Changes

  | isbn    = 0891816607

  | isbn    = 0891816607

}}

}}

Virtually all seismic data processing is aimed at ''imaging'' the earth's subsurface, that is, obtaining a picture of subsurface structure from the seismic waves recorded at the earth's surface. Deconvolution, for example, aims to sharpen reflections, and common midpoint (CMP) stacking exploits data redundancy to enhance signal-to-noise ratio while producing a seismic time section that simulates what would have been recorded in a ''zero-offset'' seismic survey, that is, one in which a single receiver, located at each seismic source position, records data generated by the source at that position.

Virtually all [[seismic data]] processing is aimed at ''imaging'' the earth's subsurface, that is, obtaining a picture of subsurface structure from the seismic waves recorded at the earth's surface. Deconvolution, for example, aims to sharpen reflections, and common midpoint (CMP) stacking exploits data redundancy to enhance signal-to-noise ratio while producing a seismic time section that simulates what would have been recorded in a ''zero-[[offset]]'' seismic survey, that is, one in which a single receiver, located at each seismic source position, records data generated by the source at that position.

Of the many processes applied to seismic data, seismic migration is the one most directly associated with the notion of imaging. Until the migration step, seismic data are merely recorded traces of echoes, waves that have been reflected from anomalies in the subsurface. In its simplest form, then, ''seismic migration'' is the process that converts information as a function of recording time to features in subsurface depth. Rather than simply stretching the vertical axes of seismic sections from a time scale to a depth scale, migration aims to put features in their proper positions in space, laterally as well as vertically.

Of the many processes applied to seismic data, seismic migration is the one most directly associated with the notion of imaging. Until the migration step, seismic data are merely recorded traces of echoes, waves that have been reflected from anomalies in the subsurface. In its simplest form, then, ''seismic migration'' is the process that converts information as a function of recording time to features in subsurface depth. Rather than simply stretching the vertical axes of seismic sections from a time scale to a depth scale, migration aims to put features in their proper positions in space, laterally as well as vertically.

The task of migration here is to convert or ''map'' reflections along the diffraction into a single point at the position of the boulder. The reverse process, by which the boulder gives rise to the observed diffraction pattern, is called ''modeling''.

The task of migration here is to convert or ''map'' reflections along the diffraction into a single point at the position of the boulder. The reverse process, by which the boulder gives rise to the observed diffraction pattern, is called ''modeling''.

While the earth's subsurface is more complicated than that shown in [[:file:seismic-migration_fig1.png|Figure 1]], the seismic data that would be obtained over the real earth can for all purposes be represented as a superposition of many diffraction curves generated by each of many boulder-like anomalies in the subsurface.

While the earth's subsurface is more complicated than that shown in [[:file:seismic-migration_fig1.png|Figure 1]], the [[seismic data]] that would be obtained over the real earth can for all purposes be represented as a superposition of many diffraction curves generated by each of many boulder-like anomalies in the subsurface.

[[:file:seismic-migration_fig2.png|Figure 2]] shows another depth section and associated seismic section for a subsurface consisting of a single dipping reflector. For a constant-velocity subsurface, the many weak diffractions from very closely spaced points along the reflector (of which five are shown in the figure) give rise, through constructive and destructive interference, to a net ''reflection'' along the straight-line envelope of the diffraction curves. Note that the reflection is displaced laterally from the true reflector position (the line connecting apexes of the diffraction curves). It is this lateral mispositioning of reflections from dipping reflectors that gave rise to the term ''migration'' for the process that corrects the positioning.

[[:file:seismic-migration_fig2.png|Figure 2]] shows another depth section and associated seismic section for a subsurface consisting of a single dipping reflector. For a constant-velocity subsurface, the many weak diffractions from very closely spaced points along the reflector (of which five are shown in the figure) give rise, through constructive and destructive interference, to a net ''reflection'' along the straight-line envelope of the diffraction curves. Note that the reflection is displaced laterally from the true reflector position (the line connecting apexes of the diffraction curves). It is this [[lateral]] mispositioning of reflections from dipping reflectors that gave rise to the term ''migration'' for the process that corrects the positioning.

[[:file:seismic-migration_fig3.png|Figure 3]] shows another perspective on this mispositioning. Reflections recorded at zero source-receiver offset follow ray paths that are perpendicular to the reflector. As a result, the reflection from the point on the reflector beneath point P, for example, would be recorded by the geophone at location G, to the right.

[[:file:seismic-migration_fig3.png|Figure 3]] shows another perspective on this mispositioning. Reflections recorded at zero source-receiver offset follow ray paths that are perpendicular to the reflector. As a result, the reflection from the point on the reflector beneath point P, for example, would be recorded by the geophone at location G, to the right.

Regardless of the migration approach implemented, the key parameter of the process is ''velocity''. Since migration involves pushing waves back to their reflecting points, it is essential that the waves be pushed backward through the same medium through which they have propagated. Clearly, waves will not get back to the correct position at a given time if the velocity used in the migration process differs from the actual subsurface velocity. Unfortunately, subsurface velocity is seldom well known, particularly in geologically complex areas. Today's migration algorithms are highly accurate when supplied with the correct subsurface velocity. Because subsurface velocity can only be estimated, however, migration yields only an estimate of the true subsurface.

Regardless of the migration approach implemented, the key parameter of the process is ''velocity''. Since migration involves pushing waves back to their reflecting points, it is essential that the waves be pushed backward through the same medium through which they have propagated. Clearly, waves will not get back to the correct position at a given time if the velocity used in the migration process differs from the actual subsurface velocity. Unfortunately, subsurface velocity is seldom well known, particularly in geologically complex areas. Today's migration algorithms are highly accurate when supplied with the correct subsurface velocity. Because subsurface velocity can only be estimated, however, migration yields only an estimate of the true subsurface.

Where lateral variation of velocity is modest (as in many places in the [[Gulf of Mexico]]), migration methods in the class called ''time migration'' have performed adequately. Where lateral velocity variation is severe (as in many overthrust areas), more computationally intensive ''depth migration is'' required. Note that the terms ''depth'' and ''time migration'' do not relate to whether the migrated results are presented as a function of time or depth. Results of both migration categories are most often displayed in time (as in the examples shown here) because of added uncertainties in results converted to depth. While depth migration is capable of accurate subsurface imaging where velocity is complex, the required accurate estimation of velocity is difficult and time consuming.

Where lateral variation of velocity is modest (as in many places in the [[Gulf of Mexico]]), migration methods in the class called ''time migration'' have performed adequately. Where lateral velocity variation is severe (as in many [[overthrust]] areas), more computationally intensive ''depth migration is'' required. Note that the terms ''depth'' and ''time migration'' do not relate to whether the migrated results are presented as a function of time or depth. Results of both migration categories are most often displayed in time (as in the examples shown here) because of added uncertainties in results converted to depth. While depth migration is capable of accurate subsurface imaging where velocity is complex, the required accurate estimation of velocity is difficult and time consuming.

==Poststack versus prestack migration==

==Poststack versus prestack migration==

[[Category:Geophysical methods]]

[[Category:Geophysical methods]]