Improvements to Kirchhoff operator

[mrava87]

Motivation

The pylops.waveeqprocessing.Kirchhoff operator has gone through a major restyling in #431 mostly aimed at including dynamic capabilities. From the discussions in the #431 and #418 we have identified the following areas that may need further improvement. Moreover some of the new features in such PR are so far only available for the 2D implementation.

Features to be implemented

• Add 2.5D wavelet filtering in _wavelet_reshaping. Current roadblock is represented by the unclear definition of σ_S and σ_R in equation 2.3.7 of Safron2018 (neither in the Bleistein and Gray, 2002 paper).
• Devise a new definition of aperture not simply based on the offset/depth ratio to be used for heterogeneous velocity models. Currently we raise a warning when a user decides to activate such aperture limitation
• Finalize discussion about the role of snell, and if this is actually needed at all.
• Finalize discussion about angle scaling and introduce cospower as discussed in the PR
• Implement angle scaling with normal (instead of vertical) when anglerefl is provided
• Implement all features when dynamic=True in 3D

[mrava87]

For more details on the above points see #431 (comment)

[dikwickley]

hello @mrava87 @cako
i would like to work on this issue. can you please assign it to me?

[mrava87]

Sure! get in touch with us (@cako) on slack so we can discuss better maybe where it is better to start :)