Target-oriented velocity analysis using Marchenko-redatumed data

Constantin Mildner

doi:10.1190/GEO2016-0280.1

2017-03-01

PRE

Constantin Mildner *

Filippo Broggini

Johan O. A. Robertsson

Dirk‐Jan van Manen

DOI：10.1190/GEO2016-0280.1

原文链接

原文求助

摘要

En 中文

Imaging a target zone and analyzing its corresponding velocities using seismic surface data can be challenging, in particular if the target zone is located below a highly scattering overburden. Marchenko redatuming enables the redatuming of the surface reflection response to an arbitrary new datum level below such an overburden. The obtained reflection response is free of internal multiples originating in the overburden, and thus, such data can be the starting point for target-oriented analyses. We develop the first application of Marchenko-redatumed data for target-oriented velocity analysis. Marchenko-redatumed data are used as the input for a conventional migration method to create angle-domain common-image gathers. Using incorrect velocities for redatuming or migration leads to residual moveout of reflections in such angle gathers. Thus, semblance analysis of the angle gathers allows us to correct an erroneous velocity profile. By combining events occurring at different depths in the angle gathers, we can correctly determine and separate the error inherent in the overburden of the velocity model used for redatuming and in the target zone used for migration. We tested the approach of velocity analysis with Marchenko-redatumed data in the simple case of a horizontally layered model and obtained the correct velocity model updates using interval velocity corrections. We then extended our approach and successfully applied a ray-based tomographic inversion to angle gathers created with Marchenko-redatumed data for a medium having lateral heterogeneity (dipping layers) and proved the advantages of velocity analysis using Marchenko-redatumed data compared with surface data.

Keyword:

GREENS-FUNCTION RETRIEVAL

COMMON-IMAGE GATHERS

MULTIDIMENSIONAL DECONVOLUTION

REFLECTION DATA

CROSS-CORRELATION

VIRTUAL SOURCE