Located at the southernmost sector of the Western Branch of the East African Rift System, the Malawi Rift exemplifies an active, magma-poor, weakly extended continental rift. This work focuses on the northern portion of the Malawi Rift, which is flanked by long (>100 km) basin-bounding border faults and crosses several significant remnant structures. This combination of characteristics makes the Malawi Rift the ideal location to investigate the controlling processes governing present-day extension throughout the lithosphere. To investigate these processes I image shallow basin- to uppermost-mantle structure beneath the region using a combination of passive- and active-source seismic datasets. I conduct passive-source imaging of the crust and upper mantle using ambient-noise and teleseismic Rayleigh-wave phase velocities between 9 and 100 s period. This study includes six lake-bottom seismometers located in Lake Malawi (Nyasa), the first time seismometers have been deployed in any of the African rift lakes. I utilize the resulting phase-velocity maps to invert for a shear velocity model of the Malawi Rift discussed below.
I utilize active-source tomographic imaging to obtain new constraints on rift basin structure in the Malawi Rift from a 3-D compressional velocity (Vp) model. The velocity model uses observations from the first wide-angle refraction study conducted using lake-bottom seismometers in one of the great lakes of East Africa. The 3-D velocity model reveals up to ~5 km of synrift sediments, which smoothly transition from eastward thickening against the Livingstone Border Fault in the North Basin to westward thickening against the Usisya Border Fault in the Central Basin. I use new constraints on synrift sediment thickness to construct displacement profiles for both faults. Both faults accommodate large throws (> 7 km) but the Livingstone Fault is ~30 km longer. The dimensions of these faults suggest they are nearing their maximum size. The presence of >4 km of sediment within the accommodation zone suggests fault length was established early pointing the "constant length" model of fault growth. The presence of an intermediate velocity unit with velocities of 3.75-4.5 km/s is interpreted to represent prior rifting (Permo-Triassic and/or Cretaceous) sedimentary deposits beneath Lake Malawi. These thick (up to 4.6 km) packages of preexisting sedimentary strata improve the understanding of the Tanganyika-Rukwa-Malawi rift system and the role of earlier stretching phases on synrift basin development.
I use the previously obtained local-scale measurements of Rayleigh wave phase velocities between 9 and 100 s combined with constraints on basin structure and crustal thickness to robustly invert for shear velocity from the surface to 135 km for the Malawi Rift. We compare our resulting 3-D model to a 3-D model of shear velocity obtained for the mature Main Ethiopian Rift and Afar Depression using commensurate datasets and identical methodologies. Comparing the Vs models for the two regions reveals markedly different seismic velocities particularly pronounced in the upper mantle (average velocities in the Malawi Rift are ~9% faster than the Main Ethiopian Rift). Our 3-D Vs model of the Malawi Rift reveals a strong, localized low velocity anomaly associated with the Rungwe Volcanic Province within the crust and upper mantle that can be explained without requiring the presence of partial melt. Away from the Rungwe Volcanic Province, velocities within the plateau regions are fast (> 4.6 km/s) and representative of depleted lithospheric mantle to depths of 100 and >135 km to the west and east of the rift, respectively. Thinned lithosphere, represented by the absence of similarly high velocities, is centered directly beneath the rift axis and footwall escarpments of the rift basins. The correlation between the localization of lithospheric thinning, the boundaries between abutting Proterozoic mobile belts, and the positions of the basin-bounding border faults may point to the controlling role of preexisting large-scale structures in localizing strain and allowing extension to occur here.
| Identifer | oai:union.ndltd.org:columbia.edu/oai:academiccommons.columbia.edu:10.7916/D89K5PS5 |
| Date | January 2018 |
| Creators | Accardo, Natalie |
| Source Sets | Columbia University |
| Language | English |
| Detected Language | English |
| Type | Theses |
Page generated in 0.008 seconds