Lower crustal seismicity, volatiles, and evolving strain fields during the initial stages of cratonic rifting

Abstract

The volcanically active East African rift system in southern Kenya and northern Tanzania transects thick cratonic lithosphere, and comprises several basins characterized by deep crustal seismicity. The US-French-Tanzania-Kenya CRAFTI project aims to understand the role of magma and volatile movement during the initiation and evolution of rifting in cratonic lithosphere. Our 38-station broadband network spans all or parts of fault-bounded rift segments, enabling comparison of lithospheric structure, fault kinematics, and seismogenic layer thickness with age and proximity to the deeply rooted Archaen craton. Seismicity levels are high in all basins, but we find profound differences in seismogenic layer thickness along the length of the rift. Seismicity in the Manyara basin occurs almost exclusively within the lower crust, and in spatial clusters that have been active since 1990. In contrast, seismicity in the ~ 5 My older Magadi basin is localized in the upper crust, and the long border fault bounding the west side of the basin is seismically inactive. Between these two basins lies the Natron rift segment, which shows seismicity between ~ 20 and ~2 km depth, and high concentrations at Oldoinyo Lengai and Gelai volcanoes. Older volcanoes on the uplifted western flank (e.g., Ngorongoro) experience swarms of activity, suggesting that active magmatism and degassing are widespread. Focal mechanisms of the frequent earthquakes recorded across the array are spatially variable, and indicate a stress field strongly influenced by (1) Holocene volcanoes, (2) mechanical interactions between adjacent rift basins, and (3) a far-field ESE-WNW extensional stress regime. We explore the spatial correlation between zones of intense degassing along fault systems and seismicity, and examine the influence of high gas pressures on lower and upper crustal seismicity in this youthful cratonic rift zone.