[Therion] Gradient out of range - does not process

[Andrew Atkinson]

On 19/12/2020 11:40, Martin Budaj wrote:
> Here is a comparison (taken from a surveying textbook by Ryšavý, 1949) of an error Δ in position of the last point on a traverse of length D, with an average leg length S, and with a mean angle error α for theodolite and β for compass (in radians):

That is interesting, in a narrow cave passage which winds lots, plus the
need to accommodate the theodalite, the choice of shot length is very
limited, we did have a section with lots of short legs, plus real
difficulty with the set up, so I suspect that is where the error
occurred. That said on a previous show cave we also got a poor loop
closure and that was long legs. A third cave where we had a loop we got
a brilliant result, 2mm for a traverse of 120m (the first cave we did,
so beginners luck!.)
The machine is borrowed, from a company that uses it for large civil
engineering projects so it should be properly calibrated, the operator
is the person borrowing it after he had used it in his day job, so he is
very competent. There are so many places where a slight error can occur,
especially if the loop is not competed in a single session, but it is
the best we can do.


>  
> When using theodolite, always use multiple directions for orienting the
> survey (shots to multiple distant points with known coordinates, the
> more distant the better). I would hesitate to use magnetic legs for this
> purpose (there is much uncertainty with the declination as the models
> are not particularly precise, there is a daily variation of declination
> (up to 0.2 degrees!) and there might be local magnetic anomalies).


In this case the cave comes out in a valley with little gps signal, the
position and orientation was set from the 2 gps positions we could get,
the a traverse to the cave. So I have very little idea how that ties in
with the declination model. I have been assuming that the direction it
gave where OS grid as that is the format of the positions, I have no way
to check this. I have set the declination to 0 to stop date and position
correction. I think this it right, but advice welcome.

thanks

Andrew