HighResolutionimaging高清晰度成像技术.ppt

test High Resolution Imaging Outline High Resolution Imaging: Motivation Convolution / Cross-correlation / Deconv Quadratic and Non-quadratic Regularization Marmousi model Synthetic Model Pre-stack data Where do we go from here??? Parametric AVP gathers  Acknowledgments Stacked Image for inline #71 Migration RLSM (4) RLSM (11) PLSM (4) Migration RLSM (11) PLSM (4) Depth slice (z=1800 m) Average amplitude (1000-1250m) Strata correlation (inline #76) Synthetic trace PLSM AVA gather (crossline 24,inline76) CIG (synthetic vs. inverted) AVA (synthetic vs. inverted) A: synthetic, B: inverted Red: synthetic Blue: inverted Synthetic CIG is computed using Aki Richards’ ap- proximation 4 adjacent CMPs RLSM Reconstruction PLSM Reconstruction RLSM / PLSM and Data Reconstruction We can also include a non-quadratic term to increase vertical resolution: Migration as an Inverse Problem: Non-quadratic regularization Fit the data, after all you paid for it.. Constrain AVP gathers to be flat/coherent Impose vertical sparseness - increase resolution True Z(m) X(m) Acoustic, Linearized, Constant V, Variable Density m(x,z) Data Shots d(s,g,t) CIG # 10 CIG # 11 CIG # 12 Z(m) 0m Offset 380m Adj RLSM HRM CIGs CIG # 10 CIG # 11 CIG # 12 Z(m) 0m Offset 380m Adj LS HR CIGs More Aperture More Band-Width Data HR Prediction LS Prediction T(s) Shots d(s,g,t) Feng J. and Sacchi M., 2004, CSEG Annual Mtg. Motivation Historical Evolution of Ideas Imaging with Quadratic Regularization Regularized Least Squares Migration (RLSM) Preconditioned Least Squares Migration (PLSM) Synthetics Case study (WCSB) Imaging with Non-quadratic Regularization Where do we go from here? Summary Acknowledgments To go beyond the resolution provided by the data (aperture and band-width) by incorporating quadratic and non-quadrat