Gravity and magnetic methods - Geophysics for the Mineral Exploration Geoscientist - page 124

similarly simpler when calculated from RTP data
( Figs. 3.26b and e ).
There are, however, some practical dif culties with the
RTP transform. The magnetic inclination and declination
of the inducing field are required for the transformation
(obtained from the IGRF; see Section 3.6.2 ) , and it is
normally assumed that the magnetism of all the rocks in
the area is parallel to the geomagnetic field and by impli-
cation is entirely induced. Remanent magnetism, and to a
lesser extent self-demagnetisation and AMS (see Section
3.2.3 ) , change a body

characteristic of gravity anomalies situated directly over
the magnetic source. In this regard it is similar (but not
identical) to the RTP response. The total horizontal gradi-
ent of pseudogravity can be used to locate contacts in the
same way as the gradient of normal gravity (see Section
3.7.4 ) . The example shown in Fig. 3.26g con rms the
formation of peaks over contacts. The mapping of contacts
is well demonstrated by the double peaked response over
stratigraphic sourced anomalies in the Broken Hill data,
e.g. (8) in Fig. 3.28h , although there is less detail than in the
various magnetic derivative images.
Comparisons between computed pseudogravity and
actual gravity data can provide information about the
magnitude and distribution of density in the magnetic
sources and the shape and size of the respective portions
of the source causing the gravity and magnetic responses.
We emphasise that pseudogravity is not the actual gravity
response and is related only to the magnetic rocks in the
survey area; it is not the total gravity response of all the
(magnetic plus non-magnetic) geology.

is magnetism so it is not parallel to the
geomagnetic field everywhere and, therefore, anomalies
will not be properly transformed. The distorted transform-
ations are most noticeable where the remanence is strong
and in a direction significantly different from that of the
Earth

'

field.
RTP algorithms often perform well when the geomag-
netic

' s

field inclination is steep, but some algorithms are
much less effective when it is shallow, i.e. close to and at
the magnetic equator (Li, 2008 ) . Prominent north south

oriented artefacts are introduced into the transformed
data, as demonstrated by Cooper and Cowan ( 2003 ). An
alternate strategy is to reduce the data to the equator
instead of the pole; but although this generally places a
response over the magnetic source, it is negative (see
Fig. 3.26a ) even though the body is more strongly magnet-
ised than its surrounds, and the anomaly does not mimic
the geometry of the source.
The Broken Hill data are notably different after reducing
to the pole ( Figs. 3.28a and b ). Note how the dipolar
anomalies in the TMI data have become monopoles (7).
The anomalies in the first vertical derivative data are also
more localised to their sources ( Figs. 3.28c and d ).

-

3.7.3 Wavelength filters

A common filtering operation applied to potential eld
data involves directly modifying the frequency content of
the data (see online Appendix 2 ) . These are the wave-
length/frequency filters described in Frequency/wavelength
in Section 2.7.4.4 .
Separation of local and regional responses is often a very
effective enhancement to potential field data. It can be an
important precursor in the application of the various
derivative filters and transforms described in the next
sections, because their ability to accurately resolve features
of gravity and magnetic anomalies can be strongly
degraded in the presence of a regional

3.7.2.2 Pseudogravity
The pseudogravity operator transforms the TMI anomaly
into the gravity-like response that would be obtained if the
body

field.

3.7.3.1 Spectral filters
Gravity or magnetic datasets contain variations with a
range of wavelengths, and usually this is a continuous
range. Deep sources and broad near-surface sources con-
tribute the longer wavelengths, whilst smaller near-surface
sources produce the shorter wavelengths (see Section 2.3 ).
It is important to appreciate that wavelength filtering to
isolate responses from particular depth intervals, so called
depth slicing, is impossible. This is due to overlap in the
frequency spectra of the responses from different depths.
Consequently, it is impossible for wavelength filtering to
completely separate the responses of a series of discrete

s magnetism were replaced with the same density
distribution, and with the density as a multiple of the
magnetisation. By treating the magnetism like density, it
aims to simplify the analysis of magnetic data. Like the
RTP operator, pseudogravity assumes that the magnetism
of the source is parallel to the geomagnetic field, and by
implication is entirely induced; so strong remanent mag-
netism and AMS are sources of error.
The pseudogravity data shown in Fig. 3.26f , obtained
from the TMI data

'

shown in Fig. 3.26a , exhibit

the smoother, longer-wavelength monopole responses