Master Thesis by Stefan Jansen – Niels Bohr Institute - University of Copenhagen

Niels Bohr Institute > Calendar > 2010 > Master Thesis by Stef...

Master Thesis by Stefan Jansen

Parameter Investigation for Subsurface Tomography with Refraction Seimic Data

By Stefan Jansen

For planned highway and railway improvements in Norway tomographic images were modelled supplementary to results obtained from a layer based interpretation tool like for example the Generalized Reciprocal Method (GRM).
In general good agreements between these methods were achieved, but some inconsistencies led to questions whether the tomographic inversions can be improved.
In this master thesis the tomographic inversion tool Rayfract has been examined thoroughly with the aim of determining settings and parameters leading to tomographic images with increased velocity contrast.
With the help of synthetic models containing various kinds of anomalies the sensitivity of the tool has been investigated. A thorough parameter investigation on one of these models lead to the determination of six combinations of settings and parameters for obtaining an improved absolute RMS-error and model resemblance.
Verification of these combinations with a model resembling one of the profiles from Norway showed unexpected responses. Finally these findings were applied to field data.
Over all this work shows that two parameters, Smoothing and wavepath width, are decisive for the quality of all conducted inversions. These parameters are both directly or indirectly linked to smoothing of the result. Even if a fixed recipe for the choice of parameters can not be given, as they are dependent on the topography and subsurface condition of each individual profile, this thesis clearly defines the parameters in focus and how to perform a quick and reliable parameter test for each new situation.

Klaus Mosegaard University of Copenhagen, Roger Wisèn, Rambøll Danmark