Earth Resources Laboratory :: Abstracts

Global Positioning System (GPS) Measurements in Turkey (1988-1992): Kinematics of the Africa-Arabia-Eurasia Plate Collision Zone

M. Burc Oral

Submitted to the Department of Earth, Atmospheric, and Planetary Sciences on March 14, 1994 in partial fulfillment of the requirements for the degree of Doctor of Philosophy

Abstract

Much of the deformation in the eastern Mediterranean is caused by the northerly motion of the African and Arabian plates against the Eurasian plate. As a result, various tectonic features are observed: collision (Bitlis-Zagros, Caucasus), subduction (the Hellenic and Cyprean arcs), strike-slip faulting (the East and North Anatolian and Dead Sea faults) and extension (the North Aegean trough, the Maramara, Gediz, and Büyük Menderes grabens). Most of the constraints for the kinematics of this actively deforming region have been provided by seismological and geological observations. Other than Satellite Laser Ranging (SLR) observations, thee are, as yet, no detailed geodetic measurements that would qualify the rates and style of deformation across active boundaries.

Our challenge was to geodetically observe motions across tectonic boundaries and in the interiors of plates, in order to gain insight into the kinematics of the Arabia-Eurasia collision zone, and to quantify rates of crustal deformation. To meet this challenge we initiated Global Positioning System (GPS) measurements in Turkey in 1988. Since then, we have conducted six GPS experiments simultaneously with the SLR observations. Between 1988 and 1992, 102 GPS sites were occupied over 121 days, and we acquired 2-4 epoch observations at 66 sites. This thesis investigates effective approaches to GPS data analyses and presents the results of these GPS measurements. By providing a GPS derived velocity field for Turkey, we quantify the motions within and across the plate boundaries, and discuss the tectonic implications of geodetically observed rates of crustal deformation.

We used GAMIT/GLOBK software packages to analyze our GPS measurements. In order to handle increasingly voluminous data, we devised and tested a "combination" approach to data analysis , and applied it to the 1991 and 1992 spring/fall experiments. Our results show that this combination of regional and global GPS measurements effectively provides the highest precision by better determining the orbital parameters of GPS satellites.

Our GPS derived velocity field indicates four major domains, each having distinct kinematics: the Pontus block, the Anatolian and Arabian plates, and the Caucasus domain. The most prominent feature in this velocity field is the Anatolian plate, which is separated from neighboring domains by strike-slip fault zones (East and North Anatolian faults, North Aegean trough), subduction (Hellenic and Cyprean arcs), and collision zones (Bitlis-Zagros suture, Caucasus domain). Given the consistency between GPS and SLR velocities, we have extended the Anatolian plate farther west to include the southern Aegean and Greece. In the east, the velocity field for the Van block, located east of Karhova, is in concert with that of Anatolia, so we consider it a part of the Anatolian plate as well. The Pontus block has a small amount of motion (8±795%mm/yr) relative to Eurasia.

As a first approximation to a rigid plate, we parameterize the velocity field observed for the Anatolian plate by Euler vectors, and find that the Pontus-Anatolia can be defined by a Euler vector (PON EANA), located at 33.4±0.5°E, 31.1±1.3°N (north of the Sinai peninsula), with a counterclockwise angular velocity of 1.25± 0.15°/Myr. We show that this Euler vector clearly delineates the North Anatolian fault as the boundary and suggests 25±895% mm/yr. slip. We also tie GPS observations on the Anatolian plate to the NUVEL-1 model by ARAEana[ø+43.2±0.7°E, l=31.2±1.2°N, w=1.22±0.15°/Myr]. Predicted slip on the East Anatolian fault is 18±12955mm/yr. We have inferred subduction rates and characterized the internal deformation of the Anatolian plate. Anatolia-Africa motion is 50±1095%mm/yr at the Hellenic arc, and decreases easterly to 20±1095% mm/yr., converging to the Anatolia-Arabia motion. Our GPS derived velocity field indicates site velocities in western Turkey which are higher by 19mm/yr than those predicted by these Euler vectors. This suggests that there is considerable amount of internal deformation. Residuals to rigid plate motion in southwestern Turkey and in the southern Aegean indicate the roll-back of the subducted African plate beneath the Hellenic arc induces the extrusion of the Anatolian plate as a whole over the African plate. Further, we quantify the partitioning of oblique convergence between the Arabian and Eurasian plates into strike-slip and shortening, and show that Arabian plate motion is transferred to the Anatolian plate. In the Eurasia fixed frame, the strike-slip component is 26±1495%mm/yr, while shortening in the Caucasus is about 13±995%mm/yr.

Our GPS observations suggest that the deformation in the eastern Mediterranean is controlled by the collision in the east at the Bitlis-Zagros/Caucasus zones and subduction at the Hellenic arc.

We compare our GPS derived velocity with geological observations by back-rotating the plate boundaries to the Pliocene. We show that the North and East Anatolian faults have sustained the boundaries for the Anatolian plate since the Pliocene. We find that contemporary motions can account for the total cumulative offset (25km) along the East Anatolian fault since the Pliocene. Along the North Anatolian fault, the calculated 75km displacement, however appears to exceed the lower bound (35km) for geologically observed cumulative slip since the Pliocene. This suggests that the present-day velocities may not completely projected back into the geological time scale for the periods longer than a couple of million years, due to frequent changes in boundary conditions.

We further compute the velocity gradient tensor from GPS and SLR observations and compare them with the strain rates observed from the seismic moment tensors. Over-all, we find a good agreement between patterns of principal strain rates across the North Anatolian fault and those of the northern Aegean.