Quick Links


SMELT - An Efficient, Flexible Method for Eliminating MultiplesNormal access

Author: A. J. Hardwick
Event name: 73rd EAGE Conference and Exhibition incorporating SPE EUROPEC 2011
Session: Seismic Noise & Multiple Suppression
Publication date: 23 May 2011
DOI: 10.3997/2214-4609.20149532
Organisations: SPE, EAGE
Language: English
Info: Extended abstract, PDF ( 1.83Mb )
Price: € 20

Two different case histories are presented, demonstrating the effectiveness and flexibility of a novel de-aliased demultiple technique termed Stepwise Multiple Elimination using Linear Transforms (SMELT). By definition, SMELT is applied in a stepwise manner passing data through a series of linear transform pairs, related to a discrete velocity interval to build multiple models which are adaptively subtracted from the input at each step. The process is effective in attenuating a wide range of free surface multiples and significantly internal multiples generated by shallow unconformities on the near to mid offsets of a CMP gather. In the first case history SRME fails to effectively remove free surface multiple on the mid to outer cables of 3D sail lines due to a lack of near offset information in shallow water. SMELT in conjunction with SRME overcomes this common problem. The second case history is in a slightly deeper marine setting where uplift and glacial erosion results in a significant impedance contrast at the Base Quaternary Unconformity. SRME was not applied to the data as the multiple models obtained contain major amplitude discrepancies. SMELT is used as an alternative to remove residual multiple on the near to mid offsets.

Back to the article list