STRESS-FIELD CHANGES IN THE AFRO-ARABIAN RIFT SYSTEM DURING THE MIOCENE TO RECENT PERIOD

We have investigated the Miocene to present-day stress fields of the K enyan rift, the northern Red Sea, and the Gulf of Suez by integrating stratigraphically constrained fault movements, alignment of radiometri cally dated dikes and volcanic vents, teleseismic earthquake focal mec hanisms, and borehole breakouts in exploratory wells. Fault analysis a nd dike orientation suggest that in the central Kenyan rift the least horizontal stress direction (S-hmin) from 12 to 6 Ma was oriented NE-S W to ENE-WSW. In the northern Kenyan rift, 4.2-3.8 Ma feeder dikes of the Tirr Tirr trachytes also indicate NE-SW S-hmin orientation. By abo ut 2.6 Ma, faulting in the central Kenyan rift occurred in response to E-W extension. After similar to 0.6 Ma, S-hmin in the central Kenyan rift experienced further clockwise rotation into NW-SE orientation. Al igned vents of the late Quaternary volcanic fields of Mt. Marsabit, Ny ambeni and Hurl Hills, limited teleseismicity, and borehole breakouts in wells drilled to the east of the Kenyan rift, indicate that the pre sent-day S-hmin orientation is approximately NW-SE throughout most of Kenya. In the Gulf of Suez during most of the Middle Miocene-Early Ple istocene, S-hmin was aligned NE-SW, normal to the rift axis. In the La te Pleistocene, but before 125 ka, the Gulf of Suez and northern Red S ea S-hmin rotated counterclockwise into a N-S direction. Extensive bre akout and limited teleseismicity data show that in the Gulf of Suez S- hmin is presently N-S oriented. Our data suggest that local stress fie lds have rotated significantly during much of the Miocene to Recent te ctonic history of the Afro-Arabian Rift System. The best documented st ress field change occurred synchronously between the Tanianian-Kenyan rifts in the south and the northern Red Sea-Gulf of Suez in the north during the Late Pleistocene. Two possible interpretations for this Lat e Pleistocene rotation of the stress field are: (1) a change in the dy namics of the Afro-Arabian Rift System affected this plate boundary ov er a region spanning 4000 km (forces generated within the rift system) , or (2) motions along one of the distant African plate boundaries cha nged dramatically, resulting in stress field changes throughout the pl ate (forces generated outside the rift system). These two models can b e tested by studies of Pleistocene to Rec:ent stress fields outside th e Afro-Arabian Rift System, and along the boundaries of the African pl ate.