S31A-01 INVITED
The Kinematic Puzzle of the Gulf of California Rift system
Extensive recent studies of the Gulf of California greatly advanced our understanding of the kinematics of this Pacific-North America plate boundary rift and its crustal structure. However, some kinematic discrepancies that need to be resolved, particularly for the time interval 12-6 Ma, after subduction had stopped but possibly before major opening began in the northern Gulf basins. Correlation of pyroclastic flow deposits across the Upper Delfin Basin segment (Oskin et al., 2001 and subsequent papers) indicates that virtually all of the opening between the coastlines in this segment (from San Felipe/Puertecitos in Baja California to the W side of Tiburon Island) occurred since ~6.1 Ma producing ~250 km of opening of the marine basin. This implies that the crust in the basin should have been brought into the region since ca. 6 Ma, perhaps by production of new igneous crust or remobilization of continental crust to fill the ~250-km gap. The total amount of post-6 Ma opening in the rift, including the onland deformation, is consistent with expected Pacific-North America displacement determined from the global plate circuit since 6 Ma, and it is also consistent with the slip history of the southern San Andreas fault. A smaller amount of motion (tens of kms) can be identified geologically post-12.5 Ma and pre-6 Ma. However, this is not sufficient to match the plate circuit results, which seem to require several hundred more km of Pacific-North America plate motion at this latitude between 12 Ma and 6 Ma. This motion has to have been located elsewhere, not between the modern marine basin boundaries. The Pliocene basin history of the Gulf has varied from place to place, as the loci of extension moved around in the rift system Similar variations in late Miocene time could explain this discrepancy, with abandoned extensional or strike-slip fault systems elsewhere, perhaps in the Sonoran coastal plain. However, further geological and geophysical work is needed to characterize this area and see if such a model is feasible and whether faults can be found that will link kinematically northward to structures of the same age, east of the San Andreas fault. A related discrepancy comes from reports of middle Miocene microfossils in marine sediments within these young northern Gulf basins. If these microfossils are primary, then the 250 km of opening would have a longer history, but it would not change the fact that some slip is missing and must be found elsewhere. In addition, the extensive volcanism in the Puertecitos Volcanic Province and Delfin Basin region since 12.5 Ma should have left a record in the marine sediments. Ash and pumice from these eruptions have been found in the onland exposures of the Gulf marine deposits (Puertecitos Fm, San Felipe marine sequence). Similar evidence should be sought in any future drill holes into the marine section.
S31A-02
Variable styles of rifting expressed in crustal structure across three rift segments of the Gulf of California
We present a summary of results from a crustal-scale seismic experiment conducted in the southern Gulf of California. This experiment, the PESCADOR experiment, imaged crustal structure across three rift segments, the Alarcon, Guaymas, and San José del Cabo to Puerto Vallarta (Cabo-PV) segments, using seismic refraction/wide-angle reflection data acquired with airgun sources and recorded by closely spaced (10-15 km) ocean-bottom seismometers (OBSs). The imaged crustal structure reveals a surprisingly large variation in rifting style and magmatism between these segments: the Alarcon segment is a wide rift with apparently little syn-rift magmatism; the Guaymas segment is a narrow, magmatically robust rift; and the Cabo-PV segment is a narrow, magmatically "normal" rift. Our explanation for the observed variability is non-traditional in that we do not invoke mantle temperature, the factor commonly invoked to explain end-member volcanic and non-volcanic rifted margins, as the source of the considerable, though non-end-member variability we observe. Instead, we invoke mantle depletion related to pre-rift arc volcanism to account for observed wide, magma-poor rifting and mantle fertility and possibly the influence of sediments to account for robust rift and post-rift magmatism. These factors may commonly vary over small lateral spatial scales in regions that have transitioned from convergent to extensional tectonics, as is the case for the Gulf of California and many other rifts. Our hypothesis suggests that substantial lateral variability may exist within the uppermost mantle beneath the Gulf of California today, and it is hoped that ongoing efforts to image upper mantle structure here will provide tests for this hypothesis.
S31A-03
Crustal Structure of Southern Baja California Peninsula, Mexico, and its Margins
Data from 6 deep 2D multichannel seismic (MCS) lines, 1 wide-angle seismic transect and gravity were used to investigate the crustal structure and stratigraphy of the southern Baja California peninsula and its margins. An array of air guns was used as seismic source shooting each 50 m. Each signal was recorded during 16 s by a 6 km long streamer with 480 channels and a spacing of 12.5 m. Seismic waves were also recorded by Ocean Bottom Seismometers (OBS) in the Pacific and the Gulf of California and by portable seismic instruments onshore southern Baja California. MCS data were conventionally processed, to obtain post-stack time-migrated seismic sections. We used a direct method for the interpretation of the wide-angle data, including ray tracing and travel times calculation. In addition to the gravity data recorded onboard, satellite and land public domain data were also used in the gravity modeling. The combined MCS, wide-angle and gravity transect between the Magdalena microplate to the center of Farallon basin in the Gulf of California, crossing the southern Baja California Peninsula to the north of La Paz, allows to verify the existence of the Magdalena microplate under Baja California. We have also confirmed an extensional component of the Tosco-Abreojos fault zone and we have calculated crustal thicknesses. We have also observed the continuation to the south of the Santa Margarita detachment. The MCS seismic sections show a number of fault scarps, submarine canyons and grabens and horsts associated to normal faults offshore southern Baja California peninsula. The normal displacement observed in the Tosco-Abreojos fault zone and some basins in the continental platform, as well as the presence of faulted acoustic basement blocks, evidence that not all extension was accommodated by the Gulf Extensional Province during the middle to late Miocene. Part of the extension was (and is) accommodated in the Baja California Pacific margin. This confirms the observations from previous seismic lines that suggest that the peninsula is a tectonic block not completely transferred to the Pacific plate. In agreement with the seismic facies and the correlations with the available stratigraphic columns of Deep Sea Drilling Program 471 and 474, we generally identify at least three seismostratigraphic units over the acoustic basement. The lower unit reflectors dip towards the palaeo-trench. We identified a Bottom Simulating Reflector (BSR) probably associated to the presence of gas hydrates, which extends at least 200 km along three seismic lines.
S31A-04
Seismic Structural Setting of Western Farallon Basin, Southern Gulf of California, Mexico.
Data from a number of high resolution 2D multichannel seismic (MCS) lines were used to investigate the structure and stratigraphy of the western Farallon basin in the southern Gulf of California. A Generator-Injector air gun provided a clean seismic source shooting each 12 s at a velocity of 6 kts. Each signal was recorded during 6- 8 s, at a sampling interval of 1 ms, by a 600 m long digital streamer with 48 channels and a spacing of 12.5 m. The MCS system was installed aboard CICESE's (Centro de Investigacion Cientifica y de Educacion Superior de Ensenada) 28 m research vessel Francisco de Ulloa. MCS data were conventionally processed, to obtain post- stack time-migrated seismic sections. The MCS seismic sections show a very detailed image of the sub-bottom structure up to 2-3 s two-way travel time (aprox. 2 km). We present detailed images of faulting based on the high resolution and quality of these data. Our results show distributed faulting with many active and inactive faults. Our study also constrains the depth to basement near the southern Baja California eastern coast. The acoustic basement appears as a continuous feature in the western part of the study area and can be correlated with some granite outcrops located in the southern Gulf of California islands. To the East, near the center of the Farallon basin, the acoustic basement changes, it is more discontinuous, and the seismic sections show a number of diffracted waves.
S31A-05
Structure of the Wagner Basin in the Northern Gulf of California From Interpretation of Seismic Reflexion Data
The northern Gulf of California straddles the transition in the style of deformation along the Pacific-North America plate boundary, from distributed deformation in the Upper Delfin and Wagner basins to localized dextral shear along the Cerro Prieto transform fault. Processing and interpretation of industry seismic data adquired by Petroleos Mexicanos (PEMEX) allow us to map the main fault structures and depocenters in the Wagner basin and to unravel the way strain is transferred northward into the Cerro Prieto fault system. Seismic data records from 0.5 to 5 TWTT. Data stacking and time-migration were performed using semblance coefficient method. Subsidence in the Wagner basin is controlled by two large N-S trending sub-parallel faults that intersect the NNW-trending Cerro Prieto transform fault. The Wagner fault bounds the eastern margin of the basin for more than 75 km. This fault dips ~50° to the west (up to 2 seconds) with distinctive reflectors displaced more than 1 km across the fault zone. The strata define a fanning pattern towards the Wagner fault. Northward the Wagner fault intersects the Cerro Prieto fault at 130° on map view and one depocenter of the Wagner basin bends to the NW adjacent to the Cerro Prieto fault zone. The eastern boundary of the modern depocenter is the Consag fault, which extends over 100 km in a N-S direction with an average dip of ~50° (up to 2s) to the east. The northern segment of the Consag fault bends 25° and intersects the Cerro Prieto fault zone at an angle of 110° on map view. The acoustic basement was not imaged in the northwest, but the stratigraphic succession increases its thickness towards the depocenter of the Wagner basin. Another important structure is El Chinero fault, which runs parallel to the Consag fault along 60 km and possibly intersects the Cerro Prieto fault to the north beneath the delta of the Colorado River. El Chinero fault dips at low-angle (~30°) to the east and has a vertical offset of about 0.5 seconds (TWTT). Seismic imaging indicates that the Wagner and Consag faults transfer most of their slip to the Cerro Prieto fault. Moreover, the 130° intersection between the Wagner and Cerro Prieto faults suggests that the Wagner fault has a significant strike-slip component. Our results indicate that most of the strain in this plate boundary is transferred along two main sub-parallel oblique faults in a narrow zone 35 km-wide.
S31A-06
A magnetotelluric transect through the Sierra San Pedro Martir, northern Baja California, Mexico
Preliminary results of a magnetotelluric transect through the Sierra San Pedro Martir, in northern Baja California, Mexico, reveal a high conductivity anomaly possibly associated with a shear zone developed during a major accretion episode in Cretaceous time. Enhancement of electrical conductivity in the rocks of the upper crust highly depends on the presence of fluids and/or conductive minerals, like those produced by metamorphism. Hence, the assessment of subsurface conductive anomalies may contribute to understand the tectonic evolution of Baja California peninsula. We use series and parallel magnetotelluric invariant impedances and a 2-D inversion technique to obtain resistivity models of the crust. The search for the best resistivity model is based on a regularization scheme, where a balance between data misfit and model roughness is looked for. At this stage, we have only a 60 km-long transect across the Peninsular Ranges batholith, but in the close future it will be completed to include measurements toward the east, across the western end of the Gulf of California extensional province.
S31A-07
Seismic Anisotropy of the Crust in the Regions of Northern Baja California and Sonora, Mexico, Using Receiver Functions
We calculated receiver functions from teleseismic P waves recorded at the NARS-Baja broad band stations to investigate the seismic anisotropy of the crust in the regions of northern Baja California and Sonora. Seismic anisotropy is typically controlled by stress state in the fragile shallow crust and by the rock strain state in the ductile intermediate and deep crust. To emphasize phases produced by the anisotropic structure of the crust we partially remove the source effect performing a frequency domain deconvolution of the radial and transverse components by the vertical one. In order to enhance coherent information and decrease noise, we stack receiver functions obtained from P-wave arrivals with similar back azimuth, since they are expected to yield consistent receiver functions. We gave particular attention to the Ps converted phase arrival, which is the most prominent wave arrival after direct P phase and is generated by the P to S conversion at the Moho due to the sharp velocity drop. When a Ps arrival appears clearly on radial and transverse components, and does not overlap with other phases, we suppose that transverse energy rose from shear wave splitting and we retrieve apparent anisotropy parameters, namely the azimuth of the fast direction and the time delay using a method based on the calculation of the covariance matrix. Using this method we characterized the crustal anisotropy under the NARS-Baja station NE75, located in the Stable Central Peninsula Province. The results obtained at this station suggest that the anisotropy of the crust must be related to the metamorphism induced by the subduction of the Farallon plate. At other stations, the results show high dispersion and back azimuthal dependency, indicating the presence of a more complex structure under the stations analyzed. We also performed forward modeling of the crust structure by matching the receiver functions obtained from observed records with synthetic seismograms calculated using a reflectivity based algorithm. At station NE75, a reasonable fit is achieved by introducing an anisotropic layer in the lower crust. A second anisotropic layer is also required in the mantle lithosphere near the Moho. This feature of the model suggests that metamorphic rocks were probably produced by the release of volatiles from the subducted Farallon plate. At station NE71 in the Transpeninsular Strike Slip Province, the anisotropy appears to lie mainly beneath the Moho. In contrast, stations NE80 and NE81 located in the Mexican Basin and Range Province show a substantial amount of crustal anisotropy, which is sufficient to account for the main features in the receiver functions calculated with the observed seismograms.
S31A-08
Seismicity in the Gulf of California Region Recorded by the NARS-Baja Array: Preliminary Results
We report the epicentral coordinates of small to moderate size earthquakes recorded by the broadband stations of the Network of Autonomously Recording Seismographs (NARS)-Baja array between April 2002 and August 2006. The NARS-Baja array consists of 19 stations installed in Baja California and Sonora, Mexico. We initially searched for earthquakes reported by the PDE in the Gulf of California region for the period of interest and we found events with magnitudes ranging between 3.2mb and 6.7Ms. Then we relocated those events using P and S wave arrivals recorded by the NARS-Baja array and the RESBAN array (a broadband seismic array with additional stations located in Sinaloa, Sonora and Baja California, Mexico). 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 steps in the NE direction near the Guaymas and Tepoca basins. 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 3.2-5.0mb differ on the average by as much as 43 km. For the magnitude range between 5 and 6.7 the discrepancy is less, differing on the average by about 25 km. To evaluate the energy released by the earthquakes analyzed, we summed the seismic energy released by individual events dividing the region in bins of three degrees wide between latitudes 23ºN and 31ºN. We found that for the time period considered, most of the seismic energy was released in the middle part of the Gulf of California between 27ºN and 29ºN.
S31A-09
Numerical Models of Oblique Rifting: Application to the Gulf of California
The modern oblique rift system in the Gulf of California was preceded by a stage of orthogonal rifting. The change to oblique rifting occurred around break up, at ca. 6-8 Ma. Pre-breakup rift segments are found in the northeastern part of Baja California. They are characterized by alternating rift asymmetry and half graben systems, similar to structures of the western branch of the East African Rift system. Numerical models of continental lithosphere extension that study the rifting process when the extension direction changes from orthogonal to oblique, show that these characterizations are typical for oblique segments. Numerical results show that the segments that are formed during oblique extension are oriented more or less orthogonal to the oblique spreading direction and show an alternating asymmetric pattern. Together the segments follow the zone of deformation, but individually they show an en echelon or stairway pattern. When extension continues, the numerical models show that some segments become the loci of mantle upwelling and magma formation. This suggests a genetic relationship with the Gulf of California oceanic spreading ridges that show a similar segmentation; the distinct basins where deformation has localised may have provided a mould for initial mid-oceanic ridge segments. After break up, an oblique system is predicted with orthogonal spreading at the ridges, offset by long transform fault systems, much as the present-day Gulf of California rift system.
S31A-10
Transition From Proto-Gulf Extension to Transtension, Coastal Sonora, Mexico
Transtensional deformation during the proto-Gulf time period (12-6Ma) may have operated to focus lithospheric rupture into the Gulf of California. Currently, it is unknown whether this oblique rupture was preceded by either: 1) partitioned deformation of dextral strike-slip on the Tosco-Abreojos fault zone (TAFZ) west of Baja California and E-W extension in the Gulf Extensional Province, or 2) a zone of distributed shear along en echelon, dextral, continental, strike-slip faults, similar to the modern plate boundary. Once the onset of transtension is recognized in the Gulf, we can infer whether the timing and the style of dextral deformation played an integral role in lithospheric necking. We have identified two well-exposed sedimentary basins in coastal Sonora that record both proto-Gulf and modern Gulf history within syn-tectonically emplaced volcanic and sedimentary rocks. These basins contain two regionally extensive pyroclastic flows that serve as marker units: the ~12.6Ma Tuff of San Felipe (Tsf) and the 6.3±0.2Ma Tuff of Mesa Cuadrada (Tmc). These markers bracket basin fill deposited pre-extension (pre-12Ma), during the proto-Gulf period (12-6Ma), and during the modern Gulf period (post-6Ma). Average dips of beds shallow from 74° in pre-extension rocks (n=20), to 56° in Tsf (n=48), to 34° in modern Gulf rocks (n=43). Modern Gulf conglomerates unconformably overlie Tmc, which in turn lies unconformably on Tsf, suggesting continuous exposure and erosion of older proto-Gulf rocks during modern Gulf time. Fault kinematics (n=22) in pre-extension rocks, which integrate all subsequent Gulf deformation, display ENE-directed (azimuth ~261°) extension with a near-vertical σ1 principal stress, after 25°±5° correction for clockwise vertical axis rotation. Fault kinematics (n=8) in modern Gulf conglomerates are dominantly strike-slip and display NW-directed transtension (azimuth ~314°), with a σ1 principal stress that plunges 40° to the NE. The azimuth of this transtension could be more westerly after correction for clockwise vertical axis rotation. Paleomagnetic analysis of Tmc and Tsf, in progress, will quantify the timing of vertical axis rotation coeval with basin development.
S31A-11
Tectonic Geomorphology of an Active Low-Angle Normal Fault, Sierra El Mayor, Northern Baja California
Low angle normal faults (LANF) are ubiquitously distributed throughout the northern Gulf of California. They commonly bound uplifted mountain ranges and are found in numerous seismic sections in the Altar Desert and Wagner Basin (A. Martin, unpublished data). The Canada David detachment (CDD) is a spectacular example of an active LANF that controls the western mountain front of Sierra El Mayor over a strike length of 60 Km. Like most LANFs, the CDD contains two prominent antiform-synform megamullion pairs that strongly control the tectonic geomorphology of the uplifted footwall block and alluvial terraces along the range flank. Quantitative morphometric analysis along the mountain front shows that drainage basins in antiformal domains have systematically higher outlet elevations, higher gradients, greater relief, and much greater hypsometric integrals. Additionally river valleys are narrower and dominated by bedrock channels that extend nearly to the outlet, which is consistent with the fact that mountain front sinuosity is almost an order of magnitude less in the antiformal domains. A sequence of as many as 8 different regional strath terraces are preserved along the range flank and reconnaissance dating of the deposits by cosmogenic isotopes suggests that they formed during the major interglacial-to-glacial climatic transitions. Strath terraces are generally much older, and relative heights between terraces is significantly lower in synformal domains. All of these geomorphologic characteristics suggest that the synformal domains have experienced much lower rates of uplift and erosion of the footwall and likewise lower rates of sedimentation in the adjacent hanging wall basin. The lack of slip gradients on the master fault between synformal and antiformal domains suggests that the megamullions formed instead by regional buckling perpendicular to the extension direction. A Quaternary scarp array extends along the entire length of the mountain front and also shows remarkable variations with megamullions. In antiformal domains, Quaternary scarps generally lie within 100 m of the trace of the CDD. The array is generally wider and contains numerous antithetic scarps that accommodate high ratios of horizontal to vertical deformation. In contrast, the scarp array in synformal domains is well removed (3.5-10 km) from the recessed trace of the CDD and is characterized by a narrow band of synthetic scarps. Our data are consistent with a basinward migration of deformation as envisioned in the "rolling hinge" model of detachment faulting, but we conclude that such a migration occurs predominantly in synformal domains as the fault adopts a straighter configuration.
S31A-12
New contraints on Rivera-Pacific relative motion from multibeam bathymetric data along the MSS and Rivera Transform
To better constrain the Euler pole describing the recent motion of the Pacific plate relative to the Rivera plate, total- field magnetic data, multi-beam bathymetric data and sidescan sonar images were collected during the BART and FAMEX campaigns of the N/O L'Atalante conducted in April and May of 2002 in the area surrounding the Moctezuma Spreading Segment of the East Pacific Rise, located offshore of Manzanillo Mexico at 106°16'W and between 17.8°N and 18.5°N, and the adjacent Rivera Transform. Better constraining Rivera-Pacific relative motion is crucial for understanding the geologic forces and associated hazards present in western Mexico. Among the main results are: (1) the principle transform displacement zone of the Rivera Transform is very narrow and well defined east of 107° 15'W and these azimuths should be used preferentially when determining plate motions in this area, and (2) spreading rates along the Moctezuma Spreading center should not be used in plate motion studies as either seafloor spreading has been accommodated at more than one location since the initiation of seafloor spreading in the area of the Moctezuma Spreading Segment, or this spreading center is not a Rivera-Pacific plate boundary as has been previously assumed. Comparison of observed transform azimuths with those predicted by six previous models of Rivera- Pacific relative motion indicate the need to redefine the current model of Rivera-Pacific relative motion. Thus, using these results, a new plate motion model, RP2005-2, is presented which better describes Rivera-Pacific relative motion occurring during the past several hundred thousand years.
S31A-13
High precision finite-differences time-domain direct modelling of wave equation for seismic oceanography experiments
Holbrook et al. (2003) demonstrated recently the possibility of visualizing fine structures in the water column, like thermohaline intrusion or internal waves, through seismic exploration experiments. Seismic exploration is becoming a popular technique for providing high-lateral resolution images of the explored area, in contrast with the classical oceanography probes, like XBT or XCDT. In this work we present a wave propagation model based upon a high order finite-differences time-domain (FDTD) scheme which includes special absorbing conditions in the boundaries. FDTD algorithms are known for presenting problems with reflections on the computational edges. Classical boundary conditions, like those of Engquist, provide reflection coefficients or the order of 10-2. However, reflection coefficients of fine structures in the water we are trying to model are about 10-4. Thus, the key point of the algorithm we present is in the implementation of Perfectly Matched Layer (PML) boundary conditions. These consist in zones with high absorption (therefore, very low reflection coefficient). The PML implemented in this scheme consists in a second order algorithm in the time domain, to take advantage of its stability and convergence properties. In this work we specify the propagation algorithm, and compare it results with the with Engquist and PML absorbing boundaries conditions. The PML condition affords reflection coefficients in the numerical edges lower than 10-4. Holbrook, W.S., Paramo, P., Pearse, S. and Schmitt, R.W., 2003. Thermohaline fine structure in an oceanographic front from seismic reflection profiling. Science, 301, 821-824.