E. Kausel’s research while affiliated with University of Santiago Chile and other places

The Chilean subduction zone is one of the most active of the world with M=8 or larger interplate thrust earthquakes occurring every 10years or so on the average. The identification and characterization of pulses propagated from dominant asperities that control the rupture of these earthquakes is an important problem for seismology and especially for seismic hazard assessment since it can reduce the earthquake destructiveness potential. A number of studies of large Chilean earthquakes have revealed that the source time functions of these events are composed of a number of distinct energy arrivals. In this paper, we identify and characterize the high frequency pulses of dominant asperities using near source strong motion records. Two very well recorded interplate earthquakes, the 1985 Central Chile (Ms=7.8) and the 2007 Tocopilla (Mw=7.7), are considered. In particular, the 2007 Tocopilla earthquake was recorded by a network with absolute time and continuos recording. From the study of these strong motion data it is possible to identify the arrival of large pulses coming from different dominant asperities. The recognition of the key role of dominant asperities in seismic hazard assessment can reduce overestimations due to scattering of attenuation formulas that consider epicentral distance or shortest distance to the fault rather than the asperity distance. The location and number of dominant asperities, their shape, the amplitude and arrival time of pulses can be one of the principal factors influencing Chilean seismic hazard assessment and seismic design. The high frequency pulses identified in this paper have permitted us to extend the range of frequency in which the 1985 Central Chile and 2007 Tocopilla earthquakes were studied. This should allow in the future the introduction of this seismological result in the seismic design of earthquake engineering.

The Concepción–Constitución area [35–37°S] in South Central Chile is very likely a mature seismic gap, since no large subduction earthquake has occurred there since 1835. Three campaigns of global positioning system (GPS) measurements were carried out in this area in 1996, 1999 and 2002. We observed a network of about 40 sites, including two east–west transects ranging from the coastal area to the Argentina border and one north–south profile along the coast. Our measurements are consistent with the Nazca/South America relative angular velocity (55.9°N, 95.2°W, 0.610°/Ma) discussed by Vigny et al. (2008, this issue) which predicts a convergence of 68mm/year oriented 79°N at the Chilean trench near 36°S. With respect to stable South America, horizontal velocities decrease from 45mm/year on the coast to 10mm/year in the Cordillera. Vertical velocities exhibit a coherent pattern with negative values of about 10mm/year on the coast and slightly positive or near zero in the Central Valley or the Cordillera. Horizontal velocities have formal uncertainties in the range of 1–3mm/year and vertical velocities around 3–6mm/year. Surface deformation in this area of South Central Chile is consistent with a fully coupled elastic loading on the subduction interface at depth. The best fit to our data is obtained with a dip of 16±3°, a locking depth of 55±5km and a dislocation corresponding to 67mm/year oriented 78°N. However in the northern area of our network the fit is improved locally by using a lower dip around 13°. Finally a convergence motion of about 68mm/year represents more than 10m of displacement accumulated since the last big interplate subduction event in this area over 170 years ago (1835 earthquake described by Darwin). Therefore, in a worst case scenario, the area already has a potential for an earthquake of magnitude as large as 8–8.5, should it happen in the near future.

Publicación ISI Email : valerie@dgf.uchile.cl Although outer rise seismicity is less common than interplate seismicity in subduction zones, a significant level of seismicity has occurred between the Nazca trench and Juan Fernandez Ridge, in central Chile, during the past 20 years. We first study the 9 April 2001 ( Mw = 7.0) event and determine its focal mechanism, depth, and source time function by body-waveform inversion from teleseismic broadband data. The results indicate tensional faulting in the upper part of the mechanical lithosphere. Its strike ( 41 degrees) is similar to those observed in events down dip of the slab at about 100 km depth, which could indicate that these earthquakes occur in preexisting structures formed at the trench. Compressive outer rise events have also occurred during the 1980s in front of the rupture zone of the 1985 Mw 7.8 Valparai so Earthquake. To understand their relation with the state of stress of the lithosphere, we construct yield stress envelopes of the oceanic lithosphere, including static and dynamic stresses. Dynamic stresses are due either to slab pull, ridge push, resistive, and drag forces. We explain the sequence of compressive and tensional events by the accumulation of stress prior to 1985 when the subduction is assumed to be locked and after by the unlockage of the subduction by the Valparai so interplate event. The yield stress envelope analysis enables us to quantify the accumulation of compressive forces before 1985 and the tensional force after.

[1] A large (Mw 7.7) intermediate-depth earthquake occurred on 13 June 2005 in the Tarapacá region of the northern Chile seismic gap. Source parameters are inferred from teleseismic broadbands, strong motions, GPS and InSAR data. Relocated hypocenter is found at ∼98 km depth within the subducting slab. The 21-days aftershock distribution, constrained by a postseismic temporary array, indicates a sub-horizontal fault plane lying between the planes of the double seismic zone and an upper bound of the rupture area of 60 km × 30 km. Teleseismic inversion shows a slab-pull down dip extension mechanism on a nearly horizontal plane. Total seismic and geodetic moments are ∼5.5 × 1020 N.m, with an averaged slip of 6.5 m from geodesy. The earthquake rupture is peculiar in that the effective velocity is slow, 3.5 Km.s−1 for a high stress-drop, 21–30 MPa. We propose that rupture was due to the reactivation by hydraulic embrittlement of a inherited major lithospheric fault within the subducting plate. The stress-drop suggests that the region of the slab between planes of the double seismic zone can sustain high stresses.

Although intraplate seismicity in the outer rise region is less common than seismicity associated with interplate subduction zone earthquakes, a significant level of seismicity has occurred between the trench and JF Ridge (JFR), in Central Chile, during the past 20 years. Our study focuses on the April 9, 2001 (Mw=6.8) event and its aftershocks. We determine its focal mechanism, depth and source time function by waveform inversion from teleseimic broadband data. The results indicate tensional faulting with a strike of 36° , a dip of 43° , and a rake of -87° . The source time function has a duration of 11 s, with a seismic moment of 2.5x1026 Dyne cm. The depth of 15 km, obtained using oceanic velocity model, localizes this event below the Moho. This result, together with the depth distribution of the first 2 days of aftershocks, determined by the permanent local seismic network, indicates that the neutral stress surface must be located below 40 km depth. The strike of this event is very well constraint by SH waveform modeling and is parallel to the main local bathymetric structures including the JF Ridge. A clear SW to NE directivity source effect is also observed, in good agreement with the 2 days aftershock distribution, confirming that the strike of the focal plane is not parallel to the main trench direction. This is a main result that shows the interaction between JF Ridge and the initiation of the subduction process at the trench. Other outer rise events have occurred during the last 20 years in front of the rupture covered by the 1985 Mw 7.8 Valpara¡so Earthquake. To understand the relation between these earthquakes and the state of stress of the lithosphere, we use detailed bathymetric data in the hypothesis of a purely elastic plate to model the shape of the outer rise oceanic lithosphere assuming that the deformation is due to the bending prior to subduction and the loading by JF Ridge. The outer rise aftershocks are located in the upper tensional part of the lithosphere, shifted seaward of the trench because of JF Ridge. Previous compression events occurred 20 years before at 30 km depth in the same area. This enables us to propose a temporal migration at depth of the tensional/compressive limit within the lithosphere.

We study the possible seismic gap in the Concepción–Constitución region of south-central Chile and the nature of the M=7.8 earthquake of January 1939. From 1 March to 31 May 1996 a seismic network of 26 short period digital instruments was deployed in this area. We located 379 hypocenters with rms travel time residuals of less than 0.50 s using an approximate velocity distribution. Using the VELEST program, we improved the velocity model and located 240 high precision hypocenters with residuals less than 0.2 s. The large majority of earthquakes occurred along the Wadati–Benioff zone along the upper part of the downgoing slab under central Chile. A few shallow events were recorded near the chain of active volcanos on the Andes; these events are similar to those of Las Melozas near Santiago. A few events took place at the boundary between the coastal ranges and the central valley. Well constrained fault plane solutions could be computed for 32 of the 240 well located events. Most of the earthquakes located on the Wadati–Benioff zone had “slab-pull” fault mechanism due to tensional stresses sub-parallel to the downgoing slab. This “slab-pull” mechanism is the same as that of eight earthquakes of magnitude around 6 that are listed in the CMT catalog of Harvard University for the period 1980–1998. This is also the mechanism inferred for the large 1939 Chilean earthquake. A very small number of events in the Benioff zone had “slab-push” mechanisms, that is events whose pressure axis is aligned with the slab. These events are found in double layered Wadati–Benioff zones, such as in northern Chile or Japan. Our spatial resolution is not good enough to detect the presence of a double layer, but we suspect there may be one.

Two campaigns of Global Positioning System (GPS) measurements were carried out in the Concepción-Constitución area of Chile in 1996 and 1999. It is very likely that this area is a mature seismic gap, since no subduction earthquake has occurred there since 1835. In 1996, 32 sites were occupied in the range 35°S-37°S, between the Pacific coast of Chile and the Andes near the Chile-Argentina border. In 1999, the network was extended by the installation of 9 new points in the Arauco region whereas 13 points among the 1996 stations were reoccupied. The analysis of this campaign data set, together with the data recorded at eight continuous GPS sites (mostly IGS stations) in South America and surrounding regions, indicates a velocity of about 40 +/- 10 mm/yr in the direction N80-90°S for the coastal sites with respect to stable cratonic South America. This velocity decreases to about 20-25 mm/yr towards the Andes. We interpret this result as reflecting interseismic strain accumulation above the Nazca-South America subduction zone, due to a locked thrust zone extending down to about 60 km depth.

We have analyzed four large to great historic earthquakes that occurred along the central Chile subduction zone from north to south on November 11, 1922 (M-s = 8.3), April, 1943 (M-s = 7.9), December 1, 1928 (M-s = 8.0) and January 25, 1939 (M-s = 7.8). Waveform modeling and P-wave first motions indicate that the 1922, 1928 and 1943 earthquakes are shallow and consistent with underthrusting of the Nazca Plate beneath the South American plate. In contrast, the 1939 earthquake is not an underthrusting event but rather a normal fault event within the down-going slab. The 1922 earthquake is by far the largest event with a complex source time function showing three pulses of moment release and a duration of 75 a. The 1943 earthquake has a simple source time function with one pulse of moment release and a duration of 24 s. This event had a local tsunami of 4 m and a far-field tsunami height in Japan of 10-30 cm. The 1928 earthquake also has a simple source time function with a duration of 28 s. The aftershocks and highest intensities are south of the epicenter indicating a southward rupture with most of the seismic moment release occurring 50-80 km south of the 1928 epicenter but still north of the adjacent 1939 earthquake region. The 1939 Chillan earthquake was not an underthrusting but rather a complex normal fault earthquake. Our preferred model is a normal fault mechanism at a depth of 80 to 100 km with two pulses of moment release and a total duration of approximately 60 s. The high intensities, lack of a tsunami, and inland location associated with the 1939 event are all consistent with an intraplate event within the down-going slab. The 1939 earthquake was clearly more destructive than the other similar size or larger events. This may in part be due to the intraplate nature of the event but also due to high amplification of the sites in the Central Valley of south central Chile.

An Mw = 8.0 earthquake occurred on 30 July 1995 in the Antofagasta region (northern Chile). The main rupture, corresponding to thrust faulting, developed from 10 to 50 km in depth along the subduction interface between the Nazca and the South American plates. The 1995 earthquake took place just south of the large seismic gap where a great earthquake (M = 9) had occurred in 1877. Most of the 1995 rupture was located within a local network consisting of nine short-period stations that had been previously installed at the southern end of the 1877 gap, and the aftershock sequence could be accurately monitored. Little destruction resulted from the 1995 earthquake in spite of its large size. Ground acceleration in Antofagasta reached 29% of gravity. A tsunami wave, 2 to 2.5 m high, was observed along the coast from Mejillones to Taltal. One strong foreshock (Mw = 6.2) occurred in the 1995 hypocentral region 6 months before the main event. Body-wave modeling of broadband seismograms from the global network, along with the analysis of the aftershock distribution, allows us to propose a well-constrained model for the whole rupture process. Some additional details of the rupture were obtained from an accelerometer record at Antofagasta. The main rupture started as a double even with thrust mechanism below the southern part of the Mejillones peninsula, and it propagated southward in a N200°E direction with an average velocity of 2.8 km/sec. It ended near the trench in normal faulting. The total rupture area and seismic moment are 185 × 90 km2 and 1.2 × 1028 dyne-cm, respectively. The aftershock distribution delineates a well-defined rupture surface along the subduction interface. The distribution of epicenters during the first 20 h of aftershock activity shows a sharp northern boundary beneath the Mejillones peninsula. Hence, the 1995 main rupture did not propagate north of the Mejillones peninsula into the 1877 gap. Aftershocks during the following 2 weeks indicate a growth of the initial rupture zone toward the north. The mechanisms of the strongest aftershocks are similar to that of the mainshock. The down-dip termination of the main rupture corresponds to the maximum depth (50 km) of the region that had been identified as the locked part of the subduction interface from the analysis of the microseismicity recorded by the local network prior to the 1995 event. A well-constrained dislocation model is proposed for the 1995 main rupture, which produces surface displacements in good agreement with available observations of coastal uplift and GPS measurements. The dislocation model, as well as Global Positioning System (GPS) measurements, indicate that the 1995 earthquake generated E-W extension in the coastal region of Antofagasta. The Atacama fault, located 40 to 50 km above the 1995 main rupture, showed small fresh surface ruptures near Sierra Remiendos (70 km to the SSE of Antofagasta) with a maximum vertical offset of 20 cm. This offset corresponds to normal faulting, which is in agreement with the E-W co-seismic extension. The Mejillones peninsula appears to be the surface expression of a barrier that interrupted the propagation of the 1995 rupture to the north into the region of the 1877 gap. Modeling of static stress changes induced by the Antofagasta earthquake indicates an increase in compressive stresses along a direction transverse to the trench immediately to the north of the 1995 rupture surface. Thus, the chances for the reactivation of the 1877 gap after this event are greater now.