PositionIdentification - softwareunderground/xsdf GitHub Wiki
Handling of position identification
A major issue when handling old SEG-Y files is what defines the position of a trace. In 3D, you have inline/crossline versus (X,Y), the two are connected via what is called in Petrel a 'SpatialLattice'. In 2D, the situation is, in a way, even worse. There is no such consensus on trace identification like the inline/crossline in 3D. Therefore, to identify a trace you have to choose a 'Trace number'. The CDP number would be a good candidate, were it not that very often users do not even recognize the numbering.
In any case, trace positions need to be identified and coupled to a position in a coordinate system. Although the exact nature of the coordinate system is crucial to the user, it is not an issue of seismic data exchange as such. The only requirements are:
- A trace has a key which identifies its position. The key can consist of multiple numbers (for example inline and crossline).
- That key will be coupled to an (X,Y) in a known coordinate system.
For pre-stack data, an extra field 'offset' enters the arena. It takes upon itself a bit of both roles: identification AND co-ordinate. To complete the positioning info the 'azimuth' would be required, but it's hardly ever available.
The general strategy to handle these issues is to make the keying mandatory and some form of positioning from that required. Any other per-trace info can be optional.