T31F-01
The Effect of Sedimentation on Extensional Style: Applications to the Gulf of California
We examine the effect of deposition of large amounts of non-locally derived sediment on extensional style. Our model differs from previous work with sedimentation and extension as it does not erode material off of rift flanks but deposits material from a third, lateral dimension, in the manner of a prograding delta, making this what is sometimes referred to as a 2.5 dimensional model. Depending on initial model conditions the onset of rifting in thick continental crust occurs in either the narrow rifting, wide rifting, or core complex mode. With continued extension, all cases eventually transition to a narrow rifting mode. We proposed that the weight of the sediments reduces the crustal buoyancy effect caused by local crustal thinning, allowing the system to extend more easily in a narrow rift mode. Two dimensional, regional scale numerical experiments of extending a plateau, represented in the models as an area of thick continental crust surrounded by areas of crust with normal thickness, are used to test this hypothesis. During extension the sediments depress the extended crust and reduce the amount of crustal flow from the still thick edges. Simultaneously, the sediments thicken the thinned area, allowing for further extension. For a system in wide or core complex mode, sedimentation would shorten the time needed to transition to a narrow rift. Sedimentation rate, extension rate, the concentration of radiogenic elements, plateau width, plateau thickness, and sediment density were varied across a suite of models to examine under what conditions sedimentation can influence the mode of extension in this way. Results indicate that sedimentation must keep up with crustal thinning at a site of concentrated extension in order to shut down the lower crustal flow and transition to a narrow rift earlier than if no sedimentation were present. A comparison is made between the numerical and analytical results to seismic data from and the geological history of the Gulf of California, where a similar system may be unfolding today.
T31F-02
Synchronous changes in the rift-margin San Jose Island basin and initiation of the Alarcón spreading ridge: implications for rift to drift transition in the Gulf of California
The rift to drift hypothesis is widely cited, but it well known in detail. The low sedimentation rate and recent rifting of the Gulf of California provides insight into the rift-to-drift process. Lizarralde et al. (2007) showed that the style of rifting, based on crustal structure, varies significantly between the central and southern Gulf of California, and this combined with the analysis of sedimentary basins shows the small-scale (~15 km) complexities of the rift-to-drift transition. The shut off of rifting on the eastern side of the plate boundary occurred at ca. 2 – 3 Ma (Aragon-Arreola etal, 2005, Aragon-Arreola & Martin-Barajas, 2007; our unpublished data). Many studies have shown that the western side of the Gulf is still active despite sea-floor spreading occurring on the Alarcón and other short spreading centers since 2 - 3 Ma. At the mouth of the Gulf, magnetic anomalies on the eastern side of the Alarcón rise show that it appears to have changed to seafloor spreading as early as 3.7 Ma. But comparatively, on the eastern side, magnetic anomalies do not indicate the formation of new oceanic crust until 2.5 Ma, so spreading was first fully established at 2.5 Ma. The San Jose Island basin (Umhoefer et al., 2007) began at approximately 4- 6 Ma; the basin had its most rapid subsidence, with faulting accompanying marine sedimentation, from 3.6 ± 0.5 Ma (Ar tuff age) to 2.5-2.4 Ma (forams). Basin margin faulting died and moved east (offshore) shortly after 2.5-2.4 Ma. Late Quaternary marine terraces suggest that faulting rates slowed by 1-2 orders of magnitude since the fault reorganization at 2.5 Ma. These observations suggest that the rift – drift transition started, but is not yet finished, on the western side of the Gulf of California, with low rates of faulting (<1? mm/yr) continuing on the continental margin for reasons that are not well understood. Our work highlights the importance of combining onshore field and MSC data and analyzing entire conjugate rifted margins to accurately assess rifting processes.
T31F-03
The Sills of Guaymas Basin, Gulf of California: Implications for Crustal Melt Distribution and Thermogenic Carbon Flux
Multi-channel (MCS) and wide-angle seismic data acquired in 2002 reveal the distribution of igneous sills emplaced into sediments and their relationship to the deeper crustal structure of Guaymas Basin, a young (<6 m.y.), heavily sedimented rift basin. The transition from extended continental to oceanic crust is abrupt across the basin's margins, and most of the crust underlying Guaymas Basin is new igneous crust that is essentially oceanic. However, as a consequence of high sedimentation rates (1-2 km/m.y.), it does not have a typical oceanic Layer 2, but instead has an upper layer of intercalated sediments and igneous intrusions. The MCS data reveal shallow (i.e. recent) sills up to 50 km away from the ~200-m-deep seafloor grabens that delineate the two spreading segments of Guaymas Basin, with no systematic relationship between sill depth and distance from the grabens. For example, sills 42 km away from the graben (1.75 m.y. spreading age) lie beneath only 50 m (or 50 k.y.) of undisturbed sediments. Thus, ongoing sill emplacement is distributed over tens of km relative to the kinematic plate boundary, and the grabens do not mark the locus of focused shallow magmatic emplacement. This implies that the distribution of deeper melt is correspondingly broad, consistent with the deeper velocity structure. Magmatic emplacement in this and perhaps other young, sedimented rift system is thus substantially different from that of unsedimented mid-ocean ridges, where hydrothermal circulation tends to focus magmatic accretion to within <10 km of the ridge axis. The MCS data also reveal 50- to 150-m-thick regions of intense sediment disruption above intruded sills. These regions probably indicate induration from alteration processes accompanying sill intrusion, including the release of thermogenic methane and carbon dioxide. New sills appear to be emplaced at the boundary between high porosity uppermost sediments and the more dense indurated sediments. Active sill intrusion over a much broader region than previously thought greatly increases the predicted carbon flux into the ocean and atmosphere from this and similarly sedimented rift basins.
T31F-04 INVITED
Capturing Magma Intrusion and Faulting Processes During Continental Rupture: Seismicity of the Dabbahu (Afar) rift
Continental rupture models emphasize the role of faults in extensional strain accommodation; magma intrusion is commonly overlooked. The volcanically and seismically active Afar depression of Ethiopia is a broad zone of extensional strain encompassing zones of incipient plate rupture. One of these highly extended rift segments entered a phase of active rifting in September, 2005. Earthquakes recorded on a temporary seismic array October 05 - April 06, direct observation of fault patterns, and geodetic data document ongoing strain and continued dyke intrusion along the ~60 km-long Dabbahu rift segment defined in earlier remote sensing studies. The depth distribution of microseismicity indicates that the dyke rises from ~15 km to <3 km from the surface along the ~60 km-long length of the tectono-magmatic segment. Focal mechanisms show rift- normal opening. The seismicity, InSAR and structural patterns all suggest that magma injection continued at least 3 months after the main episode; the mid-crustal level of the seismogenic zone suggests the deep roots of the dike may have remained open during this time. Persistent seismic swarms at 2 sites on Dabbahu volcano coincide with areas of deformation identified in the processed InSAR data: (1) a northwestward-dipping zone of seismicity interpreted as a listric fault that soles out at 8 km coincides with a broad zone of subsidence, and (2) a more diffuse, 8 km-radius zone of shallow seismicity (<2 km), a shadow zone between 2.5 and 6 km, and a narrow ‘tube' of events between 14 and 6 km. InSAR data show uplift above a shallow source in zone (2). The patterns of seismicity provide a 3D perspective of magma feeding systems maintaining the along-axis segmentation of this incipient seafloor spreading segment surrounded by transitional continental crust.
T31F-05
Melt distribution in the Ethiopian rift system: Constraints from seismic observations and modelling
Seismic observations from the EAGLE experiment in the Main Ethiopian Rift have been interpreted in terms of melt-induced anisotropy and support ideas of magma-assisted rifting in continental regions. Following the 2005 Dabbahu rifting event in Afar a further 9 broadband seismometers were installed around the newly active rift segment. These recorded more than one year of continuous data and shear-wave splitting observed in core phases (SKS/SKKS) shows considerable variability across the array. Three stations centred above the Dabbahu rift segment show markedly different splitting characteristics from the other stations. The fast direction is oriented roughly north/south and parallel to the Dabbahu magmatic segment, compared to NNE/SSW orientations at nearby stations. Also the magnitude of splitting is slightly larger at the rift stations compared to those nearby (~1s compared to 0.7-0.9s). These observations supports previous work in the Main Ethiopian Rift (MER), where fast directions change abruptly from being rift parallel on the rift flanks to magmatic-segment parallel in the rift valley. Furthermore, observations of frequency-dependent splitting in the data further suggest that the underlying cause of the anisotropy is related to aligned melt inclusions. Poroelastic modelling support mechanisms for melt-induced anisotropy due to vertically-aligned melt pockets that are on the order of centimetres in length scale. The abrupt change in splitting parameters over small lateral distances (~ 30 ° over ~30~km) suggests that the source of anisotropy is shallow. To further constrain the location of the anisotropy and study the influence of the rift transition on shear-wave splitting results, we model finite-frequency waveforms for a suite of model representations of the rift zone. In each model, the orientation of the anisotropic fabric varies laterally (i.e., the symmetry axis of the HTI symmetry rotates between 0° and 30°), but the strength of anisotropy and depth of transition differs. Waveforms are modeled using a narrow-angle one-way elastic vector wave equation to simulate finite-frequency waveform effects for an incident near-planar S-wave. The modelling is used to examine the influence of changing anisotropic symmetry across the rift as a function of wavefront curvature, and strength and depth of transition as well as lateral width of transition zone. The results show how a simple model with two regimes of anisotropy can explain the variability across the rift, in both delay time and shear-wave polarization, over short length scales.
T31F-06
Combined Seismic and Geodetic Observations of Dike Injection in the Dabbahu Rift Segment, Afar, Ethiopia
As rifting proceeds to breakup, strain localizes to discrete, narrow rift segments by a combination of diking and faulting. However, the controls on the kinematics and dynamics of rifting episodes are poorly understood. We integrate new seismic and geodetic data from the Dabbahu rift segment, Afar, to constrain the location of magma sources, and the spatial/temporal variability of diking and faulting just prior to the onset of sea-floor spreading. Since 4Ma, strain in north Afar has localized to 60-km-long rift segments that are similar in size and morphology to slow-spreading mid-ocean ridge segments. In June2006 vigorous deformation occured near AdoAle volcano, in the center of the Dabbahu rift segment. InSAR shows 2.5m opening along a 10km-long zone, with no evidence of deflation of shallow chambers beneath central volcanoes. Deformation is consistent with injection of a 2m- wide dike at 1-10km depth. Seismic and GPS data constrain the spatial/temporal evolution of deformation. Starting midday 17th June, ML<3 earthquakes occur at the southern end of the deforming zone. After 1hr, earthquake merge into tremor, increase in magnitude, and migrate 10km NNW over 4hrs. Earthquakes are likely caused by inflation induced tension at the tip of a laterally propagating dike. Prior to June, broad deformation is observed beneath AdoAle, consistent with inflation of a source zone 10-15km deep beneath the central portion of the rift segment. The June06 dike is the first of 6 major intrusions in the Dabbahu rift since Sep05, a pattern similar to the 1975-84 rifting event in Krafla, Iceland. Our results show that during continental rupture strain is accommodated by repeated injection of dikes and induced faulting. Magma may be sourced directly to the center of the rift segment from reservoirs in the upper mantle and/or lower crust and delivered into the upper 10km by the lateral injection of discrete dikes.
T31F-07
Seismological Investigation of Rift Related Uplift: The Rwenzori Network in Western Uganda
We present first results from a temporary seismic network located within the western branch of the East-African rift system. The project aims to constrain the development and uplift of the Rwenzori mountain range and its relation to the formation of the rift zone. Local and teleseismic earthquake recordings are used to image structures of the crust and upper mantle within the region. A temporary network consisting of 23 mobile broadband and short-period seismic sensors was operating during a period of about 16 months until October 2007. The stations were located mainly along two profiles in the westernmost part of Uganda near the Congo border. One profile was situated within the rift along the eastern flank of the Rwenzori Mountains, approximately between 0.2°S and 0.9°N. The second profile was located nearly perpendicular to the rift, extending from the eastern rift shoulder into the rift valley and further crossing the northern part of the Rwenzori Mountains. Additional seismic stations were placed on the rift shoulder in the South-East and the North-West to improve the localization of events. A number of seismological methods are used to study crustal and upper-mantle structure. The localization of sources and fault-plane solutions provide information on active fault zones and on current tectonic movements. Local and teleseismic tomography are applied to determine the 3D velocity structure in the area under investigation and to detect velocity anomalies in the crust and the upper mantle down to approximately 300 km depth. Moho depths are derived from converted phases using receiver functions. Furthermore, shear wave splitting is analyzed to detect regions of anisotropy and their relation to deformation processes and mantle flow. The first analysis of the data shows high seismic activity of the Rwenzori region. Approximately 500 events per month have been located. Focal depths are concentrated between 10 and 20 km. Local P and S-wave tomography using data of the first few months shows significant low velocity anomalies in the northern part of the network, correlated with active volcanic fields and hot springs. From receiver functions, crustal thicknesses of about 23 km have been found beneath the eastern flank of the rift. The interpretation of our results will provide constraints for the modeling of geodynamic processes responsible for the formation of the Rwenzori Mountains.
T31F-08
Amagmatic Extension and Thick Oceanic Crust in the Eastern Black Sea.
Volcanic rifted margins exhibit extensive syn-rift magmatism during extension and produce thick ocean crust during initial spreading, whilst magma-poor margins are characterised by limited syn-rift magmatism, highly stretched continental crust and areas of exhumed mantle. Often these two types of rifting can be seen along the same margin (e.g., off eastern Canada), but the nature of the transition from one to the other is poorly understood. New seismic refraction data collected in the Eastern Black Sea (EBS) provide provocative new constraints on this type of transition. Previous studies have revealed a thick sedimentary package (8-9 km) underlain by thin crust in the centre of the basin, but the nature of the crust has remained controversial. In spring 2005, we collected four wide-angle seismic profiles in the EBS basin. Lines 2 and 3 are approximately orientated parallel to the inferred direction of extension and located ~ 200 km apart. Line 1 connects these lines, traversing the centre of the basin and sampling some of the thinnest crust. Along Line 3, the crust thins abruptly from ~ 32 km to ~ 7-9 km in the centre of the basin. Within the thin crust the velocity increases along a shallow velocity gradient of ~ 0.35 s-1 from 5 km/s at the top to 6.5 km/s over a depth of ~ 5 km. The lower crust has a fairly constant velocity of 6.5-6.75 km/s. This velocity structure indicates that the crust along Line 3 is most likely highly stretched continental material. Mantle velocities beneath the crust are ~ 8.0 km/s, and there is no evidence for extensive syn-rift magmatism. Further southeast, the crust along Line 2 is ~ 13 km thick and has two distinct seismic velocity gradients. The top 4 km has a high velocity gradient from 4.25 km/s to 6.0 km/s. The velocity gradient gradually shallows to reach seismic velocities of 7.2-7.5 km/s at the base of the crust. This velocity structure is best interpreted as thick oceanic crust. At the eastern end of Line 1, where it crosses Line 2, the velocity structure also indicates thick oceanic crust extending for ~ 150 km along the line. A transition from oceanic to continental crust appears to occur ~ 60 km to the west of the intersection with Line 2 and coincides with a crustal fault imaged in seismic reflection data and inferred from shipboard gravity data. To the west of this fault the crust is ~ 6 km thick with a constant velocity gradient of ~ 0.35 s-1. The thickness of the crust is fairly constant over a distance of ~ 60 km before thickening slightly and decreasing again to ~ 7 km over a distance of 80 km. Line 3 intersects Line 1 at the western edge of this crustal thickening. Our data suggest that the extension of the EBS basin is greater in the SE than the NW such that rifting in the SE culminated with the onset of seafloor spreading. This supports previous hypotheses that the rift opened by rotation of the Shatsky Ridge away from the Mid Black Sea High. A remarkable feature of the observed crustal configuration is the existence of highly thinned continental crust immediately adjacent to thick oceanic crust. These two crustal domains are separated by an interpreted transform fault, implying that rift segmentation exerts control over the expression of magmatism in the rift or that there were significant along margin changes in mantle temperature and/or composition at the time of rifting.