Decomposition of seismic signal based on Hilbert-Huang transform

Hilbert-Huang transform (HHT) is a new time-frequency analytic method to get frequency variation with time of the non-linear and non-stationary signal. The basic implementation flow of this method includes two steps: empirical mode decomposition and Hilbert transform for instantaneous frequency. We can obtain the time-frequency spectrum of signal after these two steps. The key step of this method is empirical mode decomposition (EMD) with which any signal can be decomposed into a finite number of intrinsic mode functions (IMF). The IMF has clear physical significance. Applying Hilbert transform to those IMFs we obtain instantaneous frequency of each IMF. Displaying all the instantaneous frequencies on the same plane we get a time- frequency spectrum. This paper takes real seismic signal for example, presents a kind of empirical mode decomposition method which is fit for the decomposition of seismic signal and illustrates the superiority of Hilbert-Huang transform in analyzing seismic signal and its existing problem. Finally we will summarize the application of Hilbert-Huang transform in seismic data processing.