Kinematic aspects of residual prestack migration

TL;DR

This paper introduces a kinematic framework for residual prestack migration, enabling improved parameter estimation and bias correction in seismic imaging with low computational costs.

Contribution

It presents a novel approach to residual migration using aplanatic curves and raytracing, applicable to isotropic and anisotropic data, with methods to estimate migration parameters and assess the need for new migrations.

Findings

Residual migration can correct positional bias.

Aperture width over the Fresnel zone indicates the need for new migration.

Low-cost velocities derived from residual moveout analyses.

Abstract

In this contribution it is shown that various aspects of the concept of residual migration can be utilized for the case that a prestack time or depth migration has been performed for a seismic survey and a new depth is available: The concept of residual migration is introduced by determining travel times of reflected events for individual traces from aplanatic curves. These events are migrated in suitable configurations for the new velocity model for which a raytracing set as required for a Kirchhoff prestack depth migration is available. The resulting transformations can be utilized in various ways to estimate important parameters such as the positional bias of migrated events with respect to the corresponding zero offset position and the residual moveout for the new depth model ; finally a new approach is suggested to estimate the aperture width over the Fresnel zone for the first prestack migration. Various mapped prestack migrations can be estimated, including a prestack time migration with eliminated positional bias. The necessity of a new prestack depth migration can be assessed from the aperture width over the first order Fresnel zone and from the positional bias at a particular offset. Additional kinematic parameters can be acquired for a residual prestack migration, in particular the curves over which the residual summation has to be performed. The approach is applicable to isotropic and anisotropic PSTM and PSDM. Computational costs of the implementation are low when compared to a full PSDM. In the appendix a formula is derived for the migration velocities for a prestack time migration for small to intermediate source receiver offsets. The velocities can be determined either from an inhomogeneous anisotropic depth model or from residual moveout analyses of a previous migration.