Transcript Folie 1
Evidence for a low-permeability fluid trap in the Nový Kostel Seismic Zone from double-difference tomography Catrina Alexandrakis1,3, Marco Calò2, Fateh Bouchaala1 and Vaclav Vavryčuk1 1 Institute of Geophysics, CAS University of Strasbourg 3 Institute of Geophysics and Geoinformatics, TU BAF 2 EOST, 3rd Annual AIM Workshop I October 10 – 12, 2012 | Smolenice Castle, Slovakia Acknowledgements • Data: – J. Horálek, A. Boušková and other members of the WEBNET group • Funding: – European Union Research Project AIM ‘Advanced Industrial microseismic Monitoring‘ - Marie Curie Actions 2 Outline • Introduction • Methodology – Double-Difference Tomography – Weighted Average Mean Analysis • Results and Interpretation • Conclusions 3 West Bohemia Seismic Zone 4 Swarm Triggers Smrčiny Pluton Geissler et al., 2005 Babuška and Plomerová, 2008 5 Outline • Introduction • Methodology – Double-Difference Tomography – Weighted Average Mean Analysis • Results and Interpretation • Conclusion and Future Work 6 Double-Difference Tomography TomoDD (Zhang and Thurber, 2003) 7 Double-Difference Tomography TomoDD (Zhang and Thurber, 2003) 8 Double-Difference Tomography • Advantages: – Relocates hypocenter locations – 3D Vp and Vs model of focal zone – Gives the Derivative Weight Sum (DWS) at each node • Disadvantages: – No error estimate for the velocity models – Starting model parameterization introduces bias and artifacts 9 Weighted Average Mean (WAM) Analysis (Calò et al., 2011) • Solution to parameterization artifacts • Calculates the Weighted Standard Deviation (WSTD) for the final model Steps 1. Define basic model parameters (e.g. Velocity model, node locations, hypocenters) 2. Perturb the basic parameters 3. Average models together using tomoDD’s DWS 4. Calculate the standard deviation using DWS as a weighting factor 10 Single Inversions Weighted Average Mean Model Weighted Standard Deviation 11 Input Data • Absolute P and S arrival times -- WEBNET • Differential Times (two events, single station) – Catalog differential arrival times – Cross-correlated arrival times • Event Locations -- WEBNET – – – – 474 events Magnitude 0 - 3.8 Initial hypocenter locations range from 7 to 12 km depth HypoDD - relocated events • 3D Velocity Model – Initial Vp model and Vp/Vs (1.70) -- Malek et al., 2000 12 A‘ A A‘ A HRED All Stations A‘ VAC A A‘ HRC A A‘ A 13 Outline • Introduction • Methodology – Double-Difference Tomography – Weighted Average Mean Analysis • Results and Interpretation • Conclusions 14 Checkerboard Test 15 WAM Model 16 WAM Model 17 Average Velocities Average Model Base Model Average Model Average Model Base Model Base Model 18 Wave speeds and fluids • P-Velocity – Expect a decrease in fluid-filled and fractured materials – Overpressured conditions may produce a velocity increase (Ito et al., 1979; Popp and Kern, 1993) • Vp/Vs ratio: – Sensitive to the presence of fluids – Increases in fractured and fluid-filled materials 19 Average Velocities Average Model Base Model Average Model Average Model Base Model Base Model 20 Weise et al., 2001 Weise et al., 2001 Outline • Introduction • Methodology – Double-Difference Tomography – Weighted Average Mean Analysis • Results and Interpretation • Conclusions 23 • 3D velocity analysis reveals: – Layer of low Vp/Vs ratio values corresponds with the Smrčiny Pluton – May act as a low-permeability fluid trap – High Vp/Vs and P-velocities occur along the fault plane – Correspond with previously identified principal faults – High Vp/Vs values extend to the surface and may reflect fluid pathways 24 Future Work… North – South Principal Fault Across-Strike 25 Anomaly Restoration 27 Starting Model Tests Slow Model Base Model Fast Model 28