|
Models of Faults and Fluid Flow |
1 Department of Geological and Environmental Sciences, Stanford University, 450 Serra Mall, Bldg. 320, Stanford, CA, 94305, USA
2 NASA Goddard Space Flight Center, Geodynamics Branch, Mail Code 921, Greenbelt, MD, 20771, USA jmuller{at}core2.gsfc.nasa.gov
3 Department of Earth Sciences, Oxford University, Parks Road, Oxford OX1 3PR, UK
Using an elastic dislocation model, we incorporate a historical earthquake catalog, mapped Marmara Sea fault traces, and fault slip distributions for the 1999 Izmit earthquake inferred from InSAR and GPS data to determine various stress change scenarios crucial for evaluating future earthquake potential in the eastern Marmara Sea. We have tested six plausible past rupture configurations arising from the uncertainty in the location of the western termination of 1999 Izmit earthquake rupture and the location of the 1963 Yalova earthquake rupture. Coulomb stresses calculated are increased on the Princes’ Islands, Çinarcik, and Armutlu fault segments in each case. In four of the six plausible configurations of previous ruptures, the Çinarcik fault receives the greatest average stress change. In one other configuration, the average stress increase on the Princes’ Islands fault is greatest. In another, the stress changes on the Çinarcik and Princes’ Islands faults are comparable. Moreover, we show that rupture initiating on either the Princes’ Islands or Armutlu faults would be favoured to propagate onto the Central Marmara, or Imrali fault, respectively, based on the favourable geometries of the respective fault intersections. Rupture initiating on the Çinarcik fault, however would be limited to a much shorter length based on its mapped western termination. Therefore, while the earthquake-induced stress changes may, in most cases, be greatest on the Çinarcik fault, an earthquake initiating on this fault segment may produce a shorter cumulative rupture compared to rupture initiated on the two other major eastern Marmara Sea fault segments. These results are encouraging for the use of geomechanical modelling tools in addressing uncertainties inherent in most geological and geophysical data applied to earthquake-related problems.
