Good day for gloating, let's hear how good and with what equipment!
I once ran a ts traverse. Then, a compass, and suunto compass traverse, with proper declination applied. Some 1/4 mile total trav length. Closed by 0.02'!
Dead on, gnat's butt perfect. Had to prove to Ted Dura Dura it could be done by someone other than him.
I've always found such discussions amusing because the focus is on a single determination. One leg may be out 6.02 feet but another compensates 6.01 feet resulting in a final closure calculation of 0.01.
Every measurement we make is wrong. We must learn to grasp that concept.
I started working for the agency I'm with back in 1990 or so, and I honestly saw two of the guys go out and set up on one point and side-shoot a bunch of corners in. They then put coordinates on all their sideshots, and proceeded to calculate the bearings and distances between the corners. Then they ran it through a traverse program and came out of it bragging that they had "perfect" closure... like 1/1,000,000 or some such nonsense.
One of the largest subdivisions i ever worked in was traversed with amazing closure. Something on the order of 1:350k. My first day out I found a 10 foot bust in a north south line. A week later I found a stone 35 feet from the record coordinate. No doubt one of several sideshot monuments on the job.
Up through the 80s, most work flows centered around closure. A review of most statutes shows the holdovers of this line of thinking.
As we began using equipment that lends itself better to network solutions the thinking (and software) has changed. While that can be a very good thing, traversing is by no means dead. Even in the best cases, error ellipses are exceeded by some percentage of points. The key us knowing when a check is a valid check. As time has proven closure is not the end all, even if it feels good to hit one in a bazillion..
100 acre farm with about 15 stations.
Closed with an unadjusted error of closure of 0.30 feet in a forced centre traverse.
Our methodology usually is to read six angles at each station and not to have long skinny box like figures.
Cheers,
Derek
Holy Cow, post: 407942, member: 50 wrote: Dead on, gnat's butt perfect. Had to prove to Ted Dura Dura it could be done by someone other than him.
I've always found such discussions amusing because the focus is on a single determination. One leg may be out 6.02 feet but another compensates 6.01 feet resulting in a final closure calculation of 0.01.
Every measurement we make is wrong. We must learn to grasp that concept.
Was meant as more of a fun convoy, not a serious discussion. That said, we all know figures can't lie but liars can figure.
It was early in my PC career, ran a traverse around a couple of sections using a HP48 that wouldn't let me double angles, had to close the horizon on each leg. Don't recall the exact model of Topcon TS. Did a coordinate close at the end and missed by 0.16'. Needless to say I was very pleased with myself. The surveyor I was working for was convinced I'd cheated somehow.
1 in 2 million w/ two different i-men....i've always considered it a lucky fluke w/ errors balancing out somewhere.
The traverse I am most proud of was just shy of 14mi boundary of the Homer Knowles Estate in Cass County with a shape similar to the Louisiana Purchase and it crossed Black Cypress Bayou 7 times and 287 setups with perfect angular closure and was out just less than 3ft x 3ft and one leg was a stadia shot across a wide part of the river. Circa 1975 with Dietzen transit and 100ft Lufkin Chrome Super Highway.
I didn't trust a bench mark I wanted to report a recovery on, because everything around it had changed (distance to widened street, height and distance to RR, and next to a newer signal building). I really thought it had been disturbed and put back "where it was". So ran about 0.3 mile from another mark using a topo rod, estimating hundredths between 0.05 ft marks, recording 3-wire, with a construction grade Topcon auto level. I closed 0.007 on the questionable one. I called that good enough. I was hoping to be within 0.050 ft.
I used to run big traverses, and they would sometimes close on themselves 1:300-500 thousand.
Never to NAD27 monuments all that well. Running around a township with the T2 and Topcon distance meter you wanted it tight, you didn't want to try and figure out what happened. Later I would collect each point for a sanity check with a total station and also wind angles up.
We often had traverse lengths of 50000 ft. or more, so a tight looking closure wasn't unusual. Cant remember ever having one better than 1 ppm.
I did one around two sections and closed .45', it was a along a highway east-west then heading north, crossing a river and then along a county road on the other side of the river. I never did isolate that .45', still bothers me, it was a couple of inholdings in a much larger ranch, not a big deal but it was unusual to miss that much in 5 miles, the closure was ok of course but not up to our standards, we would usually see much better results.
I would bet it was the river crossing leg that was the problem. High moisture content compared to the other shots. Then again, maybe not.
Luke J. Crawford, post: 407934, member: 11382 wrote: Good day for gloating, let's hear how good and with what equipment!
The problem with that question is that in modern practice it means nothing. In high-quality surveys, the errors in measurements will be random or pseudo-random, which means that running the same traverse several times with the identical instruments and procedures will produce different closing errors.
In modern practice, of course, survey measurements are typically adjusted by least squares and usually with conventional measurements in combination with GPS vectors as both a reality check on both conventional and GPS and as an additional condition to improve the adjusted results.
Closure errors, of course, mean very little except as a very inefficient means of flagging blunders.
The modern brag would center around the positional uncertainty of points positioned by the survey. That increasingly means the uncertainties of the various points surveyed in relation to NAD83 as realized in the local segment of the CORS network, which for all practical purposes is their absolute relation to the Real World. So, to ask the question in modern terms, it would be along the lines of: "What is the positional uncertainty (at 95% confidence) that you would ordinarily expect for points positioned by a survey greater than 1km and less than 10km in extent?"
Holy Cow, post: 407942, member: 50 wrote: Dead on, gnat's butt perfect. Had to prove to Ted Dura Dura it could be done by someone other than him.
I've always found such discussions amusing because the focus is on a single determination. One leg may be out 6.02 feet but another compensates 6.01 feet resulting in a final closure calculation of 0.01.
Every measurement we make is wrong. We must learn to grasp that concept.
Can't get any better than perfect!
I'ts funny, when things work out well, you know intellectually that you've just gotten lucky by the errors balancing each other out, but at the same time, it feels good, and you can't help but thinking that it's all skill. We all suffer from the human condition.
