bedding planes, rose diagrams

[Andy Waddington on Survey stuff]

A brief note on Rose diagrams and their 3d equivalents...

When SU went through its various incarnations as a FORTRAN (on various
mainframes) and then an APL (on an IBM mainframe) program (before becoming a
BASIC program for the BBC micro), it incorporated output as a rose diagram.
I found this very useful for fairly level caves which had significant
amounts of joint controlled passage, even when that joint control was not
very obvious on the drawn-up survey. Sometimes study of the rose diagram
drew my attention to joint control I had missed, and a reevaluation of the
sketching (or another visit to the cave) showed that the drawn up survey did
not represent the cave as well as it should - joint control being a bit more
obvious than as drawn. In that sense, Rose diagrams were not only useful for
eliciting the structure of the cave, but actually contributed to the quality
of the drawing. Thus I think they are a Good Thing (TM).

In our Austria dataset, it is pretty obvious that there is a lot of joint or
fault control, but extracting the dip and strike of these features is not
straightforward (for example, think of a passage which obliquely descends a
hading fault - Boulder Alley in KH for those familiar with the Survex
example dataset. The strike of this fault is significantly different from
the direction the passage heads on the cave plan and cannot be extracted
from any drawn-up survey). Getting an answer is largely a matter of spinning
the cave in CaveRot until a plane "flattens out" and then reading off the
figures. This utilises the human eye and brain's excellent ability to
extract patterns from confused data. Where vertical series have formed on a
series of parallel or perpendicular faults, this confusion can confound even
the practiced eye. The idea that a program might be able to do better is
certainly appealing.

However, one note of caution: "Real" surveyors show a definite bias for the
vertical in such series, because vertical legs are easier to measure. A
typical pitch series in, for example, the Left Hand route in KH, will drop
vertical legs to a ledge, then a short leg up to the next bolt, then another
vertical. There is little or no chance that any of the surveyed legs will
actually be parallel to the hading fault(s) controlling the cave. This also
reflects the structure of the cave, where water has cut down below the joint
or bedding on which the passage was initiated.

If you actually want to know how the cave formed - there is a definite case
for doing a survey specifically for that purpose - taking dips and strikes
of beds and joints, for example, or surveying along the roof tube of a
passage instead of the normal cavers' route, rather than just hoping that
something useful will simply fall out of the data for a normal survey.

Older surveyors may remember a paper from the old CRG Surveying symposium
in which a stream passage in Ireland was surveyed and the meanders then
fourier-analysed to derive stream flow rate [this is a technique used on
aerial photos of surface streams]. The author had to rely on ordinary data
from a normal survey trip, which was less than satisfactory because survey
points were always on the walls, typically alternate sides. Here, a specific
purpose survey would have used points in the centre of the passage
throughout (although something close to this could be derived if good LRUD
data were recorded). More challenging - surveys were needed at several
levels through the meander in order to derive historical flow rate data -
after all, the flow rate at the present day was much easier to measure than
to derive from the survey :-)

So yes, I'd like to see these sorts of tools, but since many would need
to operate on a set of data modified from the original (perhaps a set of
"legs" down the centre of a passage at a specific height above the floor
derived from the original line and its LRUD data) they are perhaps less
useful as parts of the mainstream surveying package, which would be best
kept "lean and mean".

Andy