Best practise for surveying boundaries?

Errors can be found in total station surveys as well. It is all a matter of procedures.?ÿ

I have used GNSS equipment for legal boundary surveys for years. Prove redundancy. Prove relative accuracy. Build good procedures and it will work just fine.?ÿ

With that said, I do not have any white papers to point you to for this just years of experience.?ÿ

Best practice is to find the boundaries already established for decades instead of staking out new boundaries where they never were before.

I've found plenty of errors in total station-only boundaries. If one measures a value twice (with GNSS and total station), and they don't agree, that doesn't necessarily mean that the problem measurement is always with the GNSS value.

It's not the use of GNSS, but the methodology employed. Just like a total station, the GNSS operator needs to understand the various sources of error and apply enough redundancy and blunder checks to be sure of their work.

As far as "best practices" goes, that's going to depend on what the legal minimum standards or the contract specifications are.

We do utility pole mapping with GNSS methods, and due to the way in which the PUD sets poles along the right-of-way, we don't need to get any better than half a foot or so for the final R.O.W. lines. Two or three network RTK observations does just fine for 95% of that work.

But we also do mobile laser scanning, and when we set primary control for that type work it's lengthy static sessions with redundant observations and multiple baselines between every single point. That network then gets densified using a mix of total stations and RTK.

That being said, here's a good reference that is likely applicable to your area:

https://www.smartnetna.com/documents/TSA_GuidanceNote_NetworkRTK_BestPractices.pdf

Here in the states the National Geodetic Survey has an excellent guide for real-time GNSS surveys:

https://www.ngs.noaa.gov/PUBS_LIB/NGSRealTimeUserGuidelines.v2.1.pdf

There's a draft document of a network RTK version here:

https://www.ngs.noaa.gov/PUBS_LIB/NGSGuidelinesForRealTimeGNSSNetworksV2.2.pdf

Posted by: @fobos8

Hi guys

I survey in the UK, I don't carry out legal boundary surveys, I mainly do topos.

Someone in our area has bought a GNSS rover and is going around carrying out legal boundary surveys using his GNSS only. I was chatting to a conveyancer yesterday and told him that this practice is wrong and that the boundary points need to be measured by a total station as well. This will flag up any errors in the GNSS survey. Errors have been found in these surveys.

Can anyone direct me to a paper or article outlining the best practice for surveying boundary points? This would be useful to show to other conveyancers in the area.

Cheers, Fobos

Fobos,

?ÿ

I would suggest that you dive into a bit of reading about precision, accuracy, error, etc. GNSS is simply a tool, like a steel tape, a total station, a solar compass, etc. Depending on the distance, the conditions, etc, then the correct tool is often different from day to day.

Boundary is often less about the measurements, than about the art of finding existing corners, finding relevant records, etc.

If you measure poorly, but find the original monuments, then it could be a great survey. If you measure with great precision and accuracy, but fail to find the ancient stone, then you did a poor survey.

Lastly, IMHO, you spoke with ignorance while denigrating a fellow surveyor, when you spoke to the the conveyancer. This sounds like it was inadvertent (and an honest mistake), but the standard operating procedure is that you always talk to the other surveyor first. When we skip this step, we act in arrogance, making the assumption that we cannot learn anything from the other. I know this sounds like a bit of a rebuke, but the truth is we all fall into this trap.

https://www.e-education.psu.edu/natureofgeoinfo/book/export/html/1620

https://surveyorconnect.com/community/gnss-geodesy/rtk-for-boundary-surveying/

https://www.blm.gov/policy/pim-2020-004

https://www.gpsworld.com/accuracy-precision-and-boundary-retracement-in-surveying/

Lots of guidance documents, some specific to the UK, here:

https://www.tsa-uk.org.uk/downloads/

Fobos, concerning what you told the conveyancer, I would be interested in seeing a paper or article saying the use of GNSS in boundary surveys is wrong and that a total station must be used.

I am hesitant to comment on boundary surveys in the UK because I have absolutely no experience there, but I am anyway based on my experience in several countries that derived their boundary law from English common law...

Measurement accuracy is more important for your topos than for boundaries. Surveyors like to complain about a few centimeters/ inches/hundredths of a foot,?ÿ but really it doesn't matter for boundaries, ulsess we are talking about multiple meters/tens of feet, what matters is how the survey relates to things on the ground and how the survey determines where the boundary is, and communicates that thought process.

?ÿ

?ÿ

?ÿ

GPS/GNSS is a measuring tool, the new versions are close to foolproof, but any measurement needs checks. The stated accuracy standards of the high end receivers are very close to what can be obtained with total station/robots on small boundaries. On larger boundaries, nothing can really come close.

@fobos8

Spend a little time reading about GNSS and its use in surveying prior to condemning it, you may find it superior to your current methodology. ???? ?ÿ

Posted by: @dave-karoly

Best practice is to find the boundaries already established for decades instead of staking out new boundaries where they never were before.

Some of the best surveys I have ever done didn't involve any fancy equipment at all.

A month or 2 back a certain Javad user was bragging on how his unit returned repeatable results in the ?ñ0.19 foot range, at worst.?ÿ That is acceptable precision for rural midwest US acreages but I'm guessing that would not do for most UK real estate. GPS-RTK will do much better than that in many circumstances but now and then you run into a tough spot.?ÿ?ÿ

The penalty you pay for using the Total Station when you could be using RTK is lost time. The penalty for using RTK in circumstances where the Total Station is more appropriate is lousy data.?ÿ Guess which option the regulators are going to prefer.

@mightymoe?ÿ

I beg to differ on the foolproof comment, I've had many shots where I've double tied a point in a location suitable for GNSS and found they don't agree within expected repeatiblity or they agree with themselves but not a total station shot from the same control points for the survey.

?ÿ

GNSS surveying is much more of an art than total station surveying. With total station work if you have everything set up correctly you'll get the right result, with GNSS you can do everything right and it still throws out the odd random error or a bias that isn't obvious to spot why.

It's fascinating reading the posts, thanks for your comments and links to literature.

My tutor from college was an old-timer, who always drummed it into me that for important stuff like boundaries to use plenty of redundancy, he recommended GNSS and Total Station. That's where I'm coming from.

I'm not condemning GNSS, but I certainly don't know much about it as you guys do. How do you get your redundancy using GNSS alone, just by occupying a point more than once? Surely that's the same as measuring a point twice with a Total Station.

Also, when your GNSS controller says you have accuracy of say 10mm, do you believe it? I don't, I've run GNSS on very tight traverses with braces and found differences of up to 60mm.

To be clear I'm not dissing GNSS, I'm hoping to learn something from the forum.

?ÿ

@fobos8?ÿ

The numbers on a GNSS controller screen are only RMS values which should be repeatible 68% of the time. If you double them then close to 95% will be within that range and double again 99% will, roughly. So if controller says 10mm in horizontal CQ which is not an unusual number in good conditions then I double it in my head for most work (20mm) and triple it (30mm) for tight work. Also would never rely on a single RTK shot for any control/critical point.

?ÿ

Then you move on to environmental factors. Trees/buildings blocking satellite signals and multipath from reflective surfaces near the antenna can bias a solution so you get the same wrong result twice on a point. As a graduate GNSS was the best tool in the truck but the longer I practice the more the total station comes out for urban work, rural work can often absorb the additional errors GNSS throws out.

?ÿ

The only systematic win GNSS has over total station is usually a fixed pole height for rover which helps avoid pole height errors.

?ÿ

GNSS use is a bit like boundary surveying for me, there is no substitute for good experience whereas with modern total stations and some good training to start you can be getting repeatible results pretty much straight away.

When checking repeatability of GNSS measurements, it is important to let some time elapse between so that the satellite constellation is significantly different and any multipath effects are randomized.?ÿ If you are using long baselines then you need hours or a day between to allow the iono/tropo to change.

Posted by: @fobos8

I'm not condemning GNSS, but I certainly don't know much about it as you guys do. How do you get your redundancy using GNSS alone, just by occupying a point more than once? Surely that's the same as measuring a point twice with a Total Station.

It's not just about redundancy, but independent measurements. For total station work, that means breaking setup or observing a point from a different setup/backsight. In GNSS surveying that means observing under a minimum of two different satellite configurations. Wait a minimum of 30 minutes - an hour is better - and then re-observe. If you're getting repeatable results under different constellations, you're mostly good to go. Then it becomes a matter of how tight do you need those final positions - tighter is going to require more observations, or static work.

Running double bases, as some of our offices do, is a different form of independence, but still relies upon essentially the same satellite configuration, which can sometimes burn you.

Posted by: @fobos8

Also, when your GNSS controller says you have accuracy of say 10mm, do you believe it? I don't, I've run GNSS on very tight traverses with braces and found differences of up to 60mm.

?ÿI set my job properties to 95% or 99% confidence levels. Our crews' job templates default to 95%.

I believe those numbers, in all seriousness, about 95% of the time. As others have pointed out, GNSS is a tad more fickle than a total station, but modern receivers + good practices is perfectly reliable - good enough that I'll put my stamp on it.

Now that more receivers are able to work in what used to be impossible conditions for satellite positioning, I'm definitely seeing folks push the boundaries farther than they should, while at the same time tossing out that redundancy and independent measurements. As someone mentioned upthread, GNSS has more art to it than total station work - which is why there is so much confusion about it.

@lukenz?ÿ

Of course that's why I said close to foolproof. The system that is.

The operator is a different story.

I will say that GPS systems have been the simplest instrument I've ever used. Surveying geodetically with T2 and distance meters was wwaaayyyy more difficult and involved than GPS surveying. The very first day I got my unit was an eye-opening experience. I was wary that I wouldn't understand how it all worked but it was more like, OH this is so simple.?ÿ

Even tasks such as running line, staking offsets, topo took much more thought and effort before GPS came along. Using GPS really is simple, uncomplicated,,,,,,,,, if you have the background.?ÿ

The errors I usually see and (that is rare where I survey) are usually related to calibration or someone unable to understand how the system should be used. But I've retraced townships and townships of sectional GPS surveys and found one questionable monument. That's it!! All were done with GPS and I found nothing more than that. I doubt that one was a GPS issue. And those were late 90's early 2000s GPS surveys. Not the new equipment that is so much more advanced.?ÿ

Same with local smaller boundaries, if they were GPS they are good almost universally. Doesn't mean the boundary theory is good, but the geometry is.?ÿ

It's a tool, it's simple, it's easy, it's freed me from complicated calculations, time consuming set-ups and movements, chopping lines, it's changed surveying for the better. I don't wish to do another solar, been there, done that many, many times. I don't wish to calculate areas by hand, my DC does it in the field as I survey, I don't want to spend time at lunch adjusting traverses, not needed anymore. I don't need to set up random points to run line into ravines and through timber, that's taken care of with my radio.?ÿ

And far as bad points, don't see em now, they used to be a daily occurrence, but even though we check everything it's now become a CYA (an important) task.

If you are seeing many bad fixes or large errors in your points, you must be working in way more challenging environments than I am or my equipment is more advanced.?ÿ

?ÿ

I'd rather have 60 mm precision on an accurate old boundary mark than 10 mm precision on a shiny new inaccurate boundary mark.?ÿ