Positional Tolerance - Least Squares

Because 50ppm IS 1:20,000.:-S

True , So there saying they are allowing 1 / 20000 plus .07' to any point on the job . So in a 1000' foot line .05' plus .07'= .12' or 1/8,333.33 . :-S

I am curious how the ALTA Standards came up with .07' plus 50/1,000,000.00.

> Alta 2011 standards State .07' plus 50 parts per million added to the direct inverse between control points . So am I to understand that the further you are between control points the less likely you can meet the likely statistics set in least squares, hence the need for the 50 ppm
The statistics set in least squares usually have a provision for ppms. It is reasonable to have such statistics. The ALTA spec recognizes that such provisions are reasonable. Why 50ppm instead of 30 or 40? I don't know, except that is what ALTA was willing to accept.

Your state's standards state absolute margins of error for different types of location (urban, rural, etc.) which implies a distance to the control used, I think.

> I am curious how the ALTA Standards came up with .07' plus 50/1,000,000.00.

I don't have the answer, but note that 0.07 ft. is 2 cm.

I think the .07' applies to the main control network and the 50 ppm (1/20000) is necessary allow total station work to the main control . I have run a bunch of total station work thats hits over 1/20,000 to GPS control , but the 95% error ellipses do not hit the required accuracy for my State . The bigger the project the worse . But hit the ALTA standards , because of the allowance of 50 PPM .

To hit my States requirements of .05' for AA Control surveys , I would just about have to set up on all Static control points and locate from them . While we still can report 1/20,000 which I can do all day long .05' positional tolerance does not correlate . I think we need ppm in the equation .

I also suspect the tolerances where derived from using GPS only in using positional tolerance .

Anyone ? Least squares experts .

Also we can report either 1/20000 or positional tolerance of .05' . I am just trying to correlate the two .

> Also we can report either 1/20000 or positional tolerance of .05' . I am just trying to correlate the two .
It really can't be done. They are that different.

Georgia did the same thing. The standard simply can't be met as written on a vast majority of surveys. I don't remember exactly who (a surveyor I'm sure), but someone influential in the process of accepting new regs, made the statement "Most surveyors don't understand PPMs anyway". SMH

If the requirements can't be met on a majority of surveys, you had better get your state association busy fixing it. As it stands, you would be open to anybody with a grudge filing a complaint on one of your surveys, for which you have no defense - you didn't meet standards.

>> so for a standard control survey (AA) the positional relative error between two points is .05 for the 95% error ellipse in either major or minor direction . With GPS this seems easy . . .

??? Trimble's latest R10 receivers spec at an astounding 3mm +- 0.5ppm differential static horizontal in perfect conditions. So if the measurement is 20 miles that's 0.01' + 0.05' = 0.06' and you can't meet the relative error standards.

You are measuring less than 20 miles from tight control so you're in like Flint? Sorry, I beg to differ. Dropping down to RTK with its short occupation times, less than ideal conditions, centering errors, less than perfect control, etc., blows Trimble's specs sky high. And I doubt the glowing numbers from the manufacturer is accurate outside the lab world. They are claiming they resolve the last integer to less than 1%, a difficult thing to do at any radio frequency.

Try this: Set a point with several distant points 10' to 1000' feet from it in cardinal directions in open sky. RTK observe all the points for several days and also use an invar tape or calibrated EDM to simply measure the distances. Compare; you will be surprised. Less than a hundredth every time, I doubt it. GPS is useless for in close work; a theodolite EDM combo is much more accurate.

So, does anybody care? No matter the technique, I trust GPS static or fast static observations to +- a few inches or so, if done properly; distant observations less so. When doing high rise work or high dollar boundaries, out comes the total station. It's tough to explain to the jury you measured from some satellite 11,000 miles away when the other surveyor did a simple stretch of a tape which proves otherwise.

Psomas published an article concerning close in surveying years ago which showed GPS is not an appropriate tool for boundary surveys. Unfortunately lost in the sands of time.

We set our base on the project so we are able to get good results for static control under .05'. We are usually only setting 4 control points and running conventional to everything else. . I disagree GPS has a place in boundary surveys. I not sure why you would be 20 miles from the base.

We have the option of reporting the standard 1 /20000 or the positional accuracy of .05' for control surveys. I still use the ratio of precision method so no harm . but I started looking if I could meet the positional accuracy value stated and found that for the same survey , I could not . I trying to find out why.

> I trust GPS static or fast static observations to +- a few inches or so, if done properly

My experience puts it about a magnitude better than that.

I was at the state society meeting where they recommended standards that were identical to the ALTA specs including PPMs. The board took the PPM out of the specs.

> > I trust GPS static or fast static observations to +- a few inches or so, if done properly
>
> My experience puts it about a magnitude better than that.

Upon further rumination, I agree, small site worst case +- 0.06' in my experience. I was contemplating large scale projects, > 20 miles, route survey control and County networks where the ppm aspect comes into play.

Still, 1/10 of 2 inches is only +-0.02' = 0.04', pretty tough to meet even with terrestrial equipment best practices with intervisible stations. Are you saying GPS is good enough even for the 1/4 acre boundary survey in town?

> Still, 1/10 of 2 inches is only +-0.02' = 0.04', pretty tough to meet even with terrestrial equipment best practices with intervisible stations. Are you saying GPS is good enough even for the 1/4 acre boundary survey in town?

I generally wouldn't recommend it -- on a small survey it's almost always faster to obtain acceptable results with terrestrial gear -- but with proper equipment, procedure and conditions, GPS is a centimeter-or-better technology horizontally.

I just wrapped up the control network for a USACE flood control project. One requirement was to establish 5 pairs of intervisible monuments distributed along the project extent; distances between the pair monuments varied from about 700 feet to about 1,700 feet, with most in the 1,000- to 1,200-foot range. We ran half-hour GPS sessions on each pair, measured between them with a total station, and ran Second Order Class II levels through them. We then rolled all the data into a single adjustment. The GPS horizontal residuals were about 0.005 foot N and E, and the total station distance residuals were 0.01 foot or less.

> I just wrapped up the control network for a USACE flood control project. One requirement was to establish 5 pairs of intervisible monuments distributed along the project extent; distances between the pair monuments varied from about 700 feet to about 1,700 feet, with most in the 1,000- to 1,200-foot range. We ran half-hour GPS sessions on each pair, measured between them with a total station, and ran Second Order Class II levels through them. We then rolled all the data into a single adjustment. The GPS horizontal residuals were about 0.005 foot N and E, and the total station distance residuals were 0.01 foot or less.

Impressive.