Dynamic smoothing in crosswell traveltime tomography

• DOI: 10.1190/1.1444115
• Publication Date: 2002-10-11T19:48:56Z
• Authors: Tamas Nemeth; Egon Normark; Fuhao Qin
• Keywords: Smoothing; Regularization; Operator (biology)

Variable‐size (dynamic) smoothing operator constraints are applied in crosswell traveltime tomography to reconstruct both the smooth‐ and fine‐scale details of tomogram. In mixed underdetermined problems a large number iterations may be necessary introduce slowly varying slowness features into To speed up convergence, dynamic applies adaptive regularization prediction error function with help model covariance matrix. By so doing, term has larger weight at initial dominates final small weight. addition, it is shown that acts by reweighting adjoint modeling (preconditioning) providing additional damping. Comparisons two operators, low‐pass filter multigrid technique, fixed‐size (static) operators show yields more accurate velocity distributions greater stability for contrasts. Consequently, preferred choice regularization.