Depth imaging with crooked 2D and irregular 3D seismic data in rugged terrain

Due to permit restrictions, crooked 2D and irregular 3D seismic reflection data were acquired within a 330 km area in central coastal California to image geologic structure and faults from the near-surface to depth of > 10 km. Key steps to achieve successful depth imaging included, nonstationary coherent noise modeling and removal, 3D processing of all 2D and 3D data, 3D tomography, residual statics using first-arrivals, and adaptive fast beam migration. No sonic log velocity data were available so multiple stages of 3D velocity inversion including joint velocity-density inversion with gravity data were used to develop velocity models and statics for prestack-depthmigration velocity analyses. Improved signal-to-noise in the 3D volumes provided constraints to design wavenumber spectral-balancing and wavelet transform denoising parameters to improve imaging in the 2D data.