A revised quantitative, internally consistent, kinematic model has been determined for the present-day relative plate motions in the eastern Mediterranean and Middle Eastern regions, based on a combination of geological and GPS data. The relative motions of the brittle upper crust of the African and Arabian plates across the southern Dead Sea Fault Zone (DSFZ) are represented by relative rotation at 0.278° Ma-1 about an Euler pole at 31.1°N 26.7°E. The resulting predicted slip rate on the southern DSFZ is 4.0 mm a-1. The kinematics of the northern DSFZ are described as relative rotation at 0.243° Ma-1 about the same Euler pole, the difference in rotation rates reflecting the absorption of a small component of the relative motion by distributed shortening in the Palmyra fold belt. The northern DSFZ, in Syria and southern Turkey, is regarded as a series of transpressional stepovers, along which the rate of left-lateral slip is substantially less than the rate of relative plate motion, because this slip is oriented strongly obliquely to the relative motion between the adjoining plates. This geometry seems to result in part from some strands of the northern DSFZ reactivating older fault segments, even though they were not optimally oriented relative to the plate motion, and in part because the ideal initial geometry of the DSFZ, which would have continued northwestward offshore of the Levant coastline towards Cyprus, was precluded by the high strength of the crust along this line. The revised slip rate on the East Anatolian Fault Zone (EAFZ) is estimated as ~8 mm a-1. At this rate, restoring the observed slip requires the age of the EAFZ to be ~4 Ma. The previous phase of deformation, which involved slip on the Malatya-Ovacık Fault Zone before the EAFZ came into being, is thus dated to ~7-4 Ma, suggesting a timing of initiation for the North Anatolian Fault Zone (NAFZ) of ~7 Ma, not ~5 Ma as has previously been thought. Local evidence from the western NAFZ also supports a ~7 Ma or Early Messinian age for the NAFZ. The overall present-day kinematics of the NAFZ are described using the Euler vector determined in 2000 using GPS: involving relative rotation between the Turkish and Eurasian plates at 1.2° Ma-1 about 30.7°N 32.6°E. This Euler vector predicts a rate of relative motion between these plates of ~25 mm a-1, which when extrapolated overestimates the observed amount of localised right-lateral slip, suggesting the existence of a component of distributed right-lateral simple shear in the surroundings to the NAFZ as well. The predicted rate of left-lateral relative motion on the boundary between the Turkish and African plates is estimated as ~8 mm a-1. However, the rate of localised left-lateral slip on the onshore part of this boundary is estimated as only ~2 mm a-1, on the Yakapınar-Göksun Fault: the difference being taken up by distributed deformation within the northern "promontory" of the African plate, which appears to involve a combination of anticlockwise rotation and distributed left-lateral simple shear. It is proposed that this boundary first developed at the same time as the NAFZ, but its original geometry involving left-lateral slip on the Karataş-Osmaniye Fault has since become locked by the presence of relatively strong ophiolitic crust within this fault zone. The kinematic consistency of this model requires one to relax the assumption that brittle upper crust and mantle lithosphere are moving in step, consistent with the assumed presence of a weak layer of lower crust in between. The development of the NAFZ during the Messinian can thus be explained as a consequence of a combination of forces resulting from (a) shear tractions applied to the brittle upper crust of Turkey as a result of relative westward motion of mantle lithosphere, caused by the pre-existing relative motions between the African and Arabian plates during the earlier Miocene; and (b) the reduction in normal stress and increase in right-lateral shear stress that resulted from the dramatic water unloading during the Messinian desiccation of the Mediterranean basin. Analysis indicates that this mechanism requires the effective viscosity of the lower crust of Turkey to be ~5±3x1019 Pa s, consistent with recent estimates in other localities. The well-documented near-total absence of internal deformation within the Turkish plate thus does not result from high strength: it results from the geometry of its boundaries which allow them to slip without any need for internal deformation. The main imperfection in this pattern of boundaries results from the high-strength "patch" on the Turkey-Africa plate boundary in southern Turkey. The seismicity in this locality appears correlated with major earthquakes on the NAFZ, suggesting the possibility that this boundary behaves as a "geometrical lock" whose slip, in moderate-sized earthquakes, can permit much larger amounts of slip in much larger earthquakes on the NAFZ. Future detailed monitoring of this region may thus provide the basis for a system of advance warning of future destructive earthquakes on the NAFZ.

A revised quantitative, internally consistent, kinematic model has been determined for the present-day relative plate motions in the eastern Mediterranean and Middle Eastern regions, based on a combination of geological and GPS data. The relative motions of the brittle upper crust of the African and Arabian plates across the southern Dead Sea Fault Zone (DSFZ) are represented by relative rotation at 0.278° Ma-1 about an Euler pole at 31.1°N 26.7°E. The resulting predicted slip rate on the southern DSFZ is 4.0 mm a-1. The kinematics of the northern DSFZ are described as relative rotation at 0.243° Ma-1 about the same Euler pole, the difference in rotation rates reflecting the absorption of a small component of the relative motion by distributed shortening in the Palmyra fold belt. The northern DSFZ, in Syria and southern Turkey, is regarded as a series of transpressional stepovers, along which the rate of left-lateral slip is substantially less than the rate of relative plate motion, because this slip is oriented strongly obliquely to the relative motion between the adjoining plates. This geometry seems to result in part from some strands of the northern DSFZ reactivating older fault segments, even though they were not optimally oriented relative to the plate motion, and in part because the ideal initial geometry of the DSFZ, which would have continued northwestward offshore of the Levant coastline towards Cyprus, was precluded by the high strength of the crust along this line. The revised slip rate on the East Anatolian Fault Zone (EAFZ) is estimated as ~8 mm a-1. At this rate, restoring the observed slip requires the age of the EAFZ to be ~4 Ma. The previous phase of deformation, which involved slip on the Malatya-Ovacık Fault Zone before the EAFZ came into being, is thus dated to ~7-4 Ma, suggesting a timing of initiation for the North Anatolian Fault Zone (NAFZ) of ~7 Ma, not ~5 Ma as has previously been thought. Local evidence from the western NAFZ also supports a ~7 Ma or Early Messinian age for the NAFZ. The overall present-day kinematics of the NAFZ are described using the Euler vector determined in 2000 using GPS: involving relative rotation between the Turkish and Eurasian plates at 1.2° Ma-1 about 30.7°N 32.6°E. This Euler vector predicts a rate of relative motion between these plates of ~25 mm a-1, which when extrapolated overestimates the observed amount of localised right-lateral slip, suggesting the existence of a component of distributed right-lateral simple shear in the surroundings to the NAFZ as well. The predicted rate of left-lateral relative motion on the boundary between the Turkish and African plates is estimated as ~8 mm a-1. However, the rate of localised left-lateral slip on the onshore part of this boundary is estimated as only ~2 mm a-1, on the Yakapınar-Göksun Fault: the difference being taken up by distributed deformation within the northern "promontory" of the African plate, which appears to involve a combination of anticlockwise rotation and distributed left-lateral simple shear. It is proposed that this boundary first developed at the same time as the NAFZ, but its original geometry involving left-lateral slip on the Karataş-Osmaniye Fault has since become locked by the presence of relatively strong ophiolitic crust within this fault zone. The kinematic consistency of this model requires one to relax the assumption that brittle upper crust and mantle lithosphere are moving in step, consistent with the assumed presence of a weak layer of lower crust in between. The development of the NAFZ during the Messinian can thus be explained as a consequence of a combination of forces resulting from (a) shear tractions applied to the brittle upper crust of Turkey as a result of relative westward motion of mantle lithosphere, caused by the pre-existing relative motions between the African and Arabian plates during the earlier Miocene; and (b) the reduction in normal stress and increase in right-lateral shear stress that resulted from the dramatic water unloading during the Messinian desiccation of the Mediterranean basin. Analysis indicates that this mechanism requires the effective viscosity of the lower crust of Turkey to be ~5±3x1019 Pa s, consistent with recent estimates in other localities. The well-documented near-total absence of internal deformation within the Turkish plate thus does not result from high strength: it results from the geometry of its boundaries which allow them to slip without any need for internal deformation. The main imperfection in this pattern of boundaries results from the high-strength "patch" on the Turkey-Africa plate boundary in southern Turkey. The seismicity in this locality appears correlated with major earthquakes on the NAFZ, suggesting the possibility that this boundary behaves as a "geometrical lock" whose slip, in moderate-sized earthquakes, can permit much larger amounts of slip in much larger earthquakes on the NAFZ. Future detailed monitoring of this region may thus provide the basis for a system of advance warning of future destructive earthquakes on the NAFZ.