Gravity and magnetic methods - Geophysics for the Mineral Exploration Geoscientist - page 117

which case it is treated as a tie-line cross-over error and
removed during the levelling process (see Section 3.6.4 ).

heights is inherently unstable and can introduce noise that
degrades the quality of the data.
Where the terrain is very rugged and the hills are formed
of magnetic material, a magnetic measurement made in a
valley will be in uenced by magnetic rocks located below,
adjacent to, and above the magnetometer. In this case,
magnetic pro les measured across rugged terrain exhibit
complex responses due to the rapidly changing ground
clearance in airborne surveys, sloping survey plane and,
particularly in valleys, interference of magnetic material
contained in the neighbouring hills (Mudge, 1988 ). The
varying ground clearance and the shape of the terrain need
to be included in the modelling of target anomalies (see
Section 2.11.3 ) . Otherwise, to correct for these effects
would require detailed knowledge of the magnetism of
the terrain, which is not available (an example of the
geophysical paradox, cf. Section 1.3 ).

3.6.2 Regional variations in field strength

The geomagnetic field (see Section 3.5.1 ) appears as a very
long-wavelength regional field. Removing this increases the
resolution of magnetic anomalies associated with crustal
sources. For small surveys, the geomagnetic field varies
smoothly and by only a small amount and can be removed
as part of a single regional gradient (see Section 2.9.2 ).
For magnetic surveys that extend over large areas, such
as airborne surveys, the gradient of the geomagnetic field
can be removed by simply computing the field strength
from the IGRF model ( Fig. 3.21 ) for the time and location
of each measurement, and subtracting these from the
observations. The correction also compensates for the
slower secular variations of the geomagnetic

field, import-
ant for surveys that extend over long periods, such as
airborne surveys as these typically take days or weeks to
complete, and even months and years for national-scale
surveys.

3.6.4 Levelling

The final stage in the reduction of magnetic data is level-
ling (see Section 2.7.1.2 ) based on the repeat readings
associated with tie lines crossing the survey lines.
A detailed description of the procedures for aeromagnetic
data is provided by Luyendyk ( 1997 ) and Saul and Pearson
( 1998 ). Errors in the magnetic levels of each data point,
remaining after the application of the various corrections
described above, appear in aeromagnetic datasets as corru-
gations parallel to the survey line direction. Levelling redis-
tributes these around the whole dataset, and any further
remaining errors are usually removed with microlevelling
(see Section 2.7.1.3 ).
The quality of the data reduction can be accessed by
viewing the levelled data with shaded relief (see Section
2.8.2.3 ) , with the illumination perpendicular to the survey
line direction, optimum for highlighting across-line level-
ling errors. Derivative images, being sensitive to gradients,
often reveal residual errors when the levelled TMI data
appear to be free of corrugations.

3.6.3 Terrain clearance effects

In the absence of anomalous features, and unlike the grav-
ity field, the magnetic field varies little with distance from
the Earth. However, in the presence of crustal magnetic
features, variations in the distance between the magnetom-
eter and the Earth s surface can be important (Ugalde and

Morris, 2009 ) . This is usually not a problem for ground
surveys, but it is for airborne surveys. Also, when process-
ing and interpreting magnetic data, it is generally assumed
that the survey plane is horizontal and always above the
rocks, but this is often not the case in rugged terrain.
It is common for geological controls on topography to
produce strike-parallel ridges and valleys. Aerial surveying
across strike causes changes in ground clearance, as shown
in Fig. 2.6 . For these cases, it is possible to estimate what
the measurements would have been had the ideal ight
path with constant terrain clearance been achieved. In
order to mimic an ideal drape- own survey (see Section
2.4.1.1 ) , a drape correction is applied (Flis and Cowan,
2000 ) . The basis for this is a procedure known as continu-
ation, whereby a measurement at one height can be used to
determine the equivalent measurement at another height
(see Section 3.7.3.2 ). In practice the correction is not rou-
tinely applied because continuation of data to lower

'

3.6.5 Example of the reduction
of aeromagnetic data

Figure 3.24 shows an aeromagnetic survey over an area of
moderate terrain in the Northern Territory, Australia. The
mean terrain clearance is 80 m, but narrow ridges due to
erosion resistant units caused signi cant height variations
in their vicinity. The survey lines are spaced 100 m apart