I am looking for some advice on the vertical accuracy of Promark 3s in consistently measuring heights to within 1cm. The surveys that I have been conducting in the last year only required a few centimeters accuracy and I found that the Promark 3s are easily capable of this. However, I am now being asked by some groundwater specialists if they could use the Promark 3s instead of a total station to obtain elevation accurate to within 1cm. This is out of my realm of experience, so I was hoping someone on this board with more experience could provide some advice.
The survey situation would likely involve setting up one Promark 3 as the base and using the other as a rover. The baseline lengths would be under a kilometer. The Thales literature states that the Promark 3s have a vertical accuracy of 1cm + 2ppm. Is this true in the real world? I suppose conducting the survey as a simple network, hopping both Promarks, would result in more reliable accuracy?
Would the new Promark 3 RTK be of potential value in trying to obtain consistent vertical accuracy of 1cm??
Thanks, Peter (not a surveyor!)
I suggest the NGS-58 manual as the place to begin.
http://www.ngs.noaa.gov/PUBS_LIB/TMNOSNGS58.pdf
Accurate GPS elevation transfer is dependent upon geoid modeling -- if the model is consistently accurate in your area of usage, then over short baselines (e.g., 1km) you can obtain reliable elevations at the 1cm level with the right equipment, procedures and observing conditions.
A few things to keep in mind:
1. Most GPS manufacturers spec their equipment at 1-sigma under good observing conditions, so the quoted vertical specification means that you'll be within 1cm 68% of the time. In other words, 1 out of every 3 times you'll exceed the specification. (Under good conditions, that is; real life is often different.)
2. The only way to reasonably validate the results of a GPS-derived elevation using only GPS equipment is to repeat the observation under different conditions: different constellation configuration and different ionospheric conditions. That means doing it twice, preferably on different days.
3. The only way to know whether or not the geoid model is consistently accurate in your area of interest is to have access to high-density validation data. (Lawton, OK comes to mind as an example.)
4. With a decent autolevel and rod, you can double-run a 1km loop in an hour or so (longer in steep country) and have a high degree of confidence that you're within that 1cm allowance. With GPS, you'll spend at least that much time setting up receivers and tracking satellites -- and remember, you'll have to do it twice, preferably on 2 different days, for each point positioned -- and not know if you've got acceptable results until you post-process the data. And if the accuracy of the geoid model is an unknown quantity in your area of interest, well, you won't know what you've got.
My advice: Steer clear of GPS-derived elevations unless you really know what you're doing.
Dear Boss,
My experience is there is some difference in Z value when you compare with Auto level.But Relative Level different is exactly matching.For accurate work prepare only auto level,you can use for contouring job,where you can control all vertical control fixed by auto level.
PKS
I use GPS for elevations. PM2s and one hour minimum on traverse points. Monday last week I expanded traverse on a survey to a considerably larger area. I put a PM2 on 1 previously GPSed point and a Z-12 and PM2 on 2 new traverse points. When my OPUS point came back it was 0.08'E, 0.02'N and 0.01'Elev off of my Solutions position. There may be project twist is from using very short start baselines. I did not reconstrain my GPS solution or otherwise correct it to meet project precision.
I did adjust my traverse elevations to my GPS elevations. I was using my new robot and am totally disastified with it's trig leveling ability. It is fantasticly accurate without autopointing, so I will have to keep that in mind in the future.
Paul in PA