Questions About the 1998 FGDC Geospatial Positioning Accuracy Standards

[Sunburned_Surveyor]

I have a couple of questions about the FGDC 1998 Geospatial Positioning Accuracy Standards. I believe, based on what I've read, that these are the standards you need to use to report California State Plane Coordinate accuracy values as required by the Public Resources Code. Here are my questions:

1) I'm trying to clarify how you calculate the network accuracy of points in a control survey in a way that allows you to determine the FGDC accuracy classification of your survey. Here is what Section 2.2.2 has to say about the method of calculating the accuracy classification:

"The survey accuracy is checked by comparing minimally constrained adjustment results against established control. The result must meet a 95 percent confidence level..."

Here is what I think that may mean:

a) Perform a minimally constrained adjustment of your survey control network. Fix the published coordinate values for one of your control points. Allow the network adjustment software to calculate the measured coordinates of the other control stations in your network with published values.
b) Calculate the residuals (2D, elevation, or 3D) between the measured and published coordinate values of the stations in your network that have a published coordinate value. (Toss out the control station that you fixed in the minimum adjustment.)
c) Calculate the standard deviation of the residuals from Step "b" at the 95% confidence level.

The result of Step "c" is the network accuracy for your survey. You can then use this value to determine the accuracy classification of your survey based on Table 2.1 of the 1998 FGDC Geospatial Positioning Accuracy Standards.

Is this interpretation of the 1998 FGDC Geospatial Positioning Accuravy Standards correct? If it isn't correct, what did I get wrong? Does anyone have an article or worksheet explaining how to determine the network accuracy of a control survey as described in the 1998 FGDC Geospatial Positioning Accuracy Standards?

Once I get this figured out, I'll try to figure out how to calculate local accuracy of points in a survey.

I'm asking some questions about this topic of our GPS vendor. I'll make sure I report his responses here.

Thanks for the help.

Landon

[Sunburned_Surveyor]

Note: The steps I outlined above result in the positional accuracy needed to classify the entire survey. I don't think this would be the network accuracy of a single control point in the survey. Let me know if you understand how to calculate the local and network accuracy of a single control point in the network according to the FGDC 1998 Geospatial Positioning Accuracy Standards.

Landon

[Steve Martin]

To me a Local Accuracy is what your least squares adjustment software gives as a result of your constrained adjustment, the accuracy relative to your fixed control. NGS uses an average of vector residuals to a station to calculate local accuracies, something a little different, more akin to relative accuracies.

Network Accuracy represents the accuracy relative to the datum.

I had asked NGS how in the world can you have local accuracies on some stations smaller than the network accuracies?

I have attached some notes from NGS on that subject.

The procedure described in your post Landon seems to me to be a procedure to validate the network accuracies of your published control stations and look for systematic errors in your measurements. The actual Network Accuracies of FBN/CBN stations are calculated by NGS as part of their definition of the datum and published on their datasheets.

In order to calculate Network Accuracies for your new stations, it seems to me that you would need to propagate the published Network Accuracy by NGS of your control stations in your adjustment in order to take in to account this base error and your measurement error in determining a accuracy relative to the datum.