application of singlefrequency timespace filtering technique for seismic ground roll and random noise attenuation

Time-frequency filtering is an acceptable technique for attenuating noise in 2d (timespace) and 3-d (timespacespace) reflection seismic data. the common approach for this purpose is transforming each seismic signal from 1d time domain to a 2d timefrequency domain and then denoising the signal by a designed filter and finally transforming back the filtered signal to original time domain. the technique is efficient for ground roll and also random noise attenuation. however, if we deal with a large data set and a great number of contaminated signals with ground roll noise, a much move consuming time will be required. in this paper, timefrequency filtering is formulated and carried out by a different approach. the data is transformed from original timespace domain into several singlefrequency timespace domains, and the filters to reduce noise is designed in the new domains. the transform is easily and completely invertible. the employed time frequency analysis method is a highresolution version of stransform. application to synthetic and real shot gathers confirms the good performance and efficiency of the method for attenuating ground roll noise and random noise.

Application of SingleFrequency TimeSpace Filtering Technique for Seismic Ground Roll and Random Noise Attenuation

Timefrequency filtering is an acceptable technique for attenuating noise in 2D (timespace) and 3D (timespacespace) reflection seismic data. The common approach for this purpose is transforming each seismic signal from 1D time domain to a 2D timefrequency domain and then denoising the signal by a designed filter and finally transforming back the filtered signal to original time domain. The technique is efficient for ground roll and also random noise attenuation. However, if we deal with a large data set and a great number of contaminated signals with ground roll noise, a much move consuming time will be required. In this paper, timefrequency filtering is formulated and carried out by a different approach. The data is transformed from original timespace domain into several singlefrequency timespace domains, and the filters to reduce noise is designed in the new domains. The transform is easily and completely invertible. The employed time frequency analysis method is a highresolution version of Stransform. Application to synthetic and real shot gathers confirms the good performance and efficiency of the method for attenuating ground roll noise and random noise.