Abstarct: Time Frequency seismic attributes are widely used in the interpretation of seismic data for various exploratory purposes including the identification of subsurface structures as well as areas containing hydrocarbons. However, the accuracy of these attributes is a function of the time frequency analysis method applied to the data. In seismic signals, various time frequency transforms have been used to obtain the appropriate time and frequency analysis. These include short-time Fourier transform, S transform, and wavelet transform, each of which has its advantages and disadvantages. The S transform has been accepted by many researchers in frequency time analysis over the years. However, due to Heisenberg uncertainty principle, this transform is not able to provide acceptable time frequency analysis. So, many methods have been proposed to increase the resolution of this transform One of these methods is the synchrosqueezing method. This method, as a post-processing method, does not have the problem of uncertainty principle and is able to transfer the energy dissipated in the frequency time plane, which has caused the reduction of temporal and frequency resolution, to the main temporal and frequency position. This technique is performed on different time frequency analysis methods including S transform. In this thesis, the aim is to investigate the performance of synchrosqueezing S transform in time frequency analysis of seismic data in order to extract different time-frequency attributes. The attributes obtained from this method are used in the field of gas detection and low frequency shadows on a seismic data set. The time frequency attributes used in this method are: instantaneous amplitude, instantaneous peak frequency, sweetness factor, C80 frequency, instantaneous average frequency, instantaneous frequency bandwidth, average quality factor and iso-frequency sections. The results show that due to the high time frequency analysis by the synchrosqueezing S transform method, the resulting seismic attributes also have a higher ability to detect low frequency shadows than the standard S transform attributes.