Never trust a fix on important point
All this consideration about how to measure between points in the horizontal plane (relatively) shows the the 2D mentality of most of us surveyors. No offense intended. I am one too.
Why not use a variable height rod, particularly on challenged positions. Take the first reading and then perhaps reset RTK and make it resolve the position but before taking the second reading, change the rover rod height by a foot or two.
That accomplishes the same thing with your GPS unit by giving it another point to solve for whose position is easily related to the first solution by independent means, in this case, the HI scale on the rover rod.
If you change the rod height a couple of feet, you may still get a multipath solution but you will not get one that is different by two feet in the vertical only.
Never trust a fix on important point
>
> If you change the rod height a couple of feet, you may still get a multipath solution but you will not get one that is different by two feet in the vertical only.
Which is precisely why we like to move horizontally.
The vertical shift, you mentioned above, will help.... IF the multipath is coming from horizontal branches and the like. BUT if it's coming from Large Vertical trees, well, the environment does not shift enough, to make it a very sure thing. Shift your environment, in the direction that directly changes the location of the items yielding multipath, and now you can get genuine redundancy.
Yes, I live MOSTLY on a flat earth.... 🙂
N
Never trust a fix on important point
A better saying to consider...
Never trust a single vector without an independent check. Perhaps even...
Never publish unchecked work regardless of the tools used (or using it)...
True, Nate. I have seen and done the things you said.
I have learned a few rules when using RTK. Most of all of them have been listed in the posts above.
One shot on an important point just won't be enough as mentioned above.
The best thing is to know your equipment and its limitations.
Check, and re-check.
I like the idea of a folding check device. Maybe one that has hinged legs four feet long that will unfold out to eight feet and be easily portable? It will just be something else to carry.
I usually do what others above do: take nearby shots on a nail and compare a compass/tape reading to the inverse.
OK, I'll have to admit, I am pretty much a button pusher. I bought a VRS Rover about a year ago. Have been using it on larger boundary surveys. Has been a real time saver. I check back into points when I can and check some with a TS. So far so good. I have to admit the magic pole makes me a little nervous. Am I susceptible to the same pitfalls with VRS as with RTK? Any tips on working with VRS? I understand the idea of going back and reoccupying a point a couple of hours later and watching for obstructions but if my PDOP's are low and my Horz & Vert errors are low am I good? These discussions make me wonder.
String theory
Along the lines of a check.....
What about a piece of STRING, (Nylon, or strong good stuff) that you can HOOK on the rebar, and on the prism pole, and make the distance always the same? This would allow you to do a radius check on the point of interest, and it would not be much to carry.... Maybe even a 12' gammon reel, with a painted spot at 10' or so....
You bet you are subject to the same issues. I use Network RTK exclusively. Just yesterday, I downloaded a job and imported my points into the project, I shot a cross-section of a roadway. I never rely on a single shot ( or single average of multiple epochs). One shot in this topo was off by 15' vertically, the estimated vertical precision of that point was 0.15'. Trimble Access software, in case anyone wants to know.
A deed will lie to you, a client, even another surveyor. What else is news? You own a stamp and are responsible for using it, that's just life.
Kind of like the motto: the gun isn't the weapon, I'm the weapon.
The gear isn't the surveyor, I am.
S'like anything else: if you know how to use the proper tool for the job, properly, it's almost always better than the alternative. Been setting and processing opus points and vectors for primary control (and running hard levels through them) on several jobs lately. Was interested in leica's claims, so have been going back to the opus points and running 2x5 mins with an hour spread. So far- out of roughly 10 points- the averaged coords have residuals to opus values on the order of .02x.02. I think the biggest outlier was .035 (horizontally).
Oh- ran them through the Canadian and Australian opus too just for curiosity's sake.
I have as much confidence in rtk, personally, as any other method I've employed in my 20 years of doing this. A Gunter chain with a missing link is junk too...
String theory
I like Nates idea a set string length with a ring at one end the size of the rod so you could level the string by sliding it up and down the rod . Stake point and see if the distance checks all around in a radius . Wind up and put back in pouch . With the string level you could even check elevations as well.
Agreed, bad data properly analyzed can be useful in creating better procedures.I understand that the multi-pathing issues could be different a few feet away. Much of our discussion would be mute if the software providers would concentrate on making software that surveyors want instead of wanting surveyor to adopt to what they have. I appreciate Shawn and others giving practical input to the software developers.
A lie only comes to exists when it finds a recipient willing to accept it. Until then, it's just misinformation. The world's full of it and people who don't question it.
Hammertime
> And how about the clever use of hammers to deliberatley put dings and flaws in new furniture to make trhem look rustic and antique - sells like hotcakes to pretentious bohemian-bourgeois 😉
Fremont? Boulder? Brooklyn?
Well, Nate I think you sort proved my point about the 2D mentality.
A GNSS position is, as I understand it, a three-dimensional, earth-centered position. A spherical 3D coordinate, if you will.
A multipath solution could conceivably be off only in the horizontal plane onto which you are having your controller project those positions, but only if the offending satellite were very near the horizon. But as the elevation angle of the satellite whose signal is being reflected off some surface en route to your receiver increases, the more the height is affected in our surveyor's mindset of Northings, Eastings and Orthometric heights.
In your scenario of measuring a nearby point in the horizontal plane to check to see if the measured distance with them agrees with the inverse between the points, you are in effect doing the same thing.
The method I suggested might need more like three or even four feet boost, or perhaps a meter for the purists, but it should still reveal a multipath error. You might still be getting multipath on either or both solutions, but the magnitude would not reasonably be expected to be the same. If the horizontal positions agree within your tolerance and the elevation is different by the change in the vertical offset, you have pretty good confidence that both positions are valid. If they disagree, you do not know if either of them is valid but you do know that you need to go further to validate which if either of them is a valid position.
And for the record, I have a friend here in Tennessee who routinely uses the method I have described to help eliminate false fixes.
> I have to admit the magic pole makes me a little nervous. Am I susceptible to the same pitfalls with VRS as with RTK? Any tips on working with VRS?
I've worked for PLS's that referred to GPS as "magic" always just made me want to beat my head against the wall. There's nothing magic about it...sorry just had to vent. 😉
Check out the NGS publications there's a lot of good info there.
Also check out the Trimble Knowledge Network for some really good basic GPS online training. (It's free)
LSU Center for Geomatics has a You Tube channel with some great GPS related videos.
> I know a local fellow that builds furniture.
> He doesn't own any tools except a hatchet and a hammer.
> The furniture looks like crap.
I am going to use this quote! It is great!
> I know a local fellow that builds furniture.
> He doesn't own any tools except a hatchet and a hammer.
> The furniture looks like crap.
Never tell a man holding a hatchet and a hammer that his work product looks like crap.
The method I suggested might need more like three or even four feet boost, or perhaps a meter for the purists, but it should still reveal a multipath error.
The height change needs to exceed the 19cm wavelength. That is one reason why I've never used a fixed height rod for GPS.
Here is a link to the major manufacturers of GPS equipment:
http://gpsworld.com/rtk-gnss-receivers-a-flooded-market/
If you are a land surveyor, then you are probably using one of the tech boards manufactured by one of the companies on this list.
I am not the expert on this topic. I will simply add my experience. I have used NovAtel and I am truly happy with the experience. I have tried other receivers on a trial basis only. Yes receivers do lie. I recently tried a receiver from another major manufacturer, and it lied to me on the first shot of the test. I was 5 feet off. I would say that is a lie. There is a difference between a “lie” shot and a degraded positional shot.
I would rather be 0.1-0.3 feet off (in rural, not urban) on a degraded shot than 3-30 feet off on a shot and have the so called fixed solution, or the bad fixed solution.
After the float solution, the NovAtel tech board allows for an UNVERIFIED fix. This UNVERIFIED fixed solution is rated at being acceptable 99% of the time. If conditions allow, the receiver may provide a VERIFIED fixed solution after the UNVERIFIED fixed solution. This solution is rated at 99.9% acceptable, or you will have an unacceptable position 1 shot in 1000.
On the percentages above, I personally rate the UNVERIFIED fix as 90%, and the VERIFIED fix as 99%. Even with this kind of confidence, I check and recheck, test and retest. If it is never a fixed VERIFIED shot, I definitely come back and reshoot. I dump the receiver even on every fixed VERIFIED shot.
Another expert in the field recommended this procedure. If under canopy or in other bad conditions (shiny surfaces like lakes, steel roofs, chain link fences, silos, or whatever), get your first fixed shot and walk 20-50 feet away into more open sky. Allow it to become fixed again, and walk back and reshoot your point. Repeat this procedure until you have 3 fixed shots on your point. Examine the differences. Are the differences acceptable to you? The expert then adds that the chances of that shot being wrong are zero. (Well, almost zero.)
So I think that depending upon your manufacturer of equipment, your experience will probably be different. I have also not shelved my robotic total station, because in some conditions there will be no way to get an acceptable shot.
I hope this helps.
