"Azimuth and Bearings" in Starnet

Star*Net uses one of those values depending on which code you use to feed it your data. Each one needs to be set to be realistic for that data.

Angles are fairly obvious. Two directions could define an angle. I'd expect the standard errors to be related by the square root of half the number of directions averaged into an angle value.

But if you make a bunch of direction measurements at one setup (DB, DN, DN, ... DN, DE), the errors are correlated (particularly due to centering) and using the direction code tells Star*Net to use that correlation in its calculations.

A bearing or azimuth may be determined by various methods that have their own std errors, such as Polaris, solar, or a pair of GPS points (at what distance?). It may also be one assumed bearing to rotate everything to an old plat.

> A bearing or azimuth may be determined by various methods that have their own std errors, such as Polaris, solar, or a pair of GPS points (at what distance?). It may also be one assumed bearing to rotate everything to an old plat.

Yes, the standard error of an azimuth used as input in Star*Net would usually come into play when the azimuth of a line between two control points has actually been determined by observation. When an assumed bearing is used to orient the survey, that bearing would be fixed, error-free.

The irony in how that last bit plays out in a land survey is that it can be easier to meet the ALTA relative positional accuracy standard by using an assumed bearing basis than by using an observed azimuth since the standard error of the observed azimuth will always be larger than zero and the assumed azimuth will be considered to be free of any error in the error propagation calculations.

Kent:
I don't follow how/why you're letting an observed azimuth's standard error affect your ALTA spec analysis. Aren't you mixing accuracy of your B.O.B. with accuracy of your network which connects the boundary points? The latter being the sole subject of the ALTA spec. Will Starnet allow all bearings to be based on a non-fixed azimuth? I guess it must.
Thanks, Chris

> I don't follow how/why you're letting an observed azimuth's standard error affect your ALTA spec analysis.

The uncertainty in an azimuth from which a survey is oriented has the effect of inflating the relative positional uncertainties between points positioned by the survey, including those that are the subject of the ALTA spec.

> Aren't you mixing accuracy of your B.O.B. with accuracy of your network which connects the boundary points?

An observed azimuth is fundamentally different from an assumed azimuth when one or the other is used as a basis of bearings (to orient a survey). The observed azimuth has a quantifiable uncertainty. For example, when I used to make solar observations in the era before I had survey-grade GPS, a typical solar observation consisted of three sets on the Sun, each of which reduced to an azimuth to the reference target. The scatter of those three azimuths provided a means of estimating the standard error of the result (which turned out to be quite realistic I later discovered in going back on old work with GPS).

In the case of an assumed azimuth, one is simply designating the azimuth of the line between to certain points connected by the survey with no regard to what the actual azimuth of that line might be in relation to geodetic North or to grid North of some standard projection.

>Will Starnet allow all bearings to be based on a non-fixed azimuth? I guess it must.

Sure, of course. On larger projects, I used to observe azimuths at different points across the survey network and add all those observed azimuths, with their standard errors, to the adjustment. I later discovered when I got GPS vectors across the network to compare to the earlier conventional work that the old results were phenomenally good. Star*Net enabled that.