Abstarct: Nowadays, time series analysis is widely used in the processing and interpretation of seismic data. As the ground behaves like a low-pass filter while seismic waves are propagating, it causes frequency bandwidth changes with seismic waves with time. In most processing and interpretation applications such as deconvolution, seismic noise suppression, direct detection of hydrocarbon areas, Q factor estimation and finding different seismic attributes, simultanoues representation of time and frequency information seems necessary. Conventional methods of displaying signals in the time domain and in the Fourier domain, despite their widespread use, can not simultaneously display time and frequency information centrally. By introducing time-frequency transmissions and signaling in areas where simultaneously focused and time-sensitive information is available, signal processing has entered a new stage that greatly increased its efficiency. Transforms such as short Fourier transform, Wigner-Weill distribution, wavelet transform, and S transform have been widely used in various fields of science, especially seismology, which deal with signals and their processing, in recent decades. Each of these time-frequency transforms because of different reasons such as Heisenberg uncertainty principle and etc. has some limitations and drawbacks. On the other hand, the positive points of these transforms should not be neglected. Thus, finding ways in order to weaken the weakness points of current time-frequency transforms while keeping the strength points, was our motivation in this thesis as it can also be useful in developing their usages as well. In this thesis, conventional time-frequency transforms were investigated and analyzed and defects were also presented. In the following, empirical wavelet transform as a new tool of time-frequency transform was proposed for interpretation of seismic signals. Using this, one of the interpretive applications of reflective seismic data, such as low frequency shadow detection was analyzed. The results obtained from the application of this wavelet transform show that it has a positive performance on the interpretation of seismic sections compared with conventional time-frequency transforms. However, the inability of this method to decompose the signal in displaying the time-frequency of the synthetic signal presented in this thesis is inevitable.