Geoscience Reference
In-Depth Information
5.7.4.3
Inferred conductivity
The second form of data display is a coloured-pixel image
showing inferred conductivity variations in the form of a
map or cross-section (
Fig. 5.83b
) and occasionally as a data
volume. This form of display relies on inverse modelling of
the data to compute the conductivity at a range of depths at
each measurement location. The advantage of this kind of
display is that the results can be readily interpreted since
they resemble a cross-section, a geological map or a volume
of the subsurface. Two computational methods are in
common use: layered Earth inversion (LEI) and current-
depth imaging (CDI). Both are computationally fast, a
requirement for working with the large AEM datasets.
LEI assumes that the ground is composed of a series
of discrete horizontal conductivity layers of in
nite extent,
i.e. a 1D model (see One-dimensional model in
Section
the layers, the number of layers being chosen manually.
A two-layered model can be very useful for mapping the
thickness and conductivity of conductive overburden. CDI
determines subsurface conductivity by analysing variations
in the measured secondary field decay in terms of the
depth
LEI and CDI make important simplifying assumptions
about the subsurface. They suffer from non-uniqueness
and may contain artefacts (see
Sections 2.11.4
and
chie
y due to 2D or 3D conductivity structures, they
over-estimate depth to the source and under-estimate the
conductivity, so computed depth is often unreliable and a
particular conductivity
depth map/image will not neces-
sarily represent the true conductivity at that depth,
although relative variations may be realistic. Despite these
limitations, LEI and CDI are very convenient ways of
transforming and displaying the multichannel survey data
and they produce good results where the inherent assump-
tions are met. They are also very useful for targeting
compact zones of anomalous conductivity for more precise
analysis of their associated anomaly pro
-
les (see Pro
le
analysis in
Section 5.7.5.3
).
5.7.5
Interpretation of EM data
Multichannel EM data are interpreted using both spatial
and temporal variations in the responses, i.e. how the
amplitudes of the decay channels vary with position and
how a response decays at a speci
c location. The spatial
characteristics are the amplitude, shape and width of
anomalies, the nature of gradients in them and zero
cross-over locations. There are considerable differences in
the form of the responses of the three measured compon-
ents (X, Y and Z) (see
Section 5.7.1.5
) but, for all compon-
ents, their spatial characteristics relate to the geometry of
anomalous zones of conductivity. Temporal characteristics
provide information about the class of conductor, i.e.
unconned or conned (see
Section 5.7.2
)
, and its
conductance, and also about the vertical variation in con-
ductivity beneath the station.
Interpretation of EM data is hampered by the different
characteristics of the various loop con
time relationship of the downward moving eddy
current
-
(see
Section 5.7.2.1
)
. It also assumes
that the subsurface has a
'
smoke ring
'
flat-lying conductivity structure,
but without the restriction of a
fixed number of layers. CDI
methods can be more effective in areas where the vertical
conductivity distribution is more complex than the simple
layered-Earth model, for example where the conductivity
changes gradually with depth.
Note that for the case of the in-loop con
guration (see
Moving-loop mode in
Section 5.7.3.2
)
, which includes all
the helicopter AEM systems (see
Section 5.9.2.2
), the hori-
zontal (X and Y) components are zero over a horizontally
layered Earth; so only the vertical (Z) component can be
used to produce the 1D models. The horizontal compon-
ents contain additional information where the subsurface
conductivity distribution is 2- or 3D, in which case 2D/3D
modelling is required in order to make use of this data.
For both methods, the individual inversions (models) are
merged and smoothed to form a continuous conductivity
gurations and the
various EM systems, which determine the response of
the ground. This makes comparison between different
EM systems and con
gurations dif
cult. Also, coincidence
between the step and impulse responses cannot always be
expected; each varies differently with the quality of the
conductor (see Conductor quality in
Section 5.7.2.3
). For
example, a strong impulse response may be obtained from
the lower-conductivity disseminated part of a mineral
deposit and not at all from the much higher conductivity
of its massive zone, the latter producing only a stronger
step response.
-
depth distribution and displayed as parasections, maps for
selected
and paravolumes (see
Section 5.9.5.1
). The
choice between using LEI or CDI should be strongly guided
by the nature of the geological environment. They can be
usefully applied where the conductivity structures are suffi-
-
ciently wide so that they appear locally as 1D in nature;
otherwise artefacts are produced.
'
depths
'
Search WWH ::
Custom Search