Seismic velocity structure of the Guerrero gap, Mexico

 

Jaime Domínguez¹, Gerardo Suárez², Diana Comte³ and Luis Quintanar²

 

¹ Colegio de Ciencias y Humanidades, Plantel Sur, Universidad Nacional Autónoma de México, 04510 México, D. F., México

² Instituto de Geofísica, Universidad Nacional Autónoma de México, 04510 México, D. F., México

³ Dept. de Geofísica. Universidad de Chile, Casilla 2777, Santiago, Chile Email: jaimedr@hotmail.com

 

Received: August 26, 2005
Accepted: April 4, 2006

 

Resumen

Se obtiene la estructura bidimensional de velocidades sísmicas de la brecha de Guerrero aplicando el método de mínimos cuadrados amortiguados a los datos de la sismicidad local registrada por una red telemétrica situada en la costa, ubicada en la zona de subducción de la placa de Cocos. La región se parametrizó con una malla de 64 cubos en seis capas, un total de 384 bloques. Los resultados de esta inversión tridimensional (3–D) mostraron una diferencia de velocidad de onda P, entre bloques adyacentes y paralelos a la costa, no mayor de 0.25 km/s, mostrando una simetría bidimensional. Se utilizó una segunda inversión bidimensional (2–D), que toma en cuenta la similitud de velocidad de onda P de bandas paralelas a la costa, para generar una estructura de megabloques. Una inversión final muestra una estructura de velocidades de onda P con valores de 5.4 a 8.2 km/s, y valores para onda S entre 3.2 y 4.7 km/s, con una corteza continental de 32 km de espesor, compuesta de cuatro megabloques planos con un intervalo de velocidad de onda P de 5.4 a 7.1 km/s. El Moho se localiza a una profundidad de 32 km sobre una cuña del manto entre la corteza continental y la corteza oceánica. La corteza oceánica subducente se compone de tres capas (7.2–7.7 km/s), que presentan un ángulo de buzamiento de 26°. Un cambio importante de velocidad (7.2 a 7.6 km/s) a una profundidad de 30 km hace pensar en un cambio de fase de basalto a eclogita. El manto tiene una velocidad media de 8.2 km/s. El nuevo modelo de velocidades redujo el error de localización de la sismicidad local, la cual se ajusta mejor a las características de la brecha de Guerrero.

Palabras clave: Estructura de velocidades sísmicas, bloques de velocidad, inversión sísmica.

 

Abstract

A two–dimensional velocity structure of the Guerrero gap was obtained by applying a damped least square method to hypocenters of local seismicity recorded by a telemetric network situated on the Guerrero coast, above Cocos plate subduction zone. The region was parameterized by a mesh of 64 cubes in six layers, a total of 384 blocks. The results of 3–D inversion showed that differences of P–wave velocity values among blocks along the strike of the subduction zone were 0.25 km/s, effectively showing a two–dimensional symmetry. A 2–D inversion taking into account velocity similarities among the 2–D bands generated megablocks. A final inversion procedure yields P–wave velocity values ranging from 5.4 to 8.2 km/s, and S–wave values from 3.2 and 4.7 km/s, suggesting a continental crust with a thickness of 32 km composed of four flat megablocks with a P–wave velocity interval of 5.4 to 7.1 km/s. The Moho interface lies at 32 km depth and above a mantle wedge between continental and oceanic crust. The downgoing oceanic crust has three layers (7.2–7.7 km/s), dipping at an angle of 26°. A sharp velocity change at a depth of 30 km suggests a phase change from basalt to eclogite (7.2 to 7.6 km/s). The mantle has an average velocity of 8.2 km/s. The new velocity model reduced the error in locations and fits better the characteristics of the Guerrero gap.

Key words: Seismic velocity structure, velocity blocks, seismic inversion.

 

INTRODUCTION

The tectonic regime in the coast of Guerrero in central Mexico is dominated by subduction of the Cocos plate beneath the North American plate. The subducted slab reaches a maximum seismic depth of about 60 km (Suárez et al., 1990). The Guerrero coast region is marked by a well–defined seismic gap in which no large earthquake have occurred since 1908 (Suárez et al., 1990). A large earthquake may take place here (e.g., Nishenko and Singh, 1987), which could seriously affect Mexico City. This region is seismically defined by an active thrust interface, which extends to an anomalously shallow depth of 25 km (Nishenko and Singh, 1987; Dewey and Suárez, 1991; Suárez and Sánchez, 1996).

The Institute of Geophysics of the National Autonomous University of Mexico (UNAM) installed a telemetric seismic network composed of eight vertical–component seismometers of 1 Hz and one central station with three components (1 Hz), to monitor the seismic activity in the Guerrero gap before, during and after a large earthquake. This network spans an area between 100.24° and 101.38° W (Figure 1a). An average of five events was registered each day. Timing is controlled by a common GPS time base, P and S arrival times were read with a accuracy of one hundredth of a second. All events were routinely located using HYPO71PC (Lee and Valdés, 1985) with a flat–layer velocity model proposed by Suárez et al. (1992). A total of 6900 events have been recorded from September 1987 to July 1995, with magnitudes (M) ranging from 1.0 to 4.0. Along the coastline the distribution of earthquakes is concentrated along two seismic bands separated by 8 km (Figures 1a and 1b). The coastal band shows hypocenters with depths ranging between 10 and 25 km, dominated by thrust mechanisms with northeast dipping faults at an angle 12° (Suárez et al., 1990). The second seismic band farther inland exhibits hypocentral depths ranging between 32 and 42 km, and shows a combination of reverse and normal faulting mechanisms (Suárez et al., 1990).

The geometry of the subducted slab in this region features a shallow dipping plate (12°) beneath North American reaching a depth of 40 km. The slab is then bent upward, and follows a subhorizontal trajectory that extends inland at a depth of 50 km (Suárez et al., 1990 and Domínguez, 1991). Pardo and Suárez (1995) constructed the 3–D geometry of the entire subducting Cocos plate along the Mexican coastline.

Various workers have studied the velocity structure of this region and adjacent areas. Lewis and Snydsman (1979), Núñez–Cornú et al. (1992), Valdés–González (1993) and Valdés–González and Meyer (1996) applied different refraction ray trace methods to seismic data to obtain seismic velocity structure of the oceanic crust, along and normal to the coastline. Suárez et al. (1990), Araujo (1991) and Domínguez (1991) determined the geometry of the 2–D subducting slab along the Guerrero coastline using local information. Suárez et al. (1992) determined a one–dimensional (1D) seismic velocity model. However, there are no specific studies within the Guerrero gap that study in detail the tectonic characteristics of this region. The purpose of this work is to define the structure of the oceanic and continental lithosphere beneath the Guerrero coastline, applying a damped least square method (e.g. Aki and Lee, 1976; Roecker, 1982). We use local telemetric data to obtain the seismic velocity structure across the Guerrero gap.

 

INVERSION THEORY

The algorithm of the method proposed by Aki and Lee (1976) and Aki et al. (1977), as modified by Roecker (1982), is used to determine crustal velocity structure under the Guerrero coast. This method uses arrival times of P and S–waves of local and teleseismic events. A similar methodology has been used by several authors to study the structure of other tectonic environments (e.g., Roecker, 1982; Roecker, Yeh and Tsai, 1987; Aber and Roecker, 1991; Comte et al., 1994).

The inversion applies an iterative, non–linear formulation (Tarantola and Valette, 1982). Techniques of progressive inversion and parameter separation were also incorporated (Roecker, 1982). The earthquakes are relocated prior to perturbing the structure and the hypocentral parameters are decoupled from the velocity parameters when solving for structural perturbation. Intrinsic smoothing results from the use of several sizes blocks. The covariance matrix of the model is diagonal, and the size of this diagonal element had to be an order of magnitude larger than what might be presumed from considerations of a priori uncertainties, in order to be able to stabilize the solution.

To initiate the inversion process, the lithosphere beneath Guerrero gap was parameterized by a set of layers bound by horizontal interfaces, each of these horizontal layers is divided into a grid of rectangular prisms or blocks by two orthogonal sets of vertical interfaces. The one–dimensional or three–dimensional velocity structure is determined through minimizing, in a damped least square sense, the travel time residual of P and S–waves from earthquakes by adjusting the Earth model and the hypocentral coordinates of the earthquake events. P and S–waves velocities are determined as independent parameters within each block. Approximate ray–tracing is used to provide a reasonable estimate of travel time through a laterally varying structure, thus reducing bias introduced and by non–linearity and the choice of a particular starting model (Thurber, 1983; Aki and Lee, 1976; Aki et al., 1977; Roecker, 1982).

 

SEISMIC VELOCITY STRUCTURE INVERSION

 

1. One–dimensional inversion

The first step in the inversion process is obtaining a one–dimensional seismic velocity model. An early velocity model in the area of the Guerrero coast was obtained by Suárez et al. (1992), applying the minimum apparent velocity of refracted waves (e.g. Matumoto, 1977). This model was used as the initial velocity model for one–dimensional inversion.

The results of one–dimensional velocity inversion show P–wave values similar to those obtained by Suárez et al. (1990), particularly for the intermediate layers (12.5–45 km) which have the best resolution (between 0.69–0.87), and where most of the seismicity is located. The resolution of deeper layers (40–50 km) is low (between 0.01 and 0.06). They also show the largest errors (0.47 km/s–0.48 km/s). This is due to the fact that few rays go through them. Similarly, the top layer (0.0–10.1 km) is poorly resolved (0.36) with a resulting error of 0.20 km/s. The resulting P–wave velocity model is similar to that obtained by Suárez et al. (1992, Figure 2), with a difference of not more that 0.4 km/s.

 

2.  Three–dimensional inversion

The one–dimensional velocity structure is used as the initial model for the three–dimensional inversion process. A three–dimensional mesh is defined by 64 rectangular blocks in each one of the 6 layers. Each block has 10 km by side and a similar thickness to the one–dimensional velocity model (Figure 2). The mesh has a strike of 340°. The resulting three–dimensional velocity structure is shown in the Figure 3; we observed no large differences among velocity values of horizontally adjacent blocks parallel to the coastline. In general, the differences of P–wave velocity values are equal or less to 0.25 km/s, and 0.11 km/s for S–waves. For example, in the third layer (18.4–23.4 km, Figure 3), the difference between adjacent blocks is 0.06 to 0.18 km/s (P–wave), and 0.01 to 0.06 km/s for S–wave. The similarity of seismic velocities in Guerrero along the strike of the coastline suggest that the geological characteristics only change along the convergence between the Cocos and North America plates, in agreement with previous studies (e.g., Suárez et al., 1990; Valdés, 1993; Valdés–González and Meyer, 1996; Kostoglodov et al., 1996).

However, these results are not of uniform quality because resolution in some block is low, as they are sampled by not enough rays. To improve the resolution, we merged several cubes along the coastline to form bands. The results improved the resolution in each block or band by 40 % and reduced the error values by 10 % compared to the three–dimensional velocity structure (Figure 3). After join in blockbands parallel to the coast the cubes having the same speed values, we draw a vertical projection (following AA) which we present in the Figure 4. Grouping several blockbands from the' "megastructure" shown on this figure, to define larger blocks (hereafter named "megablocks") of similar wave speed, we inverted again to get the final velocity model. In the next section we explain the results.

 

3. Two–dimensional inversion

Different two–dimensional distributions of megablocks were tested to determine the best model parameter fit, the resolution and the error in each megablock with special attention for the velocity structures of adjacent areas, e. g, Valdés–González and Meyer (1996) and Suárez et al. (1992).

Valdés–González and Meyer (1996) found a compressional and shear velocity structure of the subduction zone between Petatlán and Mexico City (length of 450 km and depth of 110 km) and along the coast of the State of Guerrero, by applying a seismic refraction method to aftershock data of the M_s=7.6 Petatlán, Guerrero earthquake of March 14, 1979. This model consists of a two–dimensional ocean to continent structure featuring a continental block of three flat crustal layers (5.3 to 6.4 km/s V_p values) with the Moho at 45 km depth. The oceanic crust is composed of two layers with 8 km total thickness dipping 10° (6.4 to 7.4 km/s V_p values). The upper mantle–oceanic crust transition has a thickness of 55 km (7.9–8.1 km/s V_p values), and the oceanic upper mantle has a V_p of 8.2 (km/s).

Our best results are shown in Figure 5. After establishing the final distribution of megablocks, the damping value was modified, θ is a weight equal to, where σ² is the data variance and σ_m² is the model variance. The best fit was obtained for θ= 25 000.

The final seismic velocity structure (Figure 5) features a continental crust with four flat layers. The first has a thickness of 3 km (P1), which is poorly resolved (0.19) because the rays that cross it are perpendicular to the interface. The next layers (P2, P3 and P4) have a thickness of 10 km each and a resolution of between 0.84 and 0.93.

The Cocos plate is defined by three megablocks P6, P7 and P8, each with an average thickness of 5 km and resolution values of 0.83, 0.67 and 0.85 respectively. P9 is the deepest block with a poor resolution of 0.53. The velocity uncertainties are between 0.07 to 0.29 km/s, if we neglect the blocks that have poor resolution (P1 and P9), the velocity uncertainty interval is 0.07–0.25 km/s. The results for S–wave velocity are similar to those for P–wave velocity, with a resolution values less that 23 % and error values less 68 %, in relation to those values obtained for the P–wave model. The difference is due to the fact that the inversion scheme uses the S wave velocity to refine the P–wave velocity model (Roecker, 1982).

 

4. Resolution of the velocity model

In order to verify the robustness of the results, resolution and mean errors were tested using synthetic arrival time data of P and S–waves generated with the seismic velocity model, and the mesh of Figure 5. The arrival times were perturbed with random noise (time delays with a mean of 0.1 s and a standard deviation of 0.2 s) introduced in the observations on all P and S arrival times.

The inversion program was applied to these synthetic data and the seismic velocity structure was confirmed (Figure 6). In general, the resolution and errors observed were better for the original model of the Figure 5, with an interval resolution for P–waves of (0.48 – 0.99), and (0.35 – 0.97) for S–waves, and an interval error for P–waves of (0.03 – 0.13) km/s, and (0.02 – 0.05) km/s for S–waves. The observed change in P and S wave velocity values is 1 %, i.e. not more than 0.05 km/s for P–waves and 0.02 km/s for S–waves (Figure 6). These results show the robustness of the initial two–dimensional velocity structure obtained with local seismic data.

 

DISCUSSION

The distribution of seismicity relative to the network provides an adequate resolution of all megablocks in the structure. Often few seismic rays sample the blocks or cross them perpendicularly. Still, the results allow to define clearly some tectonic structures (Figure 5): (1) The oceanic crust of the subducting Cocos plate (P6, P7 and P8 blocks), with a dip of 26° as reported in large shallow thrust events by Pardo and Suárez (1995). (2) The Moho discontinuity (interface between P4 and P5) at a depth of 32 km; (3) the upper mantle (P9).

The velocity structures shown on Figure 5 agree with the distribution of megablocks for velocity structures in Oaxaca (Valdés et al., 1986) and Guerrero (Valdés–González and Meyer, 1996). The Oaxaca structure velocity is a seismic refraction and gravity model with 500 km length and 80 km in depth, composed by a continental crust (three layers with a thickness of 45 km and V_p interval of 4.3–7.0 km/s), the Moho is located at a depth of 45 km. The oceanic crust is composed of two layer with a total thickness of 8 km, over a transitional layer with a variable thickness (from 13 to 20 km), dipping with an angle of 10°. Under this structure there is the upper mantle megablock with V_p of 7.6 km/s. There are dimensional and velocity differences between both models. The transitional layer is similar to P8 block from the Figure 5. Only the distribution of the blocks is similar.

The Guerrero refraction model (Valdés–González and Meyer, 1996) is quite different. It does not provide enough detail in the contact zone beneath the Guerrero coastline. However, there are some similarities. Both models feature a continental crust of 4 flat layers and an oceanic crust of two thin layers over a high velocity block in the upper mantle.

Lewis and Snydsman (1979) defined two horizontal layers in the oceanic crust, south of Acapulco. The first layer has Vp between 4.2 and 6.2 km/s, and the second one has a Vp interval of 6.8–7.0 km/s. These two layers have a total thickness of 6 km, and may represent the upper layer of the oceanic crust (P6) shown in Figure 5, this value is possibly influenced by sediments.

The velocity change in the upper oceanic layer is possibly due to a transitional zone caused by a phase change from gabbro to eclogite, from P6 to P7 blocks (Figure 7), as in Kirby et al. (1996), Hacker (1996) and Rushmer (1996). Eclogite formation induced by dehydration of a warm slab (Heat flow > 75 mW/m²), causes the slow young descending slab (age< 15–25 Ma), to sink into the upper mantle. The age of the Cocos plate (Kostoglodov and Bandy, 1995) is of 13.5 Ma in the trench, and 17.0 Ma at the deepest seismicity. Convergence velocity between the Cocos and North America plates is 5.3 cm/yr (Bandy, 1992). A heat flow of 135 mW/m² matches the requirements of Kirby et al. (1996) to form eclogite at shallow depths within the subducted plate.

Other authors have found a phase change in the upper oceanic crust. In the Pacific plate, Northeast Japan, Hurukawa and Imoto (1992) suggest a phase change from basalt to eclogite controlling the maximum depth of coupling. In the Philippine plate, SW Japan, Hori et al. (1985) and Hori, (1990) propose a wave guide layer and a phase change from gabbro to eclogite (12–22 Ma) which reaches a maximum depth of 60 km and a thickness of less than 10 km (V_p=7.0 km/s), though there is no evidences that theses changes occur below to 60 km of depth. In Central and North Chile, Comte et al. (1994) found a phase change at 70 km of depth.

In the Mexican subduction zone, the seismogenic coupling area has a maximum depth of 25 km depth (Suárez et al., 1990; Pardo and Suárez, 1995; Kostoglodov et al., 1996; Suárez and Sánchez, 1996 and Pacheco et al., 1993). Therefore, velocity differences between blocks P6 and P7 may be responsible for the changes related to the seismogenic coupling zone. The differences between both blocks match with the change in seismotectonic regime between both bands. The first band shows only shallow–thrust faulting mechanisms over the Cocos–North America interface; the second seismic band is mainly within P7 block and shows thrust faulting for shallow seismicity and normal faulting for deep seismicity (Suárez et al., 1990; Araujo, 1991). After relocating the earthquakes, the first band shows an alignment that defines the coupling zone above the P6 block. The second seismicity band presents less dispersion in relation to hypo71 localization. The upper border of this seismicity coincides with the interface between both P6 and P7 blocks (Figure 7), after eclogite formation (Kirby et al., 1996; Figure 7).

The P8 block (Figure 7) seems to match to layer 3 of the oceanic lithosphere as proposed by Spudich and Orcutt (1980) which they defined as having average velocities between 7.2 and 7.8 km/s. This "basal layer" is defined as a transition zone between the crust and the oceanic upper mantle. Clague and Straley (1977) mention that serpentinized peridotite is commonly found to underlie the gabbro unit. Thus P8 block may represent a transitional layer between the Cocos plate (oceanic crust) and the upper mantle. The distribution of blocks of the continental crust, P1 to P4, in our model is in agreement with the continental arc tectonic province proposed by Christensen and Mooney (1995). This province is composed by three flat blocks. The first block has a thickness of 18 km and V_p interval of 5.7–6.4 km/s, the second block has a thickness of 10 km and V_p interval of 6.4–6.8 km/s, and the third block has a thickness of 10 km and V_p interval of 6.8–7.8 km/s. The last block has a thickness of 18 km and a V_p interval of 5.7–6.4 km/s. Below the continental crust there is an upper mantle block with V_p >7.8.

 

CONCLUSIONS

A two–dimensional seismic velocity model for Guerrero gap was obtained using a linearized inversion scheme, the damped least square method (Roecker, 1981 and 1982). 6955 micro–earthquakes were used. The proposed model is composed of four flat continental layers with V_p interval of 5.4 to 7.1 km/s and Vs (S–wave velocity values) interval of 3.2 to 4.0 km/s. The Moho interface is located at a depth of 32 km on top of the upper continental mantle layer. The oceanic crust dips at an angle of 26° and is composed of three megablocks. The two shallower blocks reflect the Cocos–North America plate interface and apparently show the presence of a phase change from basalt to eclogite at the bottom of the seismic coupling zone. It was found that seismic velocity of oceanic crust varies from 7.2 to 7.7 km/s for P–wave and 4.0 to 4.2 km/ s for S waves. An upper mantle block was defined with 8.2 km/s for P–wave and 4.7 km for S wave.

 

ACKNOWLEDGEMENTS

We thank Steve Roecker for providing us his computer programs, Jorge Estrada and Jesús Antonio Pérez for their technical support to the Guerrero telemetric network. Shri Krishna Singh and an anonymous reviewer for their suggestions, comments and corrections to the paper.

 

BIBLIOGRAPHY

AKI, K. and W. H. K. LEE, 1976. Determination of three–dimensional velocity anomalies under a seismic array using first P arrival times from local earthquake, a homogeneous initial model. J. Geophys. Res., 81, 23, 4381–4399.        [ Links ]

AKI, K., A. CHRISTOFFERSSON and E. S. HUSEBYE, 1977. Determination of the three dimensional seismic structure of the lithosphere. J. Geophys. Res., 82, 2, 277–296.        [ Links ]

AKI, K. and P. G. RICHARDS, 1980. Quantitative seismology: Theory and methods; W.H. Freeman, 916 pp.        [ Links ]

ARAUJO, A., 1991. Sismicidad y mecanismos focales en la brecha de Guerrero: Morfología de una placa joven, Tesis de Licenciatura, Fac. de Ingeniería, UNAM.        [ Links ]

BANDY, W., 1992. Geological and geophysical investigation of the Rivera–Cocos plate boundary: Implications for plate fragmentation, Ph.D. thesis, Tex. A&M Univ., College Station.        [ Links ]

CLAGUE, D. A. and P. F. STRALEY, 1977. Petrologic nature of the oceanic Moho. Geology, 5, 133–136, 1977.        [ Links ]

COMTE, D., S. W. ROECKER and G. SUÁREZ, 1994. Velocity structure in northern Chile: Evidence of subducted oceanic crust in the Nazca Plate. Geophys. J. Int., 117, 625–639.        [ Links ]

CHRISTENSEN, N. I. and W. D. MOONEY, 1995. Seismic velocity structure and composition of the continental crust: A global view. J. Geophys. Res., 100, B7, 9761–9788.        [ Links ]

DEWEY, J. W. and G. SUÁREZ, 1991. Seismotectonics of Middle America, Open File Rep.: The Geology of North America, Decade Map 1, Chapter 17.        [ Links ]

DOMÍNGUEZ, J., 1991. Geometría de la placa de Cocos en la región del Río Balsas, Guerrero; Tesis de Licenciatura, Fac. de Ciencias, UNAM.        [ Links ]

HACKER, B. R., 1996. Eclogite formation and the rheology, buoyancy, seismicity, and H₂O content of oceanic crust; Subduction: Top to Bottom, Geophysical Monograph 96; G. R. Bebout, D. W. Scholl, S. H. Kirby and J. P. Platt, Editors, AGU.        [ Links ]

HORI, S., H. INOUE, Y. FUKAO and M. UKAWA, 1985. Seismic detection of the untransformed 'basaltic' oceanic crust subducting into the mantle. Geophys. J. R. Astr. Soc., 83, 169–197.        [ Links ]

HORI, S., 1990. Seismic wave guided by untransformed oceanic crust subducting into the mantle: the case of the Kanto district, central Japan. Tectonophysics, 176, 355–376.        [ Links ]

HURUKAMA, N. and M. IMOTO, 1992. Subducting oceanic crust of Philippine sea and Pacific plate and weak zone normal compression in the Kanto district, Japan. Geophys. J. Int., 109, 639–652.        [ Links ]

KELLEHER J., L. SYKES and J. OLIVER, 1973. Possible criteria for predicting earthquake locations and their application to major boundaries of the Pacific and the Caribbean. J. Geophys. Res., 78, 14, 2584–2584.        [ Links ]

KIRBY, S., E. R. ENGDAHL and R. DELINGER, 1996. Intermediate–depth intraslab earthquakes and arc volcanism as physical expressions of crustal and uppermost mantle metamorphism in subducting slabs; Subduction: Top to Bottom, Geophysical Monograph 96; G. R. Bebout, D. W. Scholl, S. H. Kirby and J. P. Platt, Editors, AGU.        [ Links ]

KOSTOGLODOV, V. and W. BANDY, 1995. Seismotectonic constraints on the convergence rate between the Rivera and North America plates. J. Geophys. Res., 100, B9, 17,977–17,989.        [ Links ]

KOSTOGLODOV, V., W. BANDY, J. DOMÍNGUEZ and M. MENA, 1996. Gravity and Seismicity over the Guerrero Seismic Gap, Mexico. Geophys. Res. Lett., 23, 3385–3388.        [ Links ]

LEE, W. H. K. and C. VALDÉS, 1985. HYPO71PC: A personal computer version of the HYPO71 earthquake location program, U. S. Geol. Surv. Open File Rep., 85–749.        [ Links ]

LEWIS, T. R. B. and W. E. SNYDSMAN, 1979. Fine structure of the lower oceanic crust on the Cocos plate, J. Francheteu (Editor), Processes and Mid–ocean ridges. Tectonophysics, 55, 7–105.        [ Links ]

NISHENKO, J. and S. K. SINGH, 1987. Conditional probabilities for the recurrence of the large and great interplate earthquake along the Mexican subduction zone. Bull. Seismol. Soc. Am., 77–6, 2095–2114.        [ Links ]

NÚÑEZ–CORNÚ, F. J., D. CÓRDOBA, A. NAVA, L. GONZÁLEZ, A. GONZÁLEZ, G. CALDERÓN, M. J. BLANCO, J. LERMO, J. YAMAMOTO, F. MENDIOLA, G. CASTELLANOS, R. MOTA, E. SURIÑACH, M. CRUZ and C. SÁNCHEZ, 1992. Estudio de la estructura de la litósfera en México 1992; Boletín Informativo de la Unión Geofísica Mexicana, Época II, 12, No. 5.        [ Links ]

PACHECO, J. F., L. R. SYKES and C. H. SCHOLZ, 1993. Nature of seismic coupling along simple plate boundaries of subduction type. J. Geophys. Res., 98, 14133–14159.        [ Links ]

PARDO, M. and G. SUÁREZ, 1995. Shape of subducted Rivera and Cocos plates in southern Mexico: Seismic and tectonic implications. J. Geophys. Res., 100, 12357–12373.        [ Links ]

ROECKER, S. W., 1981. Seismicity and tectonics of the Pamir–Hindu Kush region of central Asia; Ph.D. thesis, MIT, Cambridge.        [ Links ]

ROECKER, S. W., 1982. Velocity structure of the Pamir–Hindu Kush region: Possible evidence of subducted crust. J. Geophys. Res., 87, B2, 945–959.        [ Links ]

ROECKER, S. W., Y. YEH and Y. TSAI, 1987. Three–dimensional P and S wave velocity structure beneath Taiwan: Deep structure beneath and arc–continent collision. J. Geophys. Res., 92, 10547–10570.        [ Links ]

RUSHMER, T., 1996. The influence of dehydration and partial melting reactions on the seismicity and deformation in warm subducting crust; Subduction: Top to Bottom, Geophysical Monograph 96; G. R. Bebout, D. W. Scholl, S. H. Kirby and J. P. Platt, Editors, AGU.        [ Links ]

SPUDICH, P. and J. ORCUTT, 1980. A new look at the seismic velocity structure of the oceanic crust. Rev. Geophys. Space Phys., 18, 3, 627–645.        [ Links ]

SUÁREZ, G., T. MONFRET, G. WITTLINGER and C. DAVID, 1990. Geometry of subduction and depth of the seismogenic zone in the Guerrero Gap, Mexico. Nature, 345, 6273, 336–338.        [ Links ]

SUÁREZ, G., J. P. LIGORRÍA and L. PONCE, 1992. Preliminary crustal structure of the coast of Guerrero, Mexico, using the minimum apparent velocity of refracted waves. Geofís. Int. 31, 3, 247–252.        [ Links ]

SUÁREZ, G. and O. SÁNCHEZ, 1996. Shallow depth of seismogenic coupling in southern Mexico: Implications for the maximum size of earthquakes in the subduction zone. Phys. of the Earth and Planet. Int., 93, 53–61.        [ Links ]

TARANTOLA, A. and B. VALETTE, 1982. Inverse problem = Quest for information. J. Geophys., 50, 159–170.        [ Links ]

THURBER, C., 1983. Earthquake locations and three dimensional crustal structure in the Coyote Lake area, central California. J. Geophys. Res., 88, 8226–8236.        [ Links ]

VALDÉS, C., W. D. MOONEY, S. K. SINGH, R. P. MEYER, C. LOMNITZ, J. H. LUETGERT, C. E. HELSLEY, T. R. LEWIS and M. MENA, 1986. Crustal structure of Oaxaca, Mexico. from seismic refraction measurements. Bull. Seismol. Soc. Am., 76, 2, 547–563.        [ Links ]

VALDÉS–GONZÁLEZ, C., 1993. Seismic structure, scaling, seismic attenuation and gaps using aftershocks from the major 1979 Petatlán (Surface–wave magnitude equal 7.6) earthquake. 1993. Ph. D. thesis, University of Wisconsin–Madison.        [ Links ]

VALDÉS–GONZÁLEZ, C. and R. P. MEYER, 1996. Seismic structure between the Pacific coast and Mexico City, from Petatlán earthquake (M_s=7.6) aftershocks. Geofís. Int. 35, 4, 377–401.        [ Links ]