Kent McMillan, post: 383594, member: 3 wrote: So, should one assume that RTK should be used over baselines not to exceed 2.93 ft. for best results? It's understood that at very small separations, GNSS performance improves considerable as tropo noise cancels.
Typically tropospheric errors are usually estimated to be 1 ppm or 1 mm/km when the weather conditions are homogenous at the base and rover.
"The single most important guideline to remember about the weather with RT (real-time) positioning is to never perform RT in obviously different conditions from base to rover. This would include storm fronts, precipitation, temperature or atmospheric pressure."
http://www.ngs.noaa.gov/PUBS_LIB/NGSRealTimeUserGuidelines.v2.1.pdf
I do agree that it would be beneficial to conduct similar test at longer baselines and under multipath conditions and I intend to do more of this in the future.
Kent, that's quite a leap, even for you.
Shawn Billings, post: 383604, member: 6521 wrote: You may have imported that contention into my thread, but that was not the original intention of this discussion. I've been maintaining that RTK exceeds the minimum standards for many cadastral surveying needs and can do so in many instances with more efficiency than other technologies.
Well, the point of contention is what the uncertainties of RTK-derived positions are and, more specificially, the methods by which professional surveyors derive formal estimates of those uncertainties and use them to document compliance with survey accuracy standards. It should go without saying that RTK is already widely abused, and will likely continue to be in the great rush toward cheaper and faster.
matt8200, post: 383608, member: 6878 wrote: Typically tropospheric errors are usually estimated to be 1 ppm or 1 mm/km when the weather conditions are homogenous at the base and rover.
What we are discussing is tropo noise (and ionospheric noise) that is produced by structure in the atmosphere with specific scales. At very short separations between rover and base, the paths from satellite to antenna become essentially identical and noise cancels in vector solutions. This is why you expect to get exceptionally quiet vector solutions at such short separations and why tests conducted on extremely short baselines would not be a very good indicator of much of anything encountered in actual practice over significantly greater baseline distances.
I've been surveying for 48 year (come September), and have worked in over a dozen states (nearly all in the West, but also Alaska and Maine). Over those years I have noticed that:
ALL of the BEST surveys that I have encountered, were performed using a Compass and Chain, or Transit and Tape, or Theodolite and EDMI, or a Total Station, or Static GPS, OR [dare I say it] RTK.
And, ALL of the WORST, total POS, incompetent, fraudulent, FUBAR'd surveys were TOO!.
It may very well be that RTK is the most misunderstood, misused, and/or abused advancement in Survey technology since the introduction of the magnetic compass, BUT blaming the TOOL for the results of its misuse/abuse by an operator, is disingenuous at best. Maybe we should cut to the chase, and blame the microchip!
From what I've seen, there are [unfortunately] still a few "surveyors" out there that can't properly use ANY of the above mentioned "tools."
😉
Loyal
Loyal, post: 383612, member: 228 wrote: It may very well be that RTK is the most misunderstood, misused, and/or abused advancement in Survey technology since the introduction of the magnetic compass, BUT blaming the TOOL for the results of its misuse/abuse by an operator, is disingenuous at best.
Aren't you forgetting how RTK has been marketed, though? RTK was originally marketed (in Texas, at least) as a quick way of deriving coordinates in some arbitrary system obtained via "calibration" (transformation to some arbitrary system). The appeal of that was that the user could just blow and go as they had blown and gone before.
Now, with NetworkRTK cranking out coordinates on standard projections like the various zones of the Texas Coordinates System of 1983, it isn't uncommon to see quickie surveys that are nominally oriented to grid North, but with spastic blowouts from pushing RTK where it shouldn't go. Selling RTK on the basis of speed does tend to self-select users and promote quickie ways of using it that defeat reliability and compliance with professional standards.
Kent McMillan, post: 383616, member: 3 wrote: RTK was originally marketed (in Texas, at least) as a quick way of deriving coordinates in some arbitrary system obtained via "calibration" (transformation to some arbitrary system).
Unfortunately this was all too true. I posted earlier in this thread about people who only wanted to know what buttons to push to get results; there were plenty of Brand X salesmen out there who were more than willing to accommodate them. I worked for Leica and for a large, reputable Leica dealer from 1993 - 2004, and that was definitely NOT the way we operated.
Site Calibrations, or Transformations, or Localizations, or whatever your software of choice calls them can be very dangerous, especially when created by someone lacking the proper training and/or understanding of how they work.
IMHO, many vendors pushed that approach because they had software that was developed for flat earth, plain surveying and had to be adapted to GPS. Leica and Trimble both came at GPS from a geodetic perspective, and we (at least at Leica) trained our users on that aspect of it.
Kent McMillan, post: 383616, member: 3 wrote: Aren't you forgetting how RTK has been marketed, though? RTK was originally marketed (in Texas, at least) as a quick way of deriving coordinates in some arbitrary system obtained via "calibration" (transformation to some arbitrary system). The appeal of that was that the user could just blow and go as they had blown and gone before.
Aren't YOU forgetting that I have ranted about "calibrations," the modifricking of UTM/SPC, and the widespread misuse of these tools for years?
Just because a car salesman tells you that Cruise Control makes it so "all YOU have to do is steer," doesn't make the Cruise Control liable for rolling the car in tight curve!
Loyal
Lee D, post: 383623, member: 7971 wrote: Site Calibrations, or Transformations, or Localizations, or whatever your software of choice calls them can be very dangerous, especially when created by someone lacking the proper training and/or understanding of how they work.
This^^
RTK has pushed the accuracy of surveys way up; there has been a C change. Gone are the days of following sectional surveys and seen any material differences.
I recently reviewed surveys through 40 different sections, there wasn't anything between the first surveyors plats and the guy I was reviewing who was relocating. He sent me raw data and it was as close to perfect as you can ask for.
The first surveyors used RTK as did the next surveyor. This couldn't be done any other way.
None of this was done by me, the second surveyor was using R8's, don't know what the other surveyors used, but the surveys dated back to the late 1990's.
No way, even using T2's and a good distance meter, could you replicate the accuracy I was seeing, only with RTK; static would get you fired (and rightly so).
Regarding the accuracy of RTK... sure, a static measurement is more precise, and at short distances a total station is even more precise. For that matter, at very short distances, a steel tape and a level are even more precise. However, of those methods, only static is going to give me a position that's anywhere near the precision of RTK at any kind of distance, and running static on every point wouldn't be terribly efficient. And sometimes (actually other than when setting primary control for a network almost all the time) 0.05' is plenty good enough.
No way, even using T2's and a good distance meter, could you replicate the accuracy I was seeing, only with RTK; static would get you fired (and rightly so).
Could you explain to me why someone would get fired for using Static?
Lee D, post: 383637, member: 7971 wrote: Regarding the accuracy of RTK... sure, a static measurement is more precise, and at short distances a total station is even more precise. For that matter, at very short distances, a steel tape and a level are even more precise. However, of those methods, only static is going to give me a position that's anywhere near the precision of RTK at any kind of distance, and running static on every point wouldn't be terribly efficient. And sometimes (actually other than when setting primary control for a network almost all the time) 0.05' is plenty good enough.
The point, though, remains that there is virtually no situation in which RTK will give better positional accuracy than some other surveying method will. The reason for using RTK is NOT that it is the most accurate, but that it is quicker, if dirtier, than the other higher accuracy methods.
The apparent reality is that in the race to provide the cheapest and fastest surveys, speed is more popular than accuracy or reliability. The challenge still stands to describe any land survey situation in which RTK gives a more accurate answer than some other method will.
Kent McMillan, post: 383644, member: 3 wrote: The point, though, remains that there is virtually no situation in which RTK will give better positional accuracy than some other surveying method will. The reason for using RTK is NOT that it is the most accurate, but that it is quicker, if dirtier, than the other higher accuracy methods.
The apparent reality is that in the race to provide the cheapest and fastest surveys, speed is more popular than accuracy or reliability. The challenge still stands to describe any land survey situation in which RTK gives a more accurate answer than some other method will.
This begs the question as to why you use a low order accuracy total station. Why not a more accurate total station?
Scott Ellis, post: 383639, member: 7154 wrote: Could you explain to me why someone would get fired for using Static?
I dunno about his thinking, but mine is that with being so UNcost-effective, that it would be difficult to keep a business running at a competitive price while taking a 4 hour observation on each point.
Scott Ellis, post: 383639, member: 7154 wrote: No way, even using T2's and a good distance meter, could you replicate the accuracy I was seeing, only with RTK; static would get you fired (and rightly so).
Could you explain to me why someone would get fired for using Static?
It would take too long, cost too much, these companies aren't neophytes, they understand what it costs to survey, and if they see something way out of line they will quickly get rid of you.
Imagine 260 section corner spread through 5 townships and you are going to static them?
PPK maybe as a check, ok,
But static?
How long of an occupation would we be talking about, nearest CORS is 90 miles.
For what benefit?
RTK is far more accurate than a total station if you have to traverse any distance at all. RTK is far more accurate than a total station if you can't avoid short backsights. RTK is far more accurate than a total station if a traverse is down a highly linear corridor. And yes speed does matter... we're in business to make money.
MightyMoe, post: 383656, member: 700 wrote: Could you explain to me why someone would get fired for using Static?
It would take too long, cost too much, these companies aren't neophytes, they understand what it costs to survey, and if they see something way out of line they will quickly get rid of you.
Imagine 260 section corner spread through 5 townships and you are going to static them?
PPK maybe as a check, ok,
But static?
How long of an occupation would we be talking about, nearest CORS is 90 miles.
For what benefit?
If I feel I need to setup and run a static on a point I will, I also do not let my clients set the price of my work. The benefit would be knowing my work is correct.
MightyMoe, post: 383656, member: 700 wrote: Could you explain to me why someone would get fired for using Static?
It would take too long, cost too much, these companies aren't neophytes, they understand what it costs to survey, and if they see something way out of line they will quickly get rid of you.
Imagine 260 section corner spread through 5 townships and you are going to static them?
PPK maybe as a check, ok,
But static?
How long of an occupation would we be talking about, nearest CORS is 90 miles.
For what benefit?
If I feel I need to setup and run a static on a point I will, I also do not let my clients set the price of my work. The benefit would be knowing my work is correct.
Shawn Billings, post: 383645, member: 6521 wrote: This begs the question as to why you use a low order accuracy total station. Why not a more accurate total station?
A total station that will measure angles with a standard error of 1.5" and distances with a standard error of 2mm turns out to produce excellent results when combined with GPS vectors. It uses both for what they are best at. Since the limiting factor is usually the GPS-derived coordinates, not the conventional measurements, any improvement in the conventional measurements wouldn't result in a similar improvement in the results.
By "combine", of course, what is meant is adjust them by least squares in combination with those other categories of measurements with different uncertainties and you have to have realistic estimates of the uncertainties in all of the measurements combined.
IF a surveyor knew what the uncertainties in the RTK vectors he was logging were, he could combine them with conventional measurements with a total station and/or levelling observations and also meet the same accuracy standards that a professional surveyor would want to meet, rather than just settling for whatever came out of the box. However, it would take knowing more about the uncertainties in the measurements than the typical RTK user would want to know.
