Testing of Instruments For Accuracy/Precision Statements

[Sunburned_Surveyor]

Have any of you done some basic statistical analysis of observations of your GNSS or total station instruments to prepare accuracy statements for your filed survey maps or other survey products?

If so, would you be willing to share your procedure and statement with me?

I'd like to do this for my own practice, and I'm also thinking about writing an article on the process.

Here is the type of statement I'm talking about:

This survey was conducted with GNSS and total station surveying methods. The instruments used were a Trimble R8 GNSS Receiver and a Trimble S6 Robotic Total Station. Based on the results of field tests we expect the horizontal precision of points (including line end points) shown on this survey are within 0.08 US Survey feet at the 95% confidence level. Based on the results of field tests we expect the vertical precision of points on this survey are within 0.15 US Survey feet.

Thanks.

Landon

[Steve Martin]

Hi Landon,

We always include property ties in a closed traverse, double determine them from two control points, or tie by RTK twice seperated by at least 4 hours. All of the measurements are entered into a least squares adjustment which then calculates error ellipses and relative positional tolerences. A part of that estimate is based upon your input error estimates which you should be able to refine over time by using consistent survey practices.

The statement on the map would then be "All found points were tied in a closed traverse or double determined. Results of a least squares adjustment of the measurements indicate relative positional accuracies generally on the order of +/- 0.03' or better"

[Sunburned_Surveyor]

I appreciate your response. I may have to bug you a bit more about your process later.

Landon

[Dave Karoly, PLS]

The NGS has a prescribed procedure for using an EDMI calibration base line and software for determining the errors associated with your EDMI. They also have software for processing the data gathered at the baseline.

http://www.ngs.noaa.gov/CBLINES/calibration.html

Kent McMillan posted a test to be used for determining the angular accuracy of your total station years ago on the old RPLS forum but I don't have it now.