Migration Strategies of Using Surface-related Multiples in Case of Incomplete Seismic Data

Seismic surveys are often restricted due to physical constraints resulting in missing traces, affecting the migrated image severely. To add to the plight, in traditional processing multiples are removed as noise, further reducing
the information contained in the signal. However, it has been established that multiples carry extra information that can become crucial, especially in cases of finite illumination. Therefore, it has become necessary to find the best strategy of using multiples for imaging. In this paper we discuss using surface-related multiples particularly in the case of incomplete seismic data. Current imaging methods incorporating surface-related multiples via reinjection of the measured wavefields bank heavily on dense receiver sampling, hence, falling short in case of limited receiver sampling. To tackle this issue, we introduce a migration method that incorporates a non-linear inversion approach, modelling all surface multiples from the original source field. This method gives better
imaging result and mitigates the effect of incomplete data substantially. Although both linear and on-linear approach have their pros and cons, via an example we demonstrate that a hybrid approach of both methodologies provides the best results.