Earthquakes are amongst the most significant natural hazards faced by mankind: globally, millions of fatalities are forecast this century, while in the United States half of the population is at risk from damaging shaking. However, earthquakes also provide valuable information on Earth?s structure and tectonics. In this project, we will be using information from earthquakes to learn about the structure and tectonics of the Zagros mountains in Iran. The Zagros are a classic example of a young continental ?collision zone?, caused by the Arabian plate crashing into Eurasia, and are also amongst the most seismically-active regions anywhere in the world. However, its many earthquakes are buried under a thick blanket of sediments and so the fault-lines that generate them cannot be identified using conventional field mapping. Instead, we will use cutting-edge seismological techniques to locate very precisely the past half century of earthquake activity, in doing so highlighting the locations of buried faulting at depth. The Zagros can be used as a modern analog to parts of the Rocky Mountains and Appalachians which formed tens to hundreds of millions of years ago and which are no longer seismically active. An improved understanding of the Zagros will therefore help us understand the structure and tectonic history of parts of the United States.
The Zagros range in Iran is a type example of how thick sedimentary basins are shortened during the early stages of continental collision. However, its present-day kinematics are obscured by uncertainties in the locations of steep, seismogenic reverse faults buried within or beneath the 8-16 km-thick sedimentary cover, and by conflicting interpretations of their role in controlling the growth of "whaleback" anticlines which dominate its physiography. The prevalent seismotectonic model for the Zagros has earthquakes concentrated along a small number of large, basement reverse faults which account for discrete steps in the exposed stratigraphic level, with shortening of cover rocks away from the master thrusts accommodated by detachment folding along weak salt layers. An alternative model holds that earthquake faulting is widely-distributed through the Zagros, correlating with individual anticlines through fault bend and fault propagation folding. We propose a simple test of these conflicting models that will integrate two state-of-the-art earthquake relocation procedures to yield improved epicenters for the entire, ~50 year back-catalog of several thousand teleseismically-recorded events in the Zagros. The first step repositions discrete clusters of earthquakes with respect to well-located ?calibration? events whose hypocenters are known independently from InSAR observations and modeling or from temporary local seismometer deployments. The second step uses results from the first as prior constraints in a Bayesian relocation algorithm that is better suited to larger data sets spread over wider regions, and which has well-characterized uncertainties and minimal location bias. Epicentral uncertainties will drop from a few tens of kilometers (for routine catalog estimates) to no more than a few kilometers. The existence and importance of major basement faults will be confirmed if the relocated epicenters aggregate into streaks of seismicity aligned with the known steps in stratigraphic level, whereas the opposing model will be supported if earthquakes mostly lie away from the proposed basement faults.
