Constrained 1D joint inversion of seismic surface waves and P‐refraction traveltimes

Abstract

We present a joint inversion scheme that couples P‐wave refraction and seismic surface‐wave data for a layered subsurface. An algorithm is implemented with a damped least‐squares approach. The estimated parameters are S‐ and P‐wave velocities and layer thicknesses, while densities are assumed constant during inversion. The coupling is both geometric and physical: layer thicknesses are the same for S‐ and P‐wave velocity profiles and P‐wave velocities enter in both forward algorithms. Sensitivity analysis, performed on synthetic data, reveals that surface‐wave dispersion curves can be sensitive also to P‐wave velocity of some layers (especially for Poisson's ratio values smaller than about 0.35), allowing synergic resolution of this parameter. Applications on both synthetic and field data show that the proposed approach mitigates the hidden layer problem of seismic refraction and leads to more accurate results than individual inversions also for surface waves. Additional constraints on the objective function on a priori Poisson's ratio values allow unrealistic and not admissible VP and VS values to be avoided; such constraints were applied in one field case considering the a priori information available about water‐table depth. It is also shown that estimation of porosity can help the selection of the proper constraint on a priori Poisson's ratio.

Joint inversion of Rayleigh-wave dispersion and P-wave refraction data for laterally varying layered models

“…An alternative possibility is to define a datadriven model based on the information that can be directly retrieved from the individual data sets through a preliminary simplified interpretation. Piatti et al (2012) suggest starting from the preliminary interpretation of P-wave traveltimes with simplified intercept time approach and to split the obtained layers in two, to allow the presence of further layers that could be hidden in the preliminary interpretation. This provides the initial model for P-wave velocity.…”

“…Forbriger (2003) successfully applies the joint inversion of the complex Fourier-Bessel expansion coefficient of the recorded wavefield together with Pwave arrival times on two field cases. Piatti et al (2012) present a least-squares joint inversion of surface-wave dispersion curves and P-wave first arrivals in which the two different data sets are used to retrieve V P and V S profiles where layer thickness and V P are shared by the two methods whereas V S is an unknown parameter only with respect to SWA but can be linked to V P by introducing physical laws (i.e., Poisson's ratio).…”

P- and S-wave velocity model estimation from surface wave data

“…The example reported in Fig.1 refers to a joint inversion algorithm proposed by Patti et al (2013), where P-and Swave velocity models are constrained by imposing limits to Poisson's ratio values [15] . Several applications on synthetic and field data showed the improvement introduced by the joint inversion with respect to individual inversion.…”

“…The extension to 2D/3D dataset was proposed by Re [1][2][3][4][5][6][7][8][9][10][11][12][13][14][15][16][17][18] .…”

Joint inversion of Love-wave dispersion curves and SH-wave firstarrivals for near-surface SH-wave velocity

“…Schuler made an inversion of dispersion curves of Love and Rayleigh waves and P-and Swave first arrivals to calculate parameters of a layered earth model with increasing velocity with depth. Piatti et al adopted a damped least square algorithm in inversion of surface-wave phase velocities and P-wave refraction traveltimes [9] . In order to verify the effect of joint inversion, we present here a joint inversion of Love-wave dispersion curves and SHwave first arrival traveltimes using a damped least-squares algorithm.…”