Solution to the 3D Complex Near Surface Problem by Estimation of Propagation Operators

A two-step method is proposed in this paper to solve the 3D complex near-surface problem. Based on estimating velocity-independent 3D near-surface redatuming operators, we first propose a revised 3D travel time operator that treats the average static time shift and the static time shift jitters separately. In order to make it possible to optimize these travel time operators on realistic-size 3D field data, we propose a new genetic algorithm, the Self-Adjustable Input Genetic Algorithm (SAIGA). After travel time operators are optimized by SAIGA, we propose to run another genetic algorithm optimization to fix the static time shift jitters at every source and receiver point in the field data. This two-step method has been successfully applied to a 3D land field data containing 2.4 million traces, and the results obtained are very promising.