Fibre trace registration by cross-correlations - can we successfully predict helically wound fibre pitch angle from recorded data?

Kevin W. Hall, Donald C. Lawton, Kristopher A. Innanen

The Containment and Monitoring Institutes (CaMI) Field Research Station (FRS) contains an approximately 5 km long optical fibre loop comprised of straight fibre cables and helically wound fibre cables that are spliced together end-to-end. The fibre loop traverses two wells, observation well 1 (OBS1) and observation well 2 (OBS2), and a 1 km horizontal trench along the surface with straight and helically wound fibre cables. We have interpolated fibre trace co-ordinates (x,y,z) in the past for a walk-away/walk-around VSP (Snowflake) using GPS and downhole gyroscope surveys, but it was never entirely clear if we were using the correct trace spacing for helical fibre data, or which coordinates should be assigned to what DAS trace. Quality-control checks performed by interleaving straight and helical fibre traces using interpolated coordinates plus arbitrary channel assignment shifts showed that the calculated helical fibre trace spacing was close, but not correct.

In this report, we use a cross-correlation and linear regression method on co-located seismic data to estimate a helical pseudo-pitch angle. The slope and intercept from the linear regression can be used to register datasets by fractional channel number. If one of the co-located datasets has a known trace spacing, we can also predict an unknown trace spacing for the other dataset, for example, helical fibre cable trace spacing from a known geophone, accelerometer, or straight fibre trace spacing. In future, we plan to 1) interpolate co-ordinates using trace spacings calculated from estimated pseudo-pitch angles, and 2) locate channels in one dataset that are functionally at the same location as channels in the other dataset based on the linear-regression results, and use those to better match interpolated co-ordinates to data traces without an interpretive step.