1Dublin Institute for Advanced Studies, Geophysics Section, 5 Merrion Square, Dublin 2, Ireland
2Department of Geophysics, Kandilli Observatory and Earthquake Research Institute, Boğaziçi University, Istanbul, Turkey
3British Geological Survey, Edinburgh EH9 3LA, Scotland, UK
4School of Geosciences, University of Edinburgh, Edinburgh EH9 3FW, Scotland, UK
5TUBITAK Marmara Research Center, Earth and Marine Sciences Institute (EMSI), Gebze, Turkey
Abstract. The 2006 Mb = 5.3 Manyas-Kus Golu (Manyas) earthquake has been retrospectively "stress-forecasted" using variations in time-delays of seismic shear wave splitting to evaluate the time and magnitude at which stress-modified microcracking reaches fracture criticality within the stressed volume where strain is released. We processed micro earthquakes recorded by 29 TURDEP (Multi-Disciplinary Earthquake Research in High Risk Regions of Turkey) and 33 KOERI (Kandilli Observatory and Earthquake Research Institute) stations in the Marmara region by using the aspect-ratio cross-correlation and systematic analysis of crustal anisotropy methods. The aim of the analysis is to determine changes in delay-times, hence changes in stress, before and after the 2006 Manyas earthquake. We observed that clear decreases in delay times before the impending event, especially at the station GEMT are consistent with the anisotropic poro-elasticity (APE) model of fluid-rock deformation, but we could not observe similar changes at other stations surrounding the main event. The logarithms of the duration of the stress-accumulation are proportional (self-similar) to the magnitude of the impending event. Although time and magnitude of th 2005 Manyas earthquake could have been stress-forecasted, as has been recognized elsewhere, shear-wave splitting does not appear to provide direct information about the location of impending earthquakes.