The thread about handheld GPS receivers reminded me of recent work involving staking five miles of fence for construction.
The method I used was setting temporary spikes (many of which needed to be set in drill holes in rock) on approximate line, getting accurate positions on them and then offsetting from them to true line.
It was all on foot, lugging a rock drill, so light and portable were the keywords and an El Cheapo Garmin eTrex fit the bill. It was mildly surprising how close many of the preliminary spikes were to final line.
The offsets on a few lines tell the story:
[pre]
Line: 442-2069
+2.72'
-0.65'
+7.21'
-1.24'
-1.16'
-3.76'
+0.88'
+0.14'
------
8 = n
+0.52'= mean
+/-3.11'= s.e.
Line 2069-2056
-7.29'
-4.51'
-1.76'
-6.18'
+0.69'
+0.25'
+2.48'
+2.42'
------
8 = n
-1.73'= mean
+/-3.99'= std dev
Line 2082-837
-2.48'
-1.23'
-3.52'
-2.26'
-0.21'
+0.78'
-4.19'
-1.71'
-3.03'
-2.75'
-1.70'
------
11 = n
-2.02'= mean
+/-2.45'= std dev
[/pre]
So, the pooled estimate of the std dev from the three lines would be:
[pre]
SQRT { [(8 x 3.11²) + (8 x 3.99²) + (11 x 2.45²)]/(8 + 8 + 11) } = +/-3.2'
[/pre]
That means that there is about a 68% chance that any single position will be within 3.2 ft. of final line. That was actually a bit better than I was expecting.
It's always refreshing to see math that works. Of the 27 points listed, 19 (70%) are within the calculated 68% limits. It's the right distribution and the calculation is correct.
Does your etrex access GLONASS or is it an older one?
> Does your etrex access GLONASS or is it an older one?
No, it's just WAAS-enabled GPS, a couple of years old, about $100 new, a very useful tool for the price. It sounds as if I ought to see what the current generation of El Cheapo handhelds offers.
GLONASS might speed up cold starts and may help with holding a lock under more difficult conditions, but it's doubtful that it would improve accuracy. The little fellows just don't have the guts.
One interesting and easy experiment is to let the unit collect waypoints while moving on a 100-mile or so trip. It's amazing how closely the trip is mapped when such points are uploaded to Google Earth.
Did you convert the WGS84 coordinates output by the handheld GPS to NAD83(2011)? Otherwise you will have a few feet of bias right off the bat.
> One interesting and easy experiment is to let the unit collect waypoints while moving on a 100-mile or so trip. It's amazing how closely the trip is mapped when such points are uploaded to Google Earth.
I just about always know accurate NAD83 positions for boundary markers and control points. So I have a pretty good test of accuracy once the fact that the eTrex is logging ITRF positions when the datum is set to "NAD83" is taken into account.
One of the more valuable features of the handheld is the track log that allows one to follow the same route out as he took in, as well as to "map" pasture roads for navigating around a ranch.
They love open sky, which Kent probably had. Still, that seems better than I'd expect. I did some fixed location tests with hours of data over several days and found a lot of points varying 20 ft and more. Each session was smoother than that, but morning average to afternoon average might be 20 ft different.
> Did you convert the WGS84 coordinates output by the handheld GPS to NAD83(2011)? Otherwise you will have a few feet of bias right off the bat.
Yes. The informal method I used was to enter the NAD83(2011) positions of the endpoints of the line, set up the route, and then observe the apparent offset of the endpoints at the beginning and end of the line while setting the preliminary stakes. That correction was applied to get the preliminary line. That is, the prelim stakes were set to the same nominal offset from the theoretical route as the fallings on the endpoints.
WGS84 to NAD83 is maybe 3 ft. And with all of the consumer units I've checked, it doesn't matter whether you set them to WGS84 or NAD83 because they do a null conversion based on that old military report from back before anybody could measure the difference reliably.
> They love open sky, which Kent probably had.
Oh, I wouldn't go that far. As you can see, the sky was littered with clouds. :>

Can the low-end Garmins display positions to the nearest tenth of a second? My Garmin 12CX -- which I love -- no longer works without an external antenna, which makes it kind of cumbersome to use for recon, and loading search positions into my phone is a real pain. I have a few Ashtech PM2 units, but they only display lat/long to the nearest second, which limits their usefulness as a recon tool. I also have a Garmin Nuvi mounted on my dashboard, but the form factor is awkward for handheld use. I'm looking for a low-end receiver to replace the functionality of the 12CX, and wonder if the eTrex 10 or similar would fill the bill.
> Can the low-end Garmins display positions to the nearest tenth of a second?
Yes, the bottom-of-the-line eTrex (approximately a C-note to buy) I have will do that. The caveat, though, is that you have to customize the display to call up the Lat/Long component. It took me longer to figure that one out than I'd like to think it should have, but, like so many things, it's easy to do if you know how.
I did my first geocaching this summer with a waas garmin 76. Well...actually it was our 11 yr old boy who did the locating as a summer activity.
He walked up to most of the caches within a few feet. One was hidden in an old tree stump and
the lat/long placed him right on the stump. We are planning to hide a cache of two soon.
You should place a cache out in the middle of nowhere Tx. 😛
That's kinda like the GLO/BLM old way of running line. Before they started using GPS.
Have to admit I am inadequately educated in your subject matter, however there is a big 'ole mess of prickly pear jelly waitin' to happen in that picture. 😉
To see more digits in seconds, download DNR Garmin from the Minnesota DENR. Load waypoints with the Garmin cable and then view as many digits as you want. Now, they won't duplicate control point, but they will give you a somewhat better picture of the accuracy.
You can use the program to build KML files for Google Earth as well as shape files. Viewing KML files in Notepad is a great way to learn their structure.
I was able to stake out with my PM3s in autonomous mode, it would get me within 3' pretty consistently.
The older Garmin I use gets me within 60' in the woods if I'm lucky.
> That's kinda like the GLO/BLM old way of running line. Before they started using GPS.
Operationally, it was much easier for one person to do it that way instead of lugging all the stuff, rock drill, rover, bag of laths, spikes, etc. along for the hike. When you also factor in the time spent shifting an RTK base station around to get radio coverage for different parts of the site, I doubt that my method was that much more inefficient, even if I did get in my fair share of hiking.
Operationally, it was much easier for one person to do it that way instead of lugging all the stuff, rock drill, rover, bag of laths, spikes, etc. along for the hike
I've been there for sure!!
Of course, one of my larger clients told me back when he called for staking line so he could build fence; "If you don't bring your 4 wheeler, don't show up, I tired of watching surveyors walk around the ranch".;-)
> Of course, one of my larger clients told me back when he called for staking line so he could build fence; "If you don't bring your 4 wheeler, don't show up, I tired of watching surveyors walk around the ranch".
That ranch was so rocky that I'd have preferred a burro. I don't know what the breakdown rate on the ATV's is, but I'll bet one mechanical failure a mile from the truck is all it takes to turn the economics upside down. I think they are probably great for soft terrain and jeep trails, but doubt that they cross fences as easily as I can. :>