Anything about geophysics and geoscience: jobs, events, technical or general discussions ...
Post Reply
Posts: 4
Joined: Sun Aug 23, 2020 4:20 pm


Post by Sandralin »

What is a phase shift of sqrt (i) = 45 degrees (like the square root of i) in seismic migration?
What is a phase shift and what does that "i" stand for?
I remember to work with this before but i do not remember it now.

There is also one problem I have to understand. I have seen many posts of CMP here.
In an acquisition, what are CMPs and the number of CMP? There is also CMP spacing?
Why is these number important when the source and receiver numbers and spacing are known?
What are these in use for?

Sorry this may be a stupid question, but I worked with this long time ago.

VIP Member
VIP Member
Posts: 625
Joined: Sat Mar 24, 2012 11:35 pm
Top poster of the Month


Post by GuyM »

Phase shift refers to Fourier analysis; you split the traces up into frequencies, and get an amplitude spectra (the strength of each frequency) and a phase specta (the relative position of the sine waves to each other, from zero, in degrees or more usually radians)

There's different references to phase shifts in migration, depending on the type of algorithm being used.

CMP is Common Mid Point; this is used interchangeably with CDP (common depth point) even though CMP is technically more correct.
The CMP is half-way between the source and receiver. Data is usually put into CMP "bins" that have a fixed width (spacing) to allow for small variations in acquisition. The natural CMP spacing is half of the receiver separation.

Data is collected so that there are many traces (ideally >50) in a CMP bin, all with different offsets (offset = spacing between source and receiver)
The traces then have a correction applied for offset (Normal Moveout Correction or NMO) which shifts the data in two-way time so that it is simulating what would happen if the source and receiver were co-located (zero offset). You can then sum the traces, as the reflection events will all be aligned. When you sum the traces the signal combines and the noise cancels - so you get a better signal:noise ratio and can see deeper, fainter reflections.

Without the NMO correction, the reflections from a flat layer would be hyperbolic.

Post Reply
  • Similar Topics
    Last post