest a variable dip-slip component with local slip maximums occurring near Montague Island (up to 30 m), further to the east near Kayak Island (up to 14 m), and trenchward of the northeast segment of Kodiak Island (up to 17m). A single fault plane dipping 30°NW, corresponding to the Patton Bay fault, with a slip value of 8 m modeled the localized but large uplift on Montague Island. The moment calculated on the basis of our geodetically derived slip model of 5.0×10²⁹ dyne cm is 30% less than the seismic moment of 7.5×10²⁹ dyne cm calculated from long-period surface waves (Kanamori, 1970) but is close to the seismic moment of 5.9×10²⁹ dyne cm obtained byKikuchi andFukao (1987).     

Comparison of a complex rupture model with the precursor asperities of the 1975 HawaiiM s-7.2 earthquake

A simplified multiple source model was constructed for the 1975 HawaiiM _s=7.2 earthquake by matching synthetic signals with three component accelerograms at two stations located approximately 45 km from the epicenter. Six major subevents were identified and located approximately. The signals of these are larger by factors of 1.4 to 3.2 than that of theM _L=5.9 foreshock which occurred 70 minutes before the main rupture and also triggered the SAM-1 recorders at the two stations. Dividing the rupture length (40 km) by the duration of strong ground shaking ( 50 sec) an, average rupture velocity of 0.8 km/sec (about 25% of S-velocity) is obtained. Thus it is likely that the rupture stopped between subevents. The approximate epicenters of the 6 major subevents, and of the foreshock, support the hypothesis that they were located in high stress asperities which rupture during the main shock, except for the last events which is interpreted as a stopping phase generated at a barrier. These asperities have been previously defined on the basis of differences in the precursor pattern before the mainshock. Thus, it appears that both the details of the precursors and of the main rupture depended critically on the heterogeneous tress distribution in the source volume. This suggests that main rupture initiation points and locations of high rupture accelerations may be identified before the mainshock occurs, based on precursor anomaly patterns. A satisfactory match of synthetic signals with the observations could be obtained only if the aximuth of the fault plane of subevents was rotated from N60°E to N90°E and back to N30°E. These orientations are approximately parallel to the nearest Kilauea rift segments. Hence the slip directions and greatest principal stresses were oriented perpendicular to the rifts everywhere. From this analysis and other work, it is concluded that this fault surface consisted of three types of segments with different strength: hard asperities (radius 5 km), soft but brittle segments between the asperities (radius 5 km), and a viscous half (10×40 km) which slipped during the mainshock, but where microearthquakes and aftershocks are not common.     

Source characteristics of the 1992 Nicaragua tsunami earthquake inferred from teleseismic body waves

We analyzed the broadband body waves of the 1992 Nicaragua earthquake to determine the nature of rupture. The rupture propagation was represented by the distribution of point sources with moment-rate functions at 9 grid points with uniform spacing of 20 km along the fault strike. The moment-rate functions were then parameterized, and the parameters were determined with the least squares method with some constraints. The centroid times of the individual moment-rate functions indicate slow and smooth rupture propagation at a velocity of 1.5 km/s toward NW and 1.0 km/s toward SE. Including a small initial break which precedes the main rupture by about 10 s, we obtained a total source duration of 110 s. The total seismic moment isM _o =3.4×10²⁰ Nm, which is consistent with the value determined from long-period surface waves,M _o =3.7×10²⁰ Nm. The average rise time of dislocation is determined to be 10 s. The major moment release occurred along a fault length of 160 km. With the assumption of a fault widthW=50 km, we obtained the dislocationD=1.3 m. From andD the dislocation velocity isD=D/0.1 m/s, significantly smaller than the typical value for ordinary earthquakes. The stress drop =1.1 MPa is also less than the typical value for subduction zone earthquakes by a factor of 2–3. On the other hand, the apparent stress defined by 2E _s /M _o , where andE _s are respectively the rigidity and the seismic wave energy, is 0.037 MPa, more than an order of magnitude smaller than . The Nicaragua tsunami earthquake is characterized by the following three properties: 1) slow rupture propagation; 2) smooth rupture; 3) slow dislocation motion.     

How deep was the Dudley (UK) earthquake of 22 September 2002?

On 22 September 2002, the largest UK earthquake (m_b4.3) of the last 10 years occurred near the town of Dudley in the West Midlands. Here we determine the earthquake focal mechanism and depth using data from stations at regional and teleseismic distances. Short-period teleseismic seismograms are interpreted in terms of P and surface reflections pP and sP. This analysis suggests that the source depth is deeper than the 9.7 km initially determined by the British Geological Survey (BGS). The relative amplitude method is applied to four teleseismic seismograms to support our interpretation of the surface reflections, and constrain the focal mechanism. Our preferred focal mechanism, a near vertical strike-slip with _s = 94°, = 88° and = –179°, is in reasonable agreement with a moment tensor determined by the Swiss Seismological Service. Synthetic regional surface wave seismograms match the observed seismograms for a model focal depth of 19.5 (±3.0) km and scalar moment, M₀, of 3.2 × 10¹⁵ N m. Our results emphasize that due to the well-known trade-off between depth and M₀ from inversions of long period (0.02–0.1 Hz) surface waves, it is preferable to combine long- and short-period data to constrain reliably the depth and hence estimate M₀. Our focal mechanism and depth are further validated by generating short-period synthetic seismograms that match the observations.     

Continental-array measurements ofP-wave velocities in the mantle

Summary A velocity model for theP-wave is obtained for the entire mantle, by inversion ofdT/d-data from travel times of 473 earthquakes and explosions registered at 13 seismological observatories in Sweden and Finland. As a check, the travel times resulting from this model are found to be in good agreement with our observed travel times. The observations are distributed over the entire azimuth range with respect to the recording stations. The effect of the core upon the observeddT/d-values is significant from about 95° distance outwards. The estimate of the core radius from this model is 3480 km which corresponds to about 97° of epicentral distance. The resulting model has sudden changes inP-velocity gradient at about 65, 115, 400, 780, 970, 1100, 1350, 2320, 2600 and 2710 km depth, which correspond roughly to 10°, 14°, 19.5°, 29°, 40°, 46°, 54°, 80°, 86° and 90° epicentral distance, respectively.On leave from Geophysical Institute, Tehran University, Tehran, Iran     

Measurement of lower-stratospheric aerosol by a balloon-borne photometer

Height distribution of the stratospheric aerosol extinction coefficient was measured in the altitude range 10 to 20 km by a balloon-borne multi-color sunphotometer in May 1978. It is demonstrated that detailed structures of the distribution of stratospheric aerosol can be remotely measured by the solar occultation method as well as by lidar andin situ particle counter observations. In the aerosol layer appearing at 18 km altitude the extinction coefficient at 800–1000 nm wavelength reached to 3×10^–7 m^–1, which was reasonable compared with lidar observations. Wavelength dependence of the aerosol optical depth was crudely estimated to be proportional to ^–1.5.     

Paleomagnetism of gabbros of the Early Proterozoic Blachford Lake Intrusive Suite and the Easter Island dyke,Great Slave Lake,NWT: Possible evidence for the earliest continental drift

The Caribou Lake gabbro, part of the Blachford Lake Intrusive Suite accurately dated at –2186±10 mA, has a predominantNW–/SE+ magnetization with a mean, irrespective of sign, ofD=119°,I=50°, ₉₅=5° and a palaeopole 14°N, 064°W,A ₉₅=5°; it has not proved possible to determine if the magnetization is primary. The Easter Island dyke, less well-dated in the range –2200 to –2500 Ma, has a predominantWNW+ magnetization, whose mean, when corrected for an 8° tilt, isD=288°,I=46°, ₉₅=5 and palaeopole is 32°S, 2°W,A ₉₅=5°; the magnetization is probably primary. A vertical magnetization (D), not significantly different from the present field, occurs sporadically in both units and is considered to be Late Phanerozoic in age. Palaeopoles from the Caribou Lake gabbro and the Easter Island dyke, together with those already known from Early Proterozoic intrusives of the Archaean Slave Structural Province, roughly define a swath (the Slave Track) which maps the motion of the Slave Province relative to the geomagnetic axis during this interval. The corresponding array of palaeopoles (the Superior Track) from the Superior Structural Province does not fall in the same place. Hence it would appear that Slave and Superior were not in their present relative positions in the Early Proterozoic in disagreement with arguments that have been made for a fixed supercontinent during much of the Proterozoic. Mid-Proterozoic paleomagnetic signatures indicate that Slave and Superior had assumed their present relative position by about –1750 mA. These Early Proterozoic relative motions are the earliest for which there is palaeomagnetic evidence.Earth Physics Branch Contribution No. 1111.     

Pressure enhancement associated with meridional flow in high-speed solar wind: possible evidence for an interplanetary magnetic flux rope

A sizable total-pressure (magnetic pressure plus kinetic pressure) enhancement was found within the high-speed wind stream observed by Helios 2 in 1976 near 0.3 AU. The proton density and temperature and the magnetic magnitude simultaneously increased for about 6 h. This pressure rise was associated with a comparatively large southward now velocity component (with V_z – 100 km · s^–1) and magnetic-field rotation. The pressure enhancement was associated with unusual features in the electron distribution function. It shows a wide angular distribution of electron counting rates in the low-energy (57.8 eV) channel, while previous to the enhancement it exhibits a wide angular distribution of electron count rate in the high-energy (112, 221 and 309 eV) channels, perhaps indicating the mirroring of electrons in the converging field lines of the background magnetic field. These fluid and kinetic phenomena may be explained as resulting from an interplanetary magnetic flux rope which is not fully convected by the flow but moves against the background wind towards the Sun.     

Low CodaQcin the Epicentral Region of the 2001 Bhuj Earthquake ofMw7.7

On 26 January, 2001 (03:46:55,UT) a devastating intraplate earthquake of M_w 7.7 occurred in a region about 5 km NW of Bhachau, Gujarat (23.42°N, 70.23°E). The epicentral distribution of aftershocks defines a marked concentration along an E-W trending and southerly dipping (45°) zone covering an area of (60 × 40) km². The presence of high seismicity including two earthquakes of magnitudes exceeding 7.7 in the 200 years is presumed to have caused a higher level of shallow crustal heterogeneity in the Kutch area; a site lying in the seismic zone V (zone of the highest seismicity for potentially M8 earthquakes) on the seismic zoning map of India. Attenuation property of the medium around the epicentral area of the Bhuj earthquake covering a circular area of 61,500 km² with a radius of 140 km is studied by estimating the coda-Q_c from 200 local earthquakes of magnitudes varying from 3.0–4.6. The estimated Q₀ values at locations in the aftershock zone (high seismicity) are found to be low in comparison to areas at a distance from it. This can be attributed to the fact that seismic waves are highly scattered for paths through the seismically active and fractured zone but they are well behaved outside the aftershock zone. Distribution of Q₀ values suggests that the local variation in Q₀ values is probably controlled by local geology. The estimated Q₀ values at different stations suggest a low value of Q=(102 ± 0.80)^*f^{(0.98 ± 0.02)} indicating an attenuative crust beneath the entire region. The frequency-dependent relation indicates a relatively low Q_c at lower frequencies (1–3 Hz) that can be attributed to the loss of energy due to scattering attenuation associated with heterogeneities and/or intrinsic attenuation due to fluid movement in the fault zone and fluid-filled cracks. The large Q_c at higher frequencies may be related to the propagation of backscattered body waves through deeper parts of the lithosphere where less heterogeneity is expected. Based on the attenuation curve estimated for Q₀=102, the ground acceleration at 240 km distance is 13% of 1 g i.e., 0.13 g agreeing well with the ground acceleration recorded by an accelerograph at Ahmedabad (0.11 g). Hence, it is inferred that the Q₀ value obtained from this study seems to be apt for prediction of ground motion for the region.     

Flow directions in ash-flow tuffs: a comparison of geological and magnetic susceptibility measurements, Tshirege member (upper Bandelier Tuff), Valles caldera, New Mexico, USA

This study concludes that the elongation axis (K ₁) of the ellipsoid of anisotropic magnetic susceptibility (AMS) is a suitable proxy for flow axis in ashflow tuffs. 153 oriented samples (176 specimens) were studied from 18 sites in the 1.1 Ma Tshirege member of the Bandelier Tuff. These sites are distributed around the Valles caldera at distances of 5–25 km outside of the rim.K ₁ axes correlate well with postulated radial flow axes at 13 sites.K ₁ also agrees with measured geological flow indicators, mainly imbricated larger clasts, at 7 sites. At 2 of the 5 sites where significant disagreement is seen between theoretical radial flow directions and measuredK ₁ axes, theK ₁ axes correspond well with geological flow indicators, indicating that the divergence of flow from the predicted radial flow pattern is real. Two major topographic buttresses are suggested as the cause of flow divergence for the Tshirege ash flows: the San Pedro buttress northwest of the caldera, and the San Miguel buttress in the southeast. In situK ₁ axes plunge about 7° toward the source at two-thirds of the sites; therefore the plunge ofK ₁ is a plausible in situ indicator for thedirection of flow. Multiple flow zones in sections of several meters thickness indicate changes of flow direction that are both rapid and large during ash-flow emplacement. These observations raisre the question of how best to represent mean flow directions in ash-flow sheets: by eigenvector methods, by vector-sum methods, or by modes. A method for measuring imbrication of larger clasts using apparent dips in vertical joints is outlined. Imbrication, determined in this way at one-third of the sites, dips toward the source, i.e., up-flow. The minimum (K ₃) axis of the AMS ellipsoid correlates with the flow foliation rather than with the larger clast imbrication. The flow axes of ash flows correspond with theK ₁ axes, not with the declination ofK ₃ axes as suggested by some authors. Initial dip of the sampled ash flows is not large and does not affect the paleomagnetic remanence direction, which is reversed with a mean ofD=173.5°,I=-38.4°, ₉₅=3.4°N=18. This mean is not different at the 95% confidence level from that of earlier workers. The mean pole, at 098.0°E, 74.8°N,A ₉₅=3.3°,N=18, is about 15° far-sided relative to the expected time-averaged geomagnetic pole, suggesting a history of emplacement too short to adequately average secular variation.     

Regional focal mechanisms for earthquakes in the Aegean area

The distribution of the focal mechanisms of the shallow and intermediate depth (h>40 km) earthquakes of the Aegean and the surrounding area is discussed. The data consist of all events of the period 1963–1986 for the shallow, and 1961–1985 for the intermediate depth earthquakes, withM _s 5.5. For this purpose, all published fault plane solutions for each event have been collected, reproduced, carefully checked and if possible improved accordingly. The distribution of the focal mechanisms of the earthquakes in the Aegean declares the existence of thrust faulting following the coastline of southern Yugoslavia, Albania and western Greece extending up to the island of Cephalonia. This zone of compression is due to the collision between two continental lithospheres (Apulian-Eurasian). The subduction of the African lithosphere under the Aegean results in the occurrence of thrust faulting along the convex side of the Hellenic arc. These two zones of compression are connected via strike-slip faulting observed at the area of Cephalonia island. TheP axis along the convex side of the arc keeps approximately the same strike throughout the arc (210° NNE-SSW) and plunges with a mean angle of 24° to southwest. The broad mainland of Greece as well as western Turkey are dominated by normal faulting with theT axis striking almost NS (with a trend of 174° for Greece and 180° for western Turkey). The intermediate depth seismicity is distributed into two segments of the Benioff zone. In the shallower part of the Benioff zone, which is found directly beneath the inner slope of the sedimentary arc of the Hellenic arc, earthquakes with depths in the range 40–100 km are distributed. The dip angle of the Benioff zone in this area is found equal to 23°. This part of the Benioff zone is coupled with the seismic zone of shallow earthquakes along the arc and it is here that the greatest earthquakes have been observed (M _s 8.0). The deeper part (inner) of the Benioff zone, where the earthquakes with depths in the range 100–180 km are distributed, dips with a mean angle of 38° below the volcanic arc of southern Aegean.     

Discrimination of a seismic source doublet in the Northridge, California earthquake of 17 January 1994

The January 17, 1994 Northridge earthquake (Mw = 6.7, 34.213° N, 118.537° W, depth = 18.4 km) was recorded extensively in the immediate source region by strong, ground motion accelerometers. The resulting seismograms show complex S wave patterns. Nevertheless, visual correlations of the strong-ground-motion velocity and displacement time-histories clearly identify two significant wave pulses: a secondary S pulse (called S2) arriving 3–5 seconds after the initial S wave pulse (called S1). A plausible assumption is that these phases are generated at areas on the rupturing thrust fault that experienced especially large slip. Conventional travel-time computations, relating the relative arrival times between the onsets of the primary S1 and secondary S2 phases, yield a hypocenter of the initiation point, constrained to a independently etimated fault plane, of the secondary wave source (called H2) at 34.26°N, 118.54° W, with a depth of 14.1 km; the 68% confidence error in depth is 1.3 km. This location is about 6 km up-dip and north from the estimated hypocenter, on the fault plane of the initial principal seismic source (called H1). The seismic moment for both the initial H1 and secondary source H2 was estimated from the SH displacement pulse. Values averaged over eight stations were 8.61 ± 9.56 × 10²⁴ dyne-cm and 2.49 ± 2.31 × 10²⁵ dyne-cm respectively. Reasons why the sum of the two seismic moments is smaller than the total estimated seismic moment of 1.2 × 10²⁶ dyne-cm for the Northridge earthquake are discussed. The location of the initiation point of a second source H2 in the Northridge thrust faulting is consistent with independent computations of the fault slip pattern. The estimated stress drop for the initial and secondary sources are 1 = 150 ± 15 bars and 2 = 110 ± 11 bars, respectively.     

Deformation patterns associated with the M5.9 Athens (Greece) earthquake of 7 September 1999

The static displacement field of the Athens 1999 earthquake has been numerically modeled by a BEM method and analysed from SAR interferometry images with compatible results: (a) for a fault that reaches the surface the subsidence field coincides with the hangingwall domain of the Fili neotectonic normal fault with maximum amplitude, d _max, 5.5–7 cm, which is consistent with the possibly co-seismic displacement of 6–10 cm observed in the field, the average fault dislocation of 5–8 cm found by the application of circular source models, and the displacement up to 6 cm predicted by empirical relations between magnitude and displacement; the field of uplift covers the footwall domain of the fault with d _max1.5 cm;d gradually decreases with distance from the fault at a gradient of 0.4 cm/km, (b) for a blind fault d _max is only 1.8 and 0.3 cm in the hangingwall and footwall, respectively, and the decay gradient becomes 0.15 cm/km, (c) the total deformation area is 15 km × 15 km and the Fili fault, with a preferred mean dip of 60°, constitutes the natural boundary between the subsidence and uplift areas. The macroseismic field pattern is similar with that of the static ground deformation. The majority of intensity values VI (MM and EMS-98 scales), are distributed within the hangingwall of the Fili fault, while the highest intensities (VIII and IX) concentrate very close to the Fili fault within its hangingwall domain. A gradual decrease of the intensities with the distance from the Fili fault is evident. Because of the similarity between the intensity distribution pattern and the static ground deformation pattern, we make the hypothesis that the latter predicts well enough the main characteristics of the former although the ground displacement is dominated by relatively low frequency as compared to the ground acceleration.     

Magnitude–intensity and intensity–attenuation relationships for atlas region and Algerian earthquakes

This paper presents the results of an investigation of the magnitude–intensity and intensity–attenuation relationships for earthquakes in the Atlas block and Algeria using macroseismic data. This work is based on a selected sample of isoseismal maps from 32 events which were recently revised. Surface-wave magnitudes, M_s, are recalculated using the Prague formula and range from 4·2 to 7·45. Because the Atlas mountains block is in a collision zone, earthquakes occur in general within a layer 15 km deep. Expressions of general form for the magnitude–intensity and intensity–attenuation correlations are adopted and are, respectively, and where R² = d² + h², d the source distance in km, h the focal depth in km, M_s the revised surface-wave magnitude, M_sc the predicted surface-wave magnitude, I_i the intensity at isoseismal i, I the predicted intensity, σ the standard deviation and P is zero for 50-percentile values and one for 84-percentile, and the coefficients A's and B's are determined by regression analysis. The results of this study show that the intensity–attenuation models are adequate to predict quite well the die-out of intensity with distance in the Atlas zone and coastal Algeria; it is also found that magnitude can be predicted accurately by calibrating isoseismal radii against revised instrumental surface-wave magnitude. Such magnitude–intensity relationships may be used to evaluate the magnitude of historical earthquakes in the region under survey, with no instrumental data, for which isoseismal radii and intensities are available.