Pure and Applied Geophysics

Bandy, William; Taran, Yuri; Gutiérrez, Carlos; Kostoglodov, Vladimir

doi: 10.1007/s00024-010-0241-7pmid: N/A

Hauksson, Egill; Stock, Joann; Hutton, Kate; Yang, Wenzheng; Vidal-Villegas, J.; Kanamori, Hiroo

doi: 10.1007/s00024-010-0209-7pmid: N/A

The El Mayor-Cucapah earthquake sequence started with a few foreshocks in March 2010, and a second sequence of 15 foreshocks of M > 2 (up to M4.4) that occurred during the 24 h preceding the mainshock. The foreshocks occurred along a north–south trend near the mainshock epicenter. The M w 7.2 mainshock on April 4 exhibited complex faulting, possibly starting with a ~M6 normal faulting event, followed ~15 s later by the main event, which included simultaneous normal and right-lateral strike-slip faulting. The aftershock zone extends for 120 km from the south end of the Elsinore fault zone north of the US–Mexico border almost to the northern tip of the Gulf of California. The waveform-relocated aftershocks form two abutting clusters, each about 50 km long, as well as a 10 km north–south aftershock zone just north of the epicenter of the mainshock. Even though the Baja California data are included, the magnitude of completeness and the hypocentral errors increase gradually with distance south of the international border. The spatial distribution of large aftershocks is asymmetric with five M5+ aftershocks located to the south of the mainshock, and only one M5.7 aftershock, but numerous smaller aftershocks to the north. Further, the northwest aftershock cluster exhibits complex faulting on both northwest and northeast planes. Thus, the aftershocks also express a complex pattern of stress release along strike. The overall rate of decay of the aftershocks is similar to the rate of decay of a generic California aftershock sequence. In addition, some triggered seismicity was recorded along the Elsinore and San Jacinto faults to the north, but significant northward migration of aftershocks has not occurred. The synthesis of the El Mayor-Cucapah sequence reveals transtensional regional tectonics, including the westward growth of the Mexicali Valley and the transfer of Pacific–North America plate motion from the Gulf of California in the south into the southernmost San Andreas fault system to the north. We propose that the location of the 2010 El Mayor-Cucapah, as well as the 1992 Landers and 1999 Hector Mine earthquakes, may have been controlled by the bends in the plate boundary.

Castro, Raul; Perez-Vertti, Arturo; Mendez, Ignacio; Mendoza, Antonio; Inzunza, Luis

doi: 10.1007/s00024-010-0177-ypmid: N/A

We relocated the hypocentral coordinates of small to moderate-sized earthquakes reported by the National Earthquake Information Center (NEIC) between April 2002 and August 2006 in the Gulf of California region and recorded by the broadband stations of the network of autonomously recording seismographs (NARS–Baja array). The NARS–Baja array consists of 19 stations installed in the Baja California peninsula, Sonora and Sinaloa, Mexico. The events reported by the preliminary determinations of epicenters (PDE) catalog within the period of interest have moment magnitudes (M w) ranging between 1.1 and 6.7. We estimated the hypocentral location of these events using P and S wave arrivals recorded by the regional broadband stations of the NARS–Baja and the RESBAN (Red Sismológica de Banda Ancha) arrays and using a standard location procedure with the HYPOCENTER code (Lienert and Havskov in Seism Res Lett 66:26–36, 1995) as a preliminary step. To refine the location of the initial hypocenters, we used the shrinking box source-specific station term method of Lin and Shearer (J Geophys Res 110, B04304, 2005). We found that most of the seismicity is distributed in the NW–SE direction along the axis of the Gulf of California, following a linear trend that, from north to south, steps southward near the main basins (Wagner, Delfin, Guaymas, Carmen, Farallon, Pescadero and Alarcon) and spreading centers. We compared the epicentral locations reported in the PDE with the locations obtained using regional arrival times, and we found that earthquakes with magnitudes in the range 3.2–5.0 mb differ on the average by as much as 43 km. For the M w magnitude range between 5 and 6.7 the discrepancy is less, differing on the average by about 25 km. We found that the relocated epicenters correlate well with the main bathymetric features of the Gulf.

Núñez-Cornú, Francisco; Rutz-López, Marta; Márquez-Ramírez, Víctor; Suárez-Plascencia, Carlos; Trejo-Gómez, Elizabeth

doi: 10.1007/s00024-010-0178-xpmid: N/A

We present a fresh look at the source region of the 22 January 2003 M w 7.4 Armería earthquake, which occurred off the Pacific coast of the state of Colima, Mexico, near the town of Armería. The effects of this earthquake in the neighboring states of Colima and Jalisco were different and stronger than those of previous recent major earthquakes in the region. This earthquake and its aftershocks were recorded by two local telemetered seismograph networks (RESCO and RESJAL). From 22 January to 24 January 2003, no important seismicity was located on the plates interface, or within the Rivera Plate, and most epicenters were located west of the Armería River, which is the western border of the Colima Graben, and is located outside of the Colima Gap region. From 24 January to 31 January, the seismicity recorded by both networks showed a migration in depth, with an almost vertical offshore distribution between 4 and 24 km in depth. For this period, a seven-station portable digital seismograph network, equipped with three-component seismometers, was deployed in the epicentral area to study the aftershock sequence in detail. With this denser network more than 200 M L > 2.0 aftershocks were recorded. The aftershock foci were deeper than those recorded during the early period and most of them locate on a hypothetical 12° dipping interface between the Rivera and North American Plates. Composite focal mechanism solutions for the aftershocks located during both periods indicate a reverse fault character that changes with time. Analysis of the new dataset still indicates that the earthquake was a shallow intraplate event.

Calmus, Thierry; Pallares, Carlos; Maury, René; Aguillón-Robles, Alfredo; Bellon, Hervé; Benoit, Mathieu; Michaud, François

doi: 10.1007/s00024-010-0204-zpmid: N/A

The study of the geochemical compositions and K-Ar or Ar-Ar ages of ca. 350 Neogene and Quaternary lavas from Baja California, the Gulf of California and Sonora allows us to discuss the nature of their mantle or crustal sources, the conditions of their melting and the tectonic regime prevailing during their genesis and emplacement. Nine petrographic/geochemical groups are distinguished: “regular” calc-alkaline lavas; adakites; magnesian andesites and related basalts and basaltic andesites; niobium-enriched basalts; alkali basalts and trachybasalts; oceanic (MORB-type) basalts; tholeiitic/transitional basalts and basaltic andesites; peralkaline rhyolites (comendites); and icelandites. We show that the spatial and temporal distribution of these lava types provides constraints on their sources and the geodynamic setting controlling their partial melting. Three successive stages are distinguished. Between 23 and 13 Ma, calc-alkaline lavas linked to the subduction of the Pacific-Farallon plate formed the Comondú and central coast of the Sonora volcanic arc. In the extensional domain of western Sonora, lithospheric mantle-derived tholeiitic to transitional basalts and basaltic andesites were emplaced within the southern extension of the Basin and Range province. The end of the Farallon subduction was marked by the emplacement of much more complex Middle to Late Miocene volcanic associations, between 13 and 7 Ma. Calc-alkaline activity became sporadic and was replaced by unusual post-subduction magma types including adakites, niobium-enriched basalts, magnesian andesites, comendites and icelandites. The spatial and temporal distribution of these lavas is consistent with the development of a slab tear, evolving into a 200-km-wide slab window sub-parallel to the trench, and extending from the Pacific coast of Baja California to coastal Sonora. Tholeiitic, transitional and alkali basalts of subslab origin ascended through this window, and adakites derived from the partial melting of its upper lip, relatively close to the trench. Calc-alkaline lavas, magnesian andesites and niobium-enriched basalts formed from hydrous melting of the supraslab mantle triggered by the uprise of hot Pacific asthenosphere through the window. During the Plio-Quaternary, the “no-slab” regime following the sinking of the old part of the Farallon plate within the deep mantle allowed the emplacement of alkali and tholeiitic/transitional basalts of deep asthenospheric origin in Baja California and Sonora. The lithospheric rupture connected with the opening of the Gulf of California generated a high thermal regime associated to asthenospheric uprise and emplaced Quaternary depleted MORB-type tholeiites. This thermal regime also induced partial melting of the thinned lithospheric mantle of the Gulf area, generating calc-alkaline lavas as well as adakites derived from slivers of oceanic crust incorporated within this mantle.

Andrews, Vanessa; Stock, Joann; Ramírez Vázquez, Carlos; Reyes-Dávila, Gabriel

doi: 10.1007/s00024-010-0203-0pmid: N/A

On 22 January 2003, the M w = 7.6 Tecomán earthquake struck offshore of the state of Colima, Mexico, near the diffuse triple junction between the Cocos, Rivera, and North American plates. Three-hundred and fifty aftershocks of the Tecomán earthquake with magnitudes between 2.6 and 5.8, each recorded by at least 7 stations, are relocated using the double difference method. Initial locations are determined using P and S readings from the Red Sismológica Telemétrica del Estado de Colima (RESCO) and a 1-D velocity model. Because only eight RESCO stations were operating immediately following the Tecomán earthquake, uncertainties in the initial locations and depths are fairly large, with average uncertainties of 8.0 km in depth and 1.4 km in the north–south and east–west directions. Events occurring between 24 January and 31 January were located using not only RESCO phase readings but also additional P and S readings from 11 temporary stations. Average uncertainties decrease to 0.8 km in depth, 0.3 km in the east–west direction, and 0.7 km in the north–south direction for events occurring while the temporary stations were deployed. While some preliminary studies of the early aftershocks suggested that they were dominated by shallow events above the plate interface, our results place the majority of aftershocks along the plate interface, for a slab dipping between approximately 20° and 30°. This is consistent with the slab positions inferred from geodetic studies. We do see some upper plate aftershocks that may correspond to forearc fault zones, and faults inland in the upper plate, particularly among events occurring more than 3 months after the mainshock.

Quintanar, Luis; Rodríguez-Lozoya, Héctor; Ortega, Roberto; Gómez-González, Juan; Domínguez, Tonatiuh; Javier, Clara; Alcántara, Leonardo; Rebollar, Cecilio

doi: 10.1007/s00024-010-0202-1pmid: N/A

Aftershock locations, source parameters and slip distribution in the coupling zone between the overriding North American and subducted Rivera and Cocos plates were calculated for the 22 January 2003 Tecomán earthquake. Aftershock locations lie north of the El Gordo Graben with a northwest-southeast trend along the coast and superimposed on the rupture areas of the 1932 (M w = 8.2) and 1995 (M w = 8.0) earthquakes. The Tecomán earthquake ruptured the northwest sector of the Colima gap, however, half of the gap remains unbroken. The aftershock area has a rectangular shape of 42 ± 2 by 56 ± 2 km with a shallow dip of roughly 12° of the Wadati-Benioff zone. Fault geometry calculated with the Nábělek (1984) inversion procedure is: (strike, dip, rake) = (277°, 27°, 78°). From the teleseimic body wave spectra and assuming a circular fault model, we estimated source duration of 20 ± 2 s, a stress drop of 5.4 ± 2.5 MPa and a seismic moment of 2.7 ± .7 × 1020 Nm. The spatial slip distribution on the fault plane was estimated using new additional near field strong motion data (54 km from the epicenter). We confirm their main conclusions, however we found four zones of seismic moment release clearly separated. One of them, not well defined before, is located toward the coast down dip. This observation is the result of adding new data in the inversion. We calculated a maximum slip of 3.2 m, a source duration of 30 s and a seismic moment of 1.88 × 1020 Nm.

Zobin, Vyacheslav

doi: 10.1007/s00024-010-0172-3pmid: N/A

This study examines two large thrust subduction earthquakes occurring within the Rivera-Cocos plate boundary which struck the western coast of México on 9 October 1995, Mw 8.0, and 21 (22 GMT) January 2003, Mw 7.5. The Modified Mercalli (MM) earthquake intensities observed during these earthquakes were surprising for some towns located in the Mexican coastal zone. During the smaller Mw 7.5 2003 earthquake, MM intensity VII was observed for towns of Colima, Villa de Alvarez and Ixtlahuacán, while during the larger Mw 8.0 1995 earthquake, their MM intensities were only IV–V, V and V–VI, respectively. We construct the macroseismic patterns for these two earthquakes and discuss the possible reasons for the significant difference in the outline of the MM VII isoseismals, such as the tectonic setting of epicentral zones and the directivity of rupture processes along and across the coastal line.

Michaud, François; Calmus, Thierry; Ratzov, Gueorgui; Royer, Jean-Yves; Sosson, Marc; Bigot-Cormier, Florence; Bandy, William; Mortera Gutiérrez, Carlos

doi: 10.1007/s00024-010-0193-ypmid: N/A

The relative motion of the Pacific plate with respect to the North America plate is partitioned between transcurrent faults located along the western margin of Baja California and transform faults and spreading ridges in the Gulf of California. However, the amount of right lateral offset along the Baja California western margin is still debated. We revisited multibeam swath bathymetry data along the southern end of the Tosco-Abreojos fault system. In this area the depths are less than 1,000 m and allow a finer gridding at 60 m cell spacing. This improved resolution unveils several transcurrent right lateral faults offsetting the seafloor and canyons, which can be used as markers to quantify local offsets. The seafloor of the southern end of the Tosco-Abreojos fault system (south of 24°N) displays NW–SE elongated bathymetric highs and lows, suggesting a transtensional tectonic regime associated with the formation of pull-apart basins. In such an active tectonic context, submarine canyon networks are unstable. Using the deformation rate inferred from kinematic predictions and pull-apart geometry, we suggest a minimum age for the reorganization of the canyon network.

Bartolomé, Rafael; Dañobeitia, Juanjo; Michaud, François; Córdoba, Diego; Delgado-Argote, Luis

doi: 10.1007/s00024-010-0206-xpmid: N/A

Three thousand kilometres of multichannel (MCS) and wide-angle seismic profiles, gravity and magnetic, multibeam bathymetry and backscatter data were recorded in the offshore area of the west coast of Mexico and the Gulf of California during the spring 1996 (CORTES survey). The seismic images obtained off Puerto Vallarta, Mexico, in the Jalisco subduction zone extend from the oceanic domain up to the continental shelf, and significantly improve the knowledge of the internal crustal structure of the subduction zone between the Rivera and North American (NA) Plates. Analyzing the crustal images, we differentiate: (1) An oceanic domain with an important variation in sediment thickness ranging from 2.5 to 1 km southwards; (2) an accretionary prism comprised of highly deformed sediments, extending for a maximum width of 15 km; (3) a deformed forearc basin domain which is 25 km wide in the northern section, and is not seen towards the south where the continental slope connects directly with the accretionary prism and trench, thus suggesting a different deformational process; and (4) a continental domain consisting of a continental slope and a mid slope terrace, with a bottom simulating reflector (BSR) identified in the first second of the MCS profiles. The existence of a developed accretionary prism suggests a subduction–accretion type tectonic regime. Detailed analysis of the seismic reflection data in the oceanic domain reveals high amplitude reflections at around 6 s [two way travel time (twtt)] that clearly define the subduction plane. At 2 s (twtt) depth we identify a strong reflection which we interpret as the Moho discontinuity. We have measured a mean dip angle of 7° ± 1° at the subduction zone where the Rivera Plate begins to subduct, with the dip angle gently increasing towards the south. The oceanic crust has a mean crustal thickness of 6.0–6.5 km. We also find evidence indicating that the Rivera Plate possibly subducts at very low angles beneath the Tres Marias Islands.