Full waveform inversion based on dynamic data matching of convolutional wavefields

Cycle skipping problem caused by the absent of low frequencies and inaccurate initial model makes full waveform inversion (FWI) deviate from true model. A novel method is proposed to mitigate cycle phenomenon dynamic data matching which improves synthetic observed events regulate updating in a correct direction. 1-dimentional (1-D) Gaussian convolutional kernels with different lengths are used extract features each time sample trace represents integrated properties wavefield at ranges centered on sample. According minimum Euclidean distance features, optimally matched pairs samples can be found. constraint evaluates reliability attenuating amplitude according values traveltime differences between improve accuracy matching. In addition, have capability contaminated strong noises accurately robustness propose further. The selection scheme optimal parameters discussed concluded ensure convergence method. Numerical tests Marmousi verify feasibility provides new approach tackle FWI when using field seismic data.