Progress Report on Crosscorrelogram Migration of Free-Surface Multiples in RVSP Data

According to Schuster and Rickett (2000), a shot gather of crosscorrelograms is partly equivalent to a shot gather of seismic traces, and the reflectivity distribution can be imaged by migrating free-surface multiples in the crosscorrelograms without knowing the source positions. Here, synthetic data and Exxon’s Friendswood data are used to test the feasibility of migrating free-surface multiples in crosscorrelograms obtained from RVSP data. Results show that crosscorrelogram migration is able to image some reflectors without knowing the source’s time history and location. However, spurious events are seen on the migrated crosscorrelogram and so further work is needed to improve the quality of the imaged reflectivity distribution. INTRODUCTION Schuster and Rickett (2000) presented daylight imaging in a V (x,y, z) medium 1 and showed that crosscorrelating a trace at location A with a trace located at B will yield a trace with kinematics that are partly equivalent to the kinematics of a trace generated by a shot at A and recorded by a receiver at B. They also showed that the reflectivity distribution can be imaged by migrating the crosscorrelograms without knowing the source’s time history and source location. Their scheme of migrating crosscorrelograms uses the free-surface ghost reflections, not the primary reflections. After bandpass filtering and some other preliminary processing steps in the common receiver gather (CRG) domain, the downgoing waves (including direct waves and ghost reflections) are seperated from the upgoing waves (primary reflections) using a median filter. Then, in the common shot gather (CSG) domain, any trace at g is temporally crosscorrelated with every other trace at g′ to generate a pseudo − shot gather denoted by φgg′. The pseudo − shot gather φgg′ is migrated using the free-surface reflection imaging condition for a source at g and a receiver at g′. In this report I validate the crosscorrelogram migration method by processing both synthetic data and Exxon’s Friendswood RVSP data. The results for both the synthetic example and Exxon’s Friendswood RVSP data show that crosscorrelogram migration can be used to partly image the reflectivity distribution without knowing the source signal and source positions. However, further work is needed to improve the imaging so the result has the same quality as that of seismic trace migration.