Zhang, P., Dai, Y., Wang, R. and Tan, Y., 2017. A quantitative evaluation method based on EMD for determining the accuracy of time-varying seismic wavelet extraction. Journal of Seismic Exploration, 26: 267-292. The bandwidth and amplitude of wavelet deconvolution results are the most important indicators of accuracy for time-varying wavelets. To evaluate the accuracies of extracted seismic wavelets based on these indicators, we propose a quantitative evaluation method based on empirical mode decomposition (EMD), which offers the advantages of adaptive decomposition and multi-scale analysis and can highlight local characteristics. First, time-varying seismic wavelets are extracted from a non-stationary seismogram and subjected to deconvolution or reflectivity inversion. Then, to appraise these wavelets, the amplitude spectrum from the deconvolution or inversion results is decomposed into multi-layer intrinsic mode functions (IMF) using EMD. Next, an evaluation parameter is constructed by summing the number of local extremes in all IMFs and normalizing this sum with respect to the number of frequency points in the amplitude spectrum. Larger values of this parameter indicate more accurate extracted wavelets. When applied to both synthetic and field-collected seismic data, the proposed method performs better than conventional methods for evaluating the accuracy of time-varying wavelet extraction.