As an extension of work on near-field Q-factor recovery we apply Q(z)-inversion by forward modeling to VSP model data generated from Ross Lake well-logs applying a multi-interface Sommerfeld integral algorithm. An intrinsic Q(z) model is derived from Ross Lake P-wave velocities by applying an empirical equation. When computing sweptQ forward models with the same Sommerfeld algorithm as is used for generating the VSP model data the sum of squared differences show minima at the correct intrinsic Q-values; velocities and densities are assumed to be known in these forward model computations. Applying a Golden-Section-Search algorithm for computational efficiency leads to a local-minimum problem at some depths which is solved by adding a local-minimum search extension. For this ideal noise free model situation intrinsic Q-factor recovery is very successful.
View full article as PDF (0.32 Mb)