Accuracy

In the hands of the right surveyor, it is a wonderful tool worthy of accuracy of any other tool. That being said, I have seen surveys that have had huge errors because GPS was used by the ignoramus. I have also seen the same with EDM's when compared to chaining.

I have been involved with satellite navigation systems since the the late 1980's and using GNSS in some form or another for surveying since 1993. I have used GNSS on all types of projects including establishing 'A' order control, boundary surveys, topographic and construction layout. It is great but you need to pick and choose when and where you use it. The first big factor is signal obstruction, the RTK does not work reliably in areas of heavy obstruction. The signal strength is very week with the current GPS constellation. This will change in the future but does not do you any good now. The second is your required accuracy, I always figure if I can live with +/- a tenth and the site is right then RTK is a viable option. Even then I sometimes choose my robot over the RTK if the area is small and I have to tie lots of offset features and buildings. For boundary and control I always make redundant observations with varying satellite constellations.

Yes GNSS can be very precise but I would totally disagree with the statement that its precision is similar to that of a total station. Just at the manufacture specifications for RTK on the Leica GS14. Leica is claiming 10 mm + 0.5 ppm for horizontal precision and half that or 20 mm + 1 ppm for vertical. These numbers do not compare and EDM precision of 1 to 3 mm =+ 1.5 ppm. You can get better results but you can also get worse results. Just as the case with a total station, you would have a hard time stating that your accuracy exceeded the manufacture's stated precision.

Bottom line, it is a great tool if used with the correct procedure for the right job.

John

Totally agree with Mr. Putman. Well said.

What accuracy can be expected in the measurement of elevations done in static mode?

The accuracy of a single, non-networked, GPS-derived orthometric height is a combination of the accuracy of the observed ellipsoid height plus the uncertainty of the geoid model. For a static session (e.g. OPUS) a very good rule of thumb for estimating the uncertainty of the ellipsoid height at the 68% confidence level (1 sigma) is 0.037m / Sqrt T, where T is the GPS session length in hours. If you're doing RTK then you need to factor in the uncertainty in the ellipsoid height of your base station. For 95% confidence you need to double that. For the current geoid model GEOID12A NGS provides an error estimation map that is easily overlaid in DSWorld. Depending on where you are the country the 1 sigma estimate ranges from .01 to .06 m. You can often improve the accuracy of the ellipsoid height by performing a repeat session offset in time by 4 hours from the first session to get a different geometry of the satellite constellation and combine that with processing both sessions with the rapid orbits which are typically available within 17-20 hours. Additional improvements may be made by networking or localizing with respect to the local vertical control. If heights are crucial NGS would always recommend performing the redundant measurement offset by 4 hours.

B-) :good:

I am also fairly new to GPS. I have rented units in the past and have just purchased a network rover and will be working off of a VRS Network. Seems fairly simple but I would like to have a better understanding of what exactly is going on. Any tips on how to check myself and get accurate results would be appreciated. Some of the knowledge on the forum blows me away!

Some good resources:

NGS User Guidelines for Single Base Real Time GNSS Positioning

and

NGS Webinar -- Real Time Positioning - Best Methods for the Field

That's an excellent explanation.

Allow me to borrow a translation of all of that into simple terms by quoting my late colleague and friend Ellis Veatch, P.S., M.F.A., one of the great teachers of surveyors:

"It's a side shot."

We usually find NGS published monument in the vicinity of the project site and compare observed values with these published by NGS.
http://www.metzgerwillard.us/ngscs/ngscs.html
this website has pretty neat add-on that runs on google earth. It helps to visualize location of the closest NGS monument. Hope this helps.

:good:

For a single line and distance, absolutely not

as accurate!

For 2 points 2640 feet apart, maybe.

But like everything else you need to make independent observations and establish redundancy.

Ellis Veatch, blessings and peace be upon him.

So, no one wants to commit.

In practice, whether using RTN or standalone, I get measured distances between points on the ground that are within 0.05' both horizontal and vertical. This is true for points that are 10' apart or for points 5000' apart.

Please read all the equivocations, CYA statements, qualifications, and "depends" statements by others: I agree with them all.