Geoscience Reference
In-Depth Information
Numerical techniques are available for rapidly
computing the EM responses of parametric models (see
Section 2.11.1.1
)
set in a non-conducting background,
such as a sphere, prism, dipping plate or
at sheet, and
are suitable for analysing individual anomalies in resistive
environments. The responses of multiple conductors,
including a discrete conductor in a conducting back-
ground and discrete 3D bodies, involve greater math-
ematical complexity. As with all modelling, limitations
imposed by mathematical complexities affect the accur-
acy of the result, but non-uniqueness predominates (see
Section 2.11.4
)
. Constraining one or more parameters
of the model with known geological information helps
to reduce non-uniqueness, as does combining the EM
data with the independent results of other geophysical
methods.
An alternate strategy for anomaly analysis focuses on
the eddy current
5.7.6.1
Conductive environments
The response of conductive overburden and examples
of its effect on the target response are demonstrated by
the model responses in
Figs. 5.84
to
5.87
. In addition to
the overburden response obscuring the weaker early-time
responses of underlying conductors (see Late-time
underlying conductors can induce eddy currents in the
overburden layer,
attenuating and broadening their
responses.
Variations in conductivity and thickness of the near-
surface layer can produce responses that superimpose
geological noise on the responses of underlying conduct-
ors, or can be easily mistaken for deeper conductive bodies
turn-off, producing a reduction in the high-frequency
components (see
Section 5.7.3.1
) reduces the response of
conductive overburden and conductive host rocks. The
physically larger eddy current system produced by the
elevated loop of an AEM system reduces sensitivity to
overburden and variations within it. The various effects
diminish when the receiver is distant from the overburden,
as in DHEM surveying. Note that a high-resistivity surface
layer does not present a problem for EM systems because
induction will not occur in it; only the underlying conduct-
ivity distribution produces eddy currents.
Current channelling (see
Section 5.7.2.4
) occurs in
conductive environments, and common causes are a
subsurface conductor electrically connected to conductive
overburden, conductive host rocks, fault zones and forma-
tional conductors. Its effect changes with time; a
flow in the conductive zone rather than
its geometrical form. The current
filament model is
described by a circular current-carrying loop in free space
(see
Section 5.2.2.1
), and its response is very easy and fast
to compute. It simulates the eddy current circulating in a
conductor at a speci
c delay time. The centre of the
current system can be located and its dip estimated; so
migration of the current system through the conductor
with time can be investigated (see
Section 5.7.1.4
)
. For a
plate-like conductor the eddy current circulates in the
plane of the conductor, so the attitude of the conductor
can be determined. For a sphere-like body with homoge-
neous conductivity, the plane of the circulation is perpen-
dicular to the primary field direction at the conductor.
Multiple and complex migration paths can be expected in
large complexly shaped heterogeneous bodies. The use of
multiple transmitter loop locations (see
Section 5.7.1.3
)
and three-component data (see
Section 5.7.1.5
)
aids the
investigation of current migration in this class of con-
ductor. An example of using the current
filament model
is presented in
Section 5.8.3.1
.
'
decay may appear at a later time when it has diminished.
Current channelling can usefully increase the response of
poor conductors, but the strong responses it produces
makes it difficult to distinguish between poor and good
conductors. Interpretation of a conductor
'
normal
s parameters
is likely to be erroneous unless the presence of current
channelling can be identified and corrected (usually
difficult to do).
'
5.7.6
Interpretation pitfalls
5.7.6.2
Interacting responses
Interaction between adjacent conductors produces
responses that distort genuine target responses and can
masquerade as target anomalies. The proximity of neigh-
bouring conductors can have a signi
cant effect on the
resolution of weaker conductors. For particular system
con
gurations and delay times, target conductors can be
electrically shielded by larger neighbouring conductors
EM measurements are subject to serious interpretation
pitfalls unless procedures are adopted to identify and
account for them in the interpretation of survey data. Also,
like many kinds of geophysical data, inversion of EM data
suffers from ambiguity, or non-uniqueness (see
Section
the observed data within error.
Search WWH ::
Custom Search