Current plate motions across the Red Sea

A new objective method is used to estimate precisely 64 rates of seafloor spreading since chron 2A (3.2 Ma) from all the Red Sea magnetic profiles available from 15.9 degrees to 26 degrees N. The fastest spreading rate, approximate to 16 mm yr(-1), occurs near 18 degrees N, whereas the slowest rate, 10 mm yr(-1), occurs at 25.5 degrees N and is consistent with the rate predicted from Arabia-Nubia data to the south. The standard deviation of the spreading rates is 0.8 mm yr(-1), much smaller than the median standard deviation of 4 mm yr(-1) previously assigned to spreading rates in the global plate-motion model NUVEL-1 (DeMets et al. 1990). The fit to the spreading rates, as well as the locations of earthquakes in and near the Red Sea, indicate that spreading south of approximately 17.7 degrees N is less than the full rate of spreading between the Arabian and Nubian plates. The Red Sea spreading centre is instead usefully interpreted as the boundary between the Arabian Plate and a Danakil microplate that includes the subaerial Danakil block and a larger oceanic portion of lithosphere. Despite the absence of reliable azimuths of transform faults in the Red Sea, all components of the angular velocity of Arabia relative to Nubia are usefully constrained from just the 45 relevant spreading rates. The new compact 95 per cent confidence region of the angular velocity excludes prior estimates based on only four and two spreading rates by Chase (1978) and Jestin et al. (1994), respectively. 19 spreading rates in the southern Red Sea are used to estimate the angular velocity between the Danakil microplate and Arabia. An approach based on singular value decomposition shows that without slip vectors only two of the three components of the angular velocity of the Danakil microplate relative to Arabia or Nubia are usefully constrained, but that all three components are usefully constrained if one earthquake slip vector is included.