Attenuation of Acquisition Footprint on Sparse Land and Ocean Bottom Data Using Least Squares Pre-Stack Time Migration
Richard Leggott*, Richard Wombell, and R. Gareth Williams
Seismic migration algorithms usually assume that the input data are recorded on a regular, fully populated grid. If this assumption is not met, the migration process will introduce noise in the data; this is often referred to as an acquisition footprint. This noise is usually offset- and time-dependent and hence can be partially hidden by the stacking process of high-fold data. However, this is not sufficient for low-fold and shallow data, nor for pre-stack analysis such as amplitude-versus-offset (AVO) or amplitude-versus azimuth (AVAZ). A common approach to this problem is to interpolate data on to a well-sampled, regular grid before migration. This can have problems if the signal-to-noise ratio of pre-stack data is low, as often occurs for onshore recording. An alternative approach is to perform an iterative, least-squares migration. In this approach, a forward migration is performed to obtain an estimate of the Earth’s reflectivity followed by a reverse diffraction. This diffracted data is then compared in a least-squares sense to the original field data to determine if the estimate of the reflectivity is valid. Iterating this process leads to an improved migration result with a substantially reduced acquisition footprint. This process is computationally intensive but is practical for Kirchhoff pre-stack time migration (PSTM). Decimation trials have been run on an onshore dataset that was originally shot with high fold. Reducing the original data by dropping every second shot line and every second shot point within a line causes noise to be introduced on the stack after conventional PSTM. Least squares migration reduced this noise significantly.
AAPG Search and Discovery Article #90077©2008 GEO 2008 Middle East Conference and Exhibition, Manama, Bahrain