Loop Closures

[Olly Betts]

On Wed, Nov 12, 2014 at 12:18:54AM +0000, Wookey wrote:
> +++ Olly Betts [2014-11-11 20:02 +0000]:
> > [Replying to the list too as I think this is of general interest]
> > It's also easy to let the laser beam "flick off" the intended target
> > and end up measuring to a different point to the target station
> 
> In my experience the device algorithms are pretty good at spotting
> this and a laser flicking on and off a station will either give an
> error or the nearest stable-ish point. So it is hard to get a 'far too
> far' reading. If there is something in front then you can get a 'much
> too short' reading.

That's somewhat reassuring, though the case where the reading is wrong
by a small amount is still a concern, especially when considering what
the effective accuracy is (and if the reading is metres out then alert
surveyors may twig, but a small error isn't something you can spot in
that way).
 
> > I don't know how accurate the compass and clino readings from a
> > disto-x are expected to be.
> 
> It's a good question and we should probably do some proper analysis.
> We have a fair amount of data from various training courses and
> Austria and elsewhere.
> 
> My (very much 'gut') feeling is that you can reasonably expect 1% or
> better with distoX (or SAP+disto), whereas you can expect 2% or better
> with conventional instruments. Something of that ballpark anyway.

Though as I noted before, percentage misclosure is really not a good
measure because a longer loop has a smaller expected percentage
misclosure.

If you consider N independent measurements with the same SD, the
expected percentage error is proportional to 1/sqrt(N), and with
different SDs it'll behave similarly.  So for example, a 1000 leg loop
would be expected to have about 1/10th of the percentage error of a 10
leg loop - e.g. 2% versus 0.2%.

Cheers,
    Olly