Abstract

The presence of high amplitude noise corrupts the dataset in each subsequent stage of data processing, which may lead the less reliable end product of processed seismic data. To overcome this problem, normal de-selection methods were used by taking reference to the average absolute amplitudes over different windows of analysis thus removing the high amplitude noise bursts and noisy/dead traces. This resulted in preservation of relative amplitudes which are otherwise generally disturbed by auto editing methods and also in increased signal to noise ratio thus improving the reliability of the entire data set. This present method is found to be superior as it saves time for the processor from manual editing and simultaneously avoids using statistically controlled auto editing method that fails to preserve the relative amplitude.