Experiences in pre stack and post stack wavelet processing on time lapse data

Wavelet processing techniques aim at estimating, manipulating and, optionally, removing the wavelet resulting from the filtering effect caused by acquisition on the Earth reflectivity. Two approaches are tested, based on logs from a well. In a first approach, near zero offset propagation wavelets are extracted from stacked data from different vintages using the zero offset reflectivity. The peculiarity of the acquisition of each vintage is then removed by means of inverse filtering. Time lapse differences are detected comparing the output data in terms of amplitude envelopes and pseudo acoustic impedance. A pre stack approach is also tested on a single vintage data set: offset dependent wavelets are extracted from data after calculation of non zero offset reflectivities. Inverse filtering on pre stack data provides an increase in temporal resolution and in the accuracy of the determination of stacking velocities.