Crosshole Seismic: Features of the Initial Models Design

“Crosshole seismic (CS) survey is a reliable and informative method of investigation of geological media, being a leading and sometimes unique geophysical method in geological-engineering researches.

However, there is a pitfall in processing of CS data: correctness of result model urgently depends on the correctness of input data.

Examples of CS engineering researches are considered. The main aim is to investigate dependence of the results of first breaks tomography on the initial model choice. Use of homogeneous velocity model is shown to lead often to artifact results. It is necessary to consider all a priory information, including velocities, calculated for the pairs of sources and receivers at the nearly same depths, and different wave types. Presence of contrast velocity boundaries in the lower part of investigated media and, as a rule, the first break of the head waves is shown to result in incorrect definition of velocities in the upper part. To solve this problem, the 2-stage processing is suggested. First, it is carried out only for the upper part, dealing with direct waves. Then, inversion for the whole volume is performed, using in the initial model the result of the first stage. The approach is verified by numerical modeling.”