The 1999 Izmit, Turkey earthquake sequence (M=7.5, August, Gulcuk and M=7.1, November, Duzce earthquakes) occurred within a pre-existing GPS network making it one of the most geodetically well observed seismic events on a major continental strike slip fault. Well-defined pre-earthquake strain accumulation, coseismic motions (reaching ~2 m), short-term (days - months) postseismic afterslip, and longer-term (years - 10 years) postseismic relaxation are all well imaged by geodetic observations. In addition, the earthquake appears to have altered the character of interseismic motions in unexpected ways (i.e., not easily ascribed to standard postseismic processes) as far as 200 - 300 km from the coseismic fault. As such, this event is providing fundamental constraints on the earthquake deformation cycle (inter-, pre-, co-, and postseismic) with direct implications for forecasting future earthquakes on this fault system, and by analogy for understanding the dynamics of similar strike slip faults, including the San Andreas Fault in California.
The 1999 sequence occurred along the eastern segment of a seismic gap that extends south of Istanbul across the Sea of Marmara, raising serious concerns about a future earthquake on the western Marmara segment that could have devastating effects on this vulnerable mega-city (> 20 million people). Uncertainty persists about the possible magnitude of future earthquakes in the Marmara Sea, as well as how strain is transferred from the Izmit segment to the Main Marmara Fault (MMF) within the Sea (i.e., the geometry, degree of locking, slip rate on the connecting fault). The Princes Islands are located within the Marmara south of Istanbul and approach within less than 5 km of the Izmit-Marmara connecting fault segment. Two islands (Sivriada and Büyükada) located very near the connecting segment have been selected for ICDP borehole sites (GFZ, Germany are leading this ICDP initiative). In cooperation with this work, we are focusing our initial efforts on installing GPS stations at both borehole sites. Data from these stations will be downloaded by our Turkish partners at the Marmara Research Center and by UNAVCO. Over three years, these GPS observations will provide improved constraints on the dip, locking depth, and present-day rate of strain accumulation on this fault segment. These anticipated results together with analyses of postseismic deformation throughout the Marmara region, promise to provide improved understanding of the earthquake deformation cycle, and accordingly, estimates of the timing and magnitude of future earthquakes in the Sea of Marmara.
This project is funded by the Geophysics and Tectonics Programs of NSF.
This 1-year project supported the installation of 2 GPS stations on the very small, uninhabited islands of Yussiada and Sivriada in the Sea of Marmara very near Istanbul, Turkey. These stations are monitoring the accumulation of crustal strain between the western end of the segment of the North Anatolian Fault (NAF) that caused massive damage and loss of life in the 1999, Izmit, Turkey earthquake (M7.4) and the segment of the fault immediately to the west and adjacent to the mega-city of Istanbul. Based on the >1500 year history of earthquakes in this area, the segment of the fault in the Sea of Marmara is a seismic gap capable of producing earthquakes of M~7.6. The two stations, provided by the University Navstar Consortium (UNAVCO) were successfully installed in September/October 2011 and are being downloaded daily by our collaborating partners from the TUBITAK Marmara Research Center. These two stations, located < 5 km from the NAF beneath the Sea, are designed to detect both the steady accumulation of strain that gives rise to earthquakes, and the episodic release of strain either during small earthquakes, or "aseismically" (slow fault slip that doesn’t produce earthquakes). These processes are important to clarify how earthquakes initiate (nucleation process) as well as how they transfer strain to adjacent fault segments. The quantitative information on crustal strain from these and other stations are being used to constrain probabilistic models for earthquake hazards in this critical and vulnerable region. Because the western NAF and the southern San Andreas Fault have many common characteristics, we further expect that the case study we are developing in Turkey will provide insights into earthquake processes in Southern California.
