Relationships Between Lithospheric Structures and Rifting in the East African Rift System: A Rayleigh Wave Tomography Study. (18th October 2018)

• Record Type: Journal Article
• Authors: Adams, A.; Miller, J.; Accardo, N.
• Is Part Of: Geochemistry, geophysics, geosystems. Volume 19:Number 10(2018)
• Page Start: 3793
• Page End: 3810
• Publication Date: 2018-10-18
• Subjects: East Africa -- Africa -- rifting -- tomography -- lithosphere -- craton; Geochemistry -- Periodicals; Geophysics -- Periodicals; Earth sciences -- Periodicals; 550.5
• DOI: 10.1029/2018GC007750
• Languages: English
• ISSNs: 1525-2027

Abstract

Abstract: The East African Rift System (EARS) provides a unique location for exploring factors influencing the development and maturation of continental rifting. In particular, the geographical relationships between Cenozoic rifts and Pre‐Cambrian lithospheric structures suggest that such preexisting structures exert an influence on early‐stage rift geometry and behavior. This study uses Rayleigh wave phase velocity at periods of 20 to 100 s to study lateral variability in the lithospheric structures of rift segments and preexisting structures in the central and southern EARS. The model is constructed using records of 789 earthquakes, recorded by a composite station array of 235 stations from nonconcurrent seismic networks between 1994 and 2015. In the central EARS, we observe fast velocities beneath the Tanzania Craton, isolated low‐velocity regions along the Western Rift Branch, and low velocities in all resolved portions of the Eastern Rift Branch, consistent with previous regional surface wave studies. South of the Tanzania Craton, we observe linear low‐velocity zones trending both southeast and southwest from the Tanzania Divergence Zone, suggesting a southern bifurcation of the Eastern Rift Branch. In the southern portions of the Western Rift Branch, the Malawi Rift borders fast velocities associated with the Bangweulu Block and Irumide Belt. Anomalously fast velocities in these regions persist to long periods, confirming the existence of cratonic lithosphere inferred from previous studies. Fast velocities observed beneath the Irumide Belt extend across the southernmost portion of the Malawi Rift, suggesting that strong lithosphere in this region may hinder the southern propagation of the rift. Plain Language Summary: The East African Rift System (EARS) is the best place to study how Earth's tectonic plates start to break and drift apart. The rifts here are very young and behave very differently than the better known mid‐ocean rifts. But the EARS is not a straight line and breaks into fingers, which form in weaker parts of Earth's crust and upper mantle (called the lithosphere). The goal of this study is to determine how rifts form in relation to older lithospheric features. We use waves from distant earthquakes to take pictures of the inside of the Earth, showing where those waves travel at different speed. In our images of wave speed, we see two very interesting things. First, we see slow speeds in an area where few rifts are seen at Earth's surface, suggesting that the lithosphere is hot there—or perhaps contains some melt. So, this may show the development of new rift fingers around older pieces of strong lithosphere. Second, we see fast speeds in the southern part of the rift system, which extend across the southernmost part of the EARS (where we expect slow wave speeds). This could mean that there is a block of strong lithosphere hindering the southward continuation of the rift system. Key Points: Low velocities are found along the proposed northern boundaries of the Rovuma microplate, suggesting that channelization in the mantle is leading to early‐stage rifting Fast velocities extend through the Bangweulu Block, the Irumide Belt, and across the southern Malawi Rift, suggesting that strong previously existing lithosphere is influencing the southward propagation of the East African Rift System In areas where data sets overlap, the Automated Generalized Seismic Data Function Method reproduces phase velocity results from previous, more labor‐intensive, algorithms even when using nonideal data sets