XML: was Bedding planes, rose diagrams and bug fix (fwd)

[John Halleck]

On Fri, 5 May 2000, Lev Bishop wrote:

> Date: Fri, 5 May 2000 01:28:49 +0100 (BST)
> From: Lev Bishop <lev.bishop@queens.oxford.ac.uk>
> To: John Halleck <John.Halleck@utah.edu>
> Cc: Survex User Group <survex@survex.com>
> Subject: Re: XML:  was Bedding planes, rose diagrams and bug fix (fwd)
> 
> On Wed, 3 May 2000, John Halleck wrote:
> 
> > 1) Covariance of the instrument observations.  A symmetric 3x3 matrix, so one
> >    only needs to store 6 numbers.
> >    This is generally the same for an entire cave, but can change from shot to
> >    shot in truely bizarre surveys.
> >    99.9% of the time it is the same for an entire survey... so I'd consider it
> >    a candidate for
> >    <sometag D="..." A="..." I="..." DA="..." DI="..." AI="..."> survey </sometag>>
> >    style marking.
> >    [I'm assuming Distance Asimuth Inclination as was is measured.]
> >    (D, A, and I are manditory, DA, DI, AI are optional)
> 
> Under what circumstances will the off-diagonal elements ever be nonzero?

  Anyone using an all-in-one station can have that.
  Errors on theodolites can be other than independent for various reasons.

  Is anyone likely to notice this is a cave survey?  I doubt it.
  Do I think that there is no excuse for not taking the extra effort
  do it right?  Yep.

> If you are measuring with seperate compass, clino, tape then you have 3
> independent measurements.

  Absolutely correct.

>  Perhaps if you have some kind of instrument
> which relies on a platform that needs levelling,

  Most total-stations for example.

  But the error that more effects things is the instrument and target
  centering errors.

> or you are electronically
> gimballing your compass with you electronic clino, the AI element could be
> nonzero, but I can't think how DA or DI can be nonzero.

  If the electronic range finder is also part of the same station then
  centering errors also affect it.

  If you arrived at DA or DI *indirectly*, then it can.
  (The output of a triangulation, for example.)
  I don't know about how it is done in the UK, but sometimes here someone
  will do a triangulation, and then add to the data the reduced data.

> However, I can see a few occasions under which the covariance matrix will
> be different for each shot. For example, if the instruments are
> electronic, they might output an estimated accuracy with every shot (the
> leica disto will do this). Also, you might want to take account of the
> fact that off-horizontal compass shots are likely to be less accurate,
> perhaps with a 1/cos(I) weighting, or somesuch.

  I'm philosophicly opposed to ad-hoc weighting, since I see too much of it
  in the caving literature...  But otherwise I agree.

> > 2) Covariance of the adjuted points.  Roughly the same as the above in form,
> >    but with
> >    <sometag X="...", Y="...", Z="...", XY="...", XZ="...", YZ="...">
> 
> Except, that in this case the covariance matrix will certainly be
> different for every point.

  Correct.  (And I should have stated that.)

> L

>  ----------------------------------------------------------------
> | Lev Bishop            The Queen's College, Oxford, OX1 4AW     |
> | Physics student                                                |
> | mail: lev.bishop@queens.oxford.ac.uk (quee0367@sable.ox.ac.uk) |
> | http://users.ox.ac.uk/~quee0367/      ICQ# 21136345            |
>  ----------------------------------------------------------------
>