The Tidal Signal Processing Based On Wavelet Analysis

Earth tidal and earthquake is a non-stationary process. We always want to know the frequency components and the time domain features at a given moment in data processing and analysis. The Wavelet Transform is different from the traditional Fourier Transform. It has the character of multi-resolution and it is a time-frequency analysis method that the time window and frequency window can be changed. In other words, it has higher frequency resolution and lower time resolution in the low frequency parts and higher time resolution and lower frequency resolution in the high frequency parts. So the Wavelet Transform is usually called"mathematical microscope". Therefore, we choose the Wavelet Transform to deal with the tidal signals.This paper mainly study the tidal data processing applications based on wavelet transform on HUST tide observation data. We conclude the following results by analyzing many observations.(1) The study shows that the wavelet of db6 is the best to distill and separate tide signals from original observation signals, and the wavelet transform is a good method for discriminating and separating the solid tides and high-frequency signals from original observation data with wavelet of db6.(2) The main signals of LP channel are solid tides by Nakai testing and Venedikov harmonic analysis.(3) The earthquake can be distilled and the high-frequency noise can be removed by the discrete wavelet decomposition, namely increasing the accuracy of observation.It is a new useful method for Earth information of broadband domain observation and it has the important significance for researching earth observation information and earthquake prediction.