Abstract: This paper takes the natural earthquakes recorded in the Shanghai area, the "4·20" Shanghai Central Blasting and the "4·26" Pudong Blasting as examples to analyze the standard time-frequency transform, and study the obtained standard time spectrum characteristics. By comparing the time-frequency spectrum, it can be seen that the resolution of the time-frequency domain of the NTFT-transformed data is significantly improved compared to the wavelet transform, and the signal characteristics of the seismic signal and background noise can be displayed more clearly. To a certain extent, it can effectively identify unobvious seismic waveforms and even seismic signals that are submerged by noise. The frequencies of natural earthquakes and blasting events in Shanghai are mainly distributed in the range of 20 ~ 40 Hz, with little difference in frequency range; natural seismic energy signals are mostly concentrated around 20 Hz, while blasting event energy signals are roughly concentrated between 30 ~ 40 Hz. Natural seismic energy distribution presents "multi-peak" characteristics, while blasting events show relatively "few peaks", mainly showing the characteristics of high-energy, small-scale impulse forms.