Geophysical data acquisition, processing and interpretation - Geophysics for the Mineral Exploration Geoscientist - page 60

low-amplitude ones. Shaded relief also creates spurious
apparent cross-cutting relationships ( Fig. 2.34 ) .

or larger features of less interest, i.e. the background or
regional response. The regional response is a form of
geological environmental noise (see Section 2.4 ). The
target anomaly and the regional field can be separated in
a process known as regional removal ( Fig. 2.37 ) . The
variation remaining after the regional has been removed
is known as the residual response. This is the shorter-
wavelength variation correlating with the shallower geol-
ogy, and is usually the response of interest. In Fig. 2.37 it is
the response originating from the mineralised environ-
ment. Removing the regional response is important for
quantitative analysis (see Section 2.11 ) and can also greatly
facilitate qualitative interpretation (see Section 2.10 ) , as
shown in Section 3.11.1 .
As with any consideration of signal and noise, the vari-
ations of interest depend on the interpreter

Information about infrastructure that might produce
artefacts in the geophysical data, e.g. powerlines,
pipelines, railways, buildings or open pits, needs to be
available. This can usually be obtained from cadastral
maps and aerial photography. Responses from these
features re ect their form: powerlines create linear arte-
facts, buildings localised responses etc.

￿

When several kinds of geophysical data are being inter-
preted together, which is generally a good strategy, it is
very important to consider the geological controls on
their individual responses. For example, a geological
feature may have a wide gravity response, but only a
particular region of the feature may contain minerals
that produce a magnetic response, electrically conductive
minerals may be con

￿

s requirements
of the data. Consider the hypothetical example of a massive
nickel sulphide deposit located at the contact between
ma c and ultrama c rocks in a greenstone belt, e.g.
the komatiitic peridotite-hosted deposits that occur in the
Kambalda area in Western Australia ( Fig. 2.38 ). In the
early stages of exploration, prospective contacts may be
the target being sought. In this case, the contact response
is a residual (signal) superimposed on the regional (noise)
response of, say, the base of the greenstone succession.
When the contact has been located and exploration focuses
along it, responses from nickel sulphide mineralisation
become the target. The contact response is now the
regional (noise), and the response of the mineralisation is
the residual (signal).
Ideally, the responses due to the deeper and/or larger
geological features are computed (modelled) and sub-
tracted from the data. However, accurately defining the
regional response requires detailed knowledge of the local
geology, which depends on the geological interpretation of
the area (an example of the geophysical paradox, cf.
Section 1.3 ). Non-uniqueness (see Section 2.11.4 ) also con-
tributes to the problem because interference between adja-
cent local anomalies may produce a longer-wavelength
composite response which could be mistaken for part of
the regional response. Also, the local and the regional
responses must both be properly de ned in the original
data series, i.e. the sampling interval needs to satisfy the
conditions of the sampling theorem for the higher-
frequency local response ( Section 2.6.1 ), and the data series
must extend far enough in distance to define the longer-
wavelength regional response. Unfortunately, then, the
regional variation cannot be accurately calculated or

'

ned to other parts of the body, and
radioactivity would only be observed where radioactive
minerals are exposed at the surface. Often there is partial
correspondence between gravity and magnetic datasets,
both primarily re ecting bedrock lithological variations.
However, there is no requirement for their responses to
correspond exactly with, say, the distributions of radio-
elements in the near-surface as derived from radiometric
data, which are much more likely to correlate with
spectral remote sensing data. Sometimes the different
depths

by the different geophysical methods
cause lateral offsets between the responses of the differ-
ent data types if the source is dipping. For example,
electrical measurements may respond well to the con-
ductive (weathered) near-surface geology, whilst gravity
and magnetic data may contain responses from the
unweathered deeper regions.

' probed '

To create an interpretation which accounts for the above
requires knowledge that is method speci c, together with a

good understanding of the nature of geophysical responses,
and also how the data were acquired and processed.

2.9.2 Removing the regional response

The measured variations in the geophysical parameter of
interest almost invariably consist of a series of superim-
posed responses from geologic features of different dimen-
sions and/or depths in the vicinity of the measurement.
A common manifestation of this is the interference (see
Appendix 2 ) between an anomaly of interest and the
longer-wavelength variations associated with deeper and/