Abstract
Seismic and geoelectric methods are often used in the exploration
of near-surface structures. Generally, these two methods give, independently
of one other, a sufficiently exact model of the geological structure.
However, sometimes the inversion of the seismic or geoelectric data
fails. These failures can be avoided by combining various methods
in one joint inversion which leads to much better parameter estimations
of the near-surface underground than the independent inversions.
In the companion paper (Part I: basic ideas), it was demonstrated
theoretically that a joint inversion, using dispersive Rayleigh and
Love waves in combination with the well-known methods of DC resistivity
sounding, such as Schlumberger, radial dipole-dipole and pole-pole
arrays, provides a better parameter estimation. Two applications
are shown: a five layer structure in Borsod County, Hungary, and
a three-layer structure in Thüringen, Germany. Layer thicknesses,
wave velocities and resistivities are determined. Of course, the
field data sets obtained from the 'real world' are not as complete
and as good as the synthetic data sets in the theoretical Part I.
In both applications, relative model distances, in percentages, serve
as quality control factors for the different inversions; the lower
the relative distance, the better the inversion result. In the Borsod
field case, Love wave group slowness data and Schlumberger, radial
dipole-dipole and pole-pole (i.e two-electrode) data sets are processed.
The independent inversion performed using the Love wave data leads
to a relative model distance of 155\%. An independent Schlumberger
inversion results in 41\%, a joint geoelectric inversion of all data
sets in 15\%, a joint inversion of Love wave data and all geoelectric
data sets in 15\% and the robust joint inversion of Love wave data
and the three geoelectric data sets in 10\%. In the Thüringen
field case, only Rayleigh wave group slowness data and Schlumberger
data were available. The independent inversion using Rayleigh wave
data results in a relative model distance of 19\%. The independent
inversion performed using Schlumberger data leads to 34\%, the joint
and robust joint inversion of Rayleigh wave and Schlumberger data
gave results of 18\% and 20\%, respectively.
Links and resources
Tags