Image processing - magnetometry

[Simon Cox]

At  1:00 PM 4/7/95 +0100, Mr PM van Leusen wrote:
>I have a file of magenetometer data taken at 1 m res (ew) and .25 m res (ns)
>for a 14 ha area of the Roman town of Wroxeter (UK). It shows quite a lot
>of detail but the image
                    ^^^
Do you still have the original "site" observations or only a gridded version?

>also contains a) spikes caused by metal objects in
>the soil, b) some horizontal striping caused by the measurement technique,
>and c) 'shading' effect caused by the angle of polar magnetics (about 53
>degrees).

These are standard problems most commonly encountered in airborne magnetic
surveys in exploration geophysics.  I'm not an expert, but the following
issues probably apply.

(i)  the spikes - these are real information and should be treated as such.
If they are not what you are looking for then you have a number of
choices:   you can model them and subtract the effect from the image;  if
they are small and shallow and what you are looking for is large and deep
then you can re-survey at a higher "altitude" [or filter ("upward
continue") the data to get the same effect]; if the signal from these
really swamps the signal you are looking for, then bad luck.

(ii)  there are a couple of possible explanations for the striping - either
related to (a) levelling (ie mistakes in the correction made for the
elevation of the magnetometer) or (b) the gridding technique.   Do you have
"tie lines" as well as survey lines?  it is common to use a few orthogonal
tie-lines to correct the levelling on the main survey lines, though it is
also common to actually discard the tie-line data after it has been used
for levelling but before gridding (I only found this out this week and was
a bit surprised!).  Alternatively, if a segmented method was used for
gridding, then an inadequate overlap can cause artefacts which look like
stripes.  Try increasing the segmentation overlap.

(iii)  It is common to use a technique called "reduction to the pole" to
correct for the magnetic inclination at the survey latitude.  I'm not quite
sure what the algorithm is, but the effect is to fix up the anomalies so
that they are symmetric over the source, rather than having an assymmetric
peak-and-trough in a N-S direction.  I suggest you either find a friendly
exploration geophysicist or go to the library, but I think that many of the
key words are in my chat here.

>Has anyone ver dealt with such data in Grass before and would they be
>willing to share their experience with me?

No. I've not actually processed magnetics in GRASS  (but I did go to
Wroxeter/Uriconium on a primary school trip almost 30 years ago!).

Good luck                       Simon Cox

___________________________________________________________
Dr Simon Cox                          __  \
CSIRO Exploration & Mining         ,~'  L_|\    Australian
39 Fairway, PO Box 437,         ;-'         \   Geodynamics
Nedlands, WA  6009  Australia   (            \  Cooperative
      Phone +61 9 389 8421      +    ___     /  Research
      Fax   +61 9 389 1906       L~~'   "\__/   Centre
simon at ned.dem.csiro.au                     W
AGCRC info>>   http://www.ned.dem.csiro.au/AGCRC/
___________________________________________________________