170 likes | 301 Views
Interferometric Prediction and Least Squares Subtraction of Surface Waves. Shuqian Dong and Ruiqing He University of Utah. Land Field Data Test. OUTLINE. Motivation: Surface Wave Filtering. Interfer. Surface Wave Theory. Conclusions. Land Field Data Test. OUTLINE.
E N D
Interferometric Prediction and Least Squares Subtraction of Surface Waves Shuqian Dong and Ruiqing He University of Utah
Land Field Data Test OUTLINE • Motivation: Surface Wave Filtering • Interfer. Surface Wave Theory • Conclusions
Land Field Data Test OUTLINE • Motivation: Surface Wave Filtering • Interfer. Surface Wave Theory • Conclusions
Motivation A CSG with Strong Surface Waves A CSG with Strong Surface Waves 0 0 Time (s) Time (s) • Solution: Interfer. Predict. + Least Squares Subtraction. Accounts for dispersion. 1.0 1.0 7200 7200 0 Offset (m) 0 Offset (m) • Problem: Surface waves = strong coherent noise blurs seismogram. Moveout-based filtering not always effective for dispersive waves.
Land Field Data Test OUTLINE • Motivation: Surface Wave Filtering • Interfer. Surface Wave Theory • Conclusions
A B B C A B C • Prediction of multiples by convolution (SRME) * • Prediction of Primaries by Crosscorrelation (Interferometry) ⊕
e e ikx ikx u (s,g) u (s,g’) u (s,g’)= A(s,g’) u (s,g)= A(s,g) sg sg’ ⊕ τ g g’ S ik(x x gg’ Sg’ u(g,g’) } * u (s,g’) u (s,g) u(g,g’) = -x sg e ) τ = A(s,g’) A(s,g) g g’ • Predict Surface Waves by Crosscorrelation ⊕
A B C B B B C C A A’ C B B A’ • Predict Surface Waves by Crosscorrelation ⊕ + ⊕
… S S 2 N S 1 g g’ g g’ • Coherent Stacking: surface waves (all src pts = stationary) Incoherent Stacking: primaries ⊕ • Coherent Stacking: FS Multiples? Avoid stationary source points
Offset (m) 3600 0 Offset (m) 3600 0 0 0 Time (s) Time (s) 2.0 2.0 Predcted Surface Waves Original Data 1 Amplitude 0 -1 Time (s) 2.0 0 Surface Waves Prediction
Refl. Surf. d d d (t) (t) (t) = + Pred. Refl. d d f - d (t) (t) * (t) ≈ (t) f (t) * = - Least Square Matching Filter
Surface Waves Filtering Results Original Data Filtered Data 0 0 Time (s) Time (s) 2.0 2.0 0 Offset (m) 7200 0 Offset (m) 7200
Result Comparison Results of f-k method Results of interferometric method 0 0 Time (s) Time (s) 2.0 2.0 Offset (m) 7200 Offset (m) 0 0 7200
Conclusions • Preliminary results promising for interfer. Prediction + subtraction surface waves. • Future work: iterative prediction + subtraction.
Can Interferometric Prediction+Subtraction work for Irregular 3D Arrays? Answer?: Irregular S. Calif. Earthquake Array Predicted Surface Waves 400 38 N Predicted Surface Waves Stations Station Offset (km) Latitude 32 N 0 -200 200 Time (s) Longitude 120 W 115 W (Andrew Curtis, The Leading Edge, 2006)
Acknowledgements We thank the UTAM sponsors for the support of the research.