@rover83?ÿ
This is exactly what I've been looking for.?ÿ Duncan-Parnell gave me a Trimble style sheet to calculate and export positional accuracy on a point by point basis.?ÿ It's nice, but nowhere near as streamlined as your method.?ÿ?ÿ
@bill93?ÿ
Are all accuracy standards defined in terms of error ellipses nowadays? Why are PPMs included in accuracy standards? How are the number of PPMs determined for each standard?
@wal1170 The relationship for the two points in a pair would be the correction source (the base).
ppm stands for "parts-per-million". It's just another way to express a ratio, like 1:10,000. Incorporating ppm into an accuracy standard allows for the distance between two points to contribute to the allowable error. 50ppm is the same as 1:200,000 ratio. If two points are 350 feet apart, then 50ppm would allow for an additional 0.0175 foot of error above the fixed allowable error of 3cm/0.10'. So at a 95% confidence level, the error ellipse should not exceed 0.1175' on that pair of points. As Bill93 pointed out, the way the error ellipse is described as the axis instead of the semi-axis is a bit unusual.?ÿ
50ppm is the same as 1:20,000 ratio
50ppm or expressed as 1:20,000 is an impossible standard to meet. It equals <0:0025' in 50'.?ÿ
Ever set an instrument up over a CP, level up, and forget to center over the point? I have and felt terrible when I realized my mistake 20 minutes later (great backsight numbers, terrible foresight check). The point being, if 2 CPs are 400.000 feet apart, you can set anywhere as long as the instrument and backsight are 400.00 feet apart and get a good horizontal backsight residual (your only potential check would be vertical distance error for a single setup).?ÿ
If you have your crew set on CP1 and CP2, then check a similar point from each setup, I would recommend the point NOT be on line in between CP1 and CP2; they should turn at least a 90 degree angle from one of the setups. Like you said, this does not constitute a traverse, so meeting the Boards requirement on closure would be interesting but at least it provides some peace of mind.
Norman Oklahoma mentioned resection. For Board requirements, how would resections play into meeting the "ratio of precision shall not exceed an error of closure of one foot per 10,000 feet" (for urban surveys)??ÿ
Have always puzzled over a particular mentor's position on precision ratio:
"How can we prove one in five thousand if it's only 1320 feet to go around the block?"
?ÿ
@bill93 you're right. Thank you for the correction.
@nate-the-surveyor You're telling me that nobody on your Board understands elementary stats? Have they stripped all licensees from the Board? There is a lot of deregulation going on but I haven't heard anything on this level..
@mr bionic man,
It's been small town Polly ticks for a long tyme. The future may be brighter now, but it was a battle to put objective numbers on things was a future step. When plats didn't even close, or if they did, it was pretty obvious that the numbers on the plat, and the actual numbers had no relationship... It's been a rodeo here.?ÿ
N
@shawn-billings have you ever been able to create GNSS control point that is repeatable?
@wal1170?ÿ
Our crews do it daily.
If you are claiming that it's impossible to establish repeatable positions on control points using GNSS, you should probably read up on a little organization called the National Geodetic Survey, which does exactly that, and by doing so maintains the reference frame that we all operate in.
@wal1170 yes, but "repeatable" is a bit subjective, yes?
@rover83 some one has their panties bunched. If you go around a site and throw down pairs and and have no real connection other than a base, you my friend will have problems. On a large scale project those errors of the GNSS are negated. But, on a small urban setting watch out if pairs are 2cm accurate watch out.
If two points are 350 feet apart, then 50ppm would allow for an additional 0.0175 foot of error above the fixed allowable error of 3cm/0.10'.
Is this additional error allowed because error increases with distance, and the standards can be modified for it?
@field-dog yes. That's correct. Although I would submit that scalar errors are becoming very very small with GNSS, so I suspect in the future it will no longer be a consideration. We'll just be looking to not exceed some fixed absolute error tolerance relative to the datum (globe).
@wal1170?ÿ
Well, you made the statement:
"have you ever been able to create GNSS control point that is repeatable?"
...implying that control points established with GNSS are not in fact repeatable. That's incorrect.
If the workflow is poor i.e., holding absolute coordinate values from single-determined GNSS vectors for conventional control, and holding those values throughout your project, then yes, one would indeed need to "watch out".
But that's poor practice, and is not what we're talking about. We are talking about performing weighted network adjustments with error propagation, computing error ellipses, and analyzing the results to ensure they meet specs.
It is indeed possible to establish a control network - or boundary ties - that both meets relative accuracy specifications and does not connect every single control point with a conventional observation.
Have always puzzled over a particular mentor's position on precision ratio:
"How can we prove one in five thousand if it's only 1320 feet to go around the block?"
Because it's 13200 tenths or 132000 hundredths.