GNSS cadastral survey representation

[Ian Gillespie]

In a recent NZIS article the opinion was expressed that a GNSS line was measured when both end of the line were occupied at the same time.


Recently in speaking to other surveyors I was told that their procedure for GNSS cadastral surveys was to take out the joins between points measured by GNSS (traverse points and pegs points) and show these in Landonline as measured lines. This means that the completed plan has the appearance of a survey done by total station rather than the spider’s web appearance some GNSS surveys have. As I understand it you wouldn’t be able to tell by looking at such a plan that was the method used. The reason seems to be that the plan is much clearer, and it certainly would be.


This technique doesn’t seem right to me. If we take a GNSS line as having an accuracy of around 20mm (at the 95% level and including setup errors) then the error in a measured line would be +-20mm but in a calculated join it would be estimated at around 28mm. For lines under 100m between traverse marks this is getting a bit close to what the rules allow.


The SG guidelines (page 269) suggest that the reason for distinguishing between measured and calculated vectors to so that future users can see how the survey was undertaken. They also say that what constitutes a measured vector is up to the surveyor.


Does anyone have any thoughts on this matter, does it matter? Is there a consensus as to what constitutes best practice?

Example of a GPS landonline plan

[Dave Manson]

My first GPS cadastral surveys had the "spiders web" type presentation and frankly they were bloody awful things to read. I have more recently had to work over others plans that took the same approach and they were quite trying to use.

I have since applied the "what constitutes a measured vector is up to the surveyor" approach and I think my plans are better for it. They might look like the surveys were done in the old fashioned way but are easy to interpret and for those that really need to know the dataset does include the methodology used to determine each vector. I guess the other local surveyors down here will know by now that lines on my plans are not necessarily inter visible.

The vectors shown between marks on my plans are calculated from a least squares adjustment of a network that includes at least two vectors to each point from a different base station (we always use our own base). I frequently re-measure lines between traverse marks with a total station at pegging time and the closeness to the GPS derived distance gives me confidence with the method.

[Mark Geddes]

Good work Dave! I agree the spiders web plans are horrible to read and more often than not result in surveys that are hard to reproduce the measurements from i.e. lines are built out and (at times) less care seems to be taken in placing the marks. As for my current approach to this....it may a little old school, but I put text on the plan face saying if a line is not visible and where I have a mixed methods job (i.e. jigger and GPS) I put the old 'all lines by EDM unless otherwise shown' and then annotate the other lines with 'GPS'. While I initially struggled with the idea of calc'd joins being shown as measured - my compromise is to show on the plan or state in the survey report where the base station position(s) was/were so it is obvious to other surveyors what method was used to determine the brg/dist. This may effectively be similar to the content you are showing in explaining your methodology in datasets.