I understand the Reliance RTK system will provide real-time 1 meter horizontal accuracy assuming good satellite coverage but intermittent poor satellite visibility. i.e.- moderate tree cover.
What is the range of vertical accuracy that we could expect in that same real time?
What if we added short occupation times-for example 30-60 seconds and then post processed the Reliance data against either 1-second CORS data from some 20 miles away or against a nearby reference station? Could we expect a significant improvement in vertical accuracy with those short occupation times and post processing?
Finally how about using a Locus RTK and a Reliance RTK for reconnaissance type topographic surveys?
Could one crew member quickly record the good visibility points using the Locus and the second man record the tougher points using 30-60 second occupation times and then both parties post process their data against the reference from the Locus?
Thanks in advance for your assistance. We are looking to do stream inventory work for NRCS.
First, we should get some terminology straight....Reliance is "Real Time Differential", meaning it uses code or (depending on your base station ability) carrier smoothed code corrections for real time. It is NOT "Real Time Kinematic", which implies that ambiguity resolution of integers for near cm accuracy is the processing method...
Because we are dealing with code, your verticals will not be particularly good. My experience in stream/river side work here in Maine is that verticals are at best 1.5 meters RMS, which is really a RADIUS. You are in the center of an "point cloud", if you will, with a scattering of dots 3 meters or so in diameter. I don't put much faith in that method for serious work.
You may get better results with a Racal style correction system OR having your own base station providing corrections via radio link. That way the broadcast corrections are more closely linked in baseline length that using some Coast Guard station 50 miles away....Ashtech has that technology for mine surveying and seismic work, and the Reliance is capable of generating base corrections for use by another Reliance as a rover unit....Check with your dealer....
As for short occupation post processed Reliance, you can get better verticals with 1) 3 to 4 minute occupations using a decimeter Reliance, with 2) relatively open canopy conditions and 3) at least 15 minutes of running time before and AFTER your field session. Post processing in decimeter mode requires mostly open conditions so that the carrier phase data can be recorded without major interruptions for use in the software. Canopy puts a crimp in those plans....and 4) that your baseline lengths between base and rover be very short, such as a mile or 2, so that the unmodeled errors are equal at both base and rover and drop out of the error budget.
If you have open conditions a Locus system in kinematic mode will provide 0.1 foot horiz and vert shots with 20 second observation times, IF you can maintain uninterrupted lock on at least 5 sats. HArd to do streamside.......and not for the faint of heart, since your real accuracies/precisions are not known until you get back to the office and process. You have to know your site conditions and sat visiblities ahead of time to assess if the sat lock conditions can be met....
Two solutions come to mind....First, consider setting control points with GPS in relatively open locations on the streams with Reliance and using a hand held laser rangefinder for bearing/distance/change in elev. to each streamside point....See an article in POB magazine about 5 years ago where Reliance style equipment was used with a laser for streambank locations......Can't remember if they did verticals or not, but the water level was fairly static.....
Another solution is an RTK system where you can use full RTK in many areas where canopy is minimal and then switch the units to RTDiff for code solutions where canopy is tougher.....Pricing versus usage is an issue...
Finally, consider setting control points and then traversing between them with either laser rangefinder/fluxgate compass or traditional Suunto/hipchain/abney level type work. That way you can put the Reliance on a point, let it chug for 10 minutes while you work the next 200 feet or so, then pick up the Reliance, move it ahead for another control point, work ahead, etc. That way you have control every few hundred feet to check in with your conventional work in between where the canopy is tough.....And it is still one man work, for the most part......bob
First..I don't know anything about the Reliance system. But you said it was sub-meter (much like the Trimble ProXRS) system...Okay...Here's a cheat..
Download and process your data (or use Omnistar, ext to get real time data)..
Then export the Lat/Long/Ellips...
Use geoid 99 (on the NGS site) to compute the elevations...You'll get much better results. I've found (at least with Trimble that their results don't always agree with the pure NGS geoid program)..
Might not be the same with Ashtech...So just try a sample point on something that you already know the answer too (a page from Phil's book)...
TM-and if this doesn't work, then at least consider the source!
Modified By Trimble Man on 2/11/2002 at 7:01 PM
You can use Reliance in real time but as Robert told you it is not RTK. Locus does not have a real time capability. For work in the trees you may do better with Reliance than with Ashtech Solutions.
Some secrets:
Time cures most defects. Work in the trees is always difficult. Your 30 to 60 second occupation times may not be enough to get the job done. It depends on your accuracy expectations.
Get a base close by. Some people call this a TCORS (temporary CORS). Make sure your base is in the best location you can find for a clear view of the sky. Your own receiver in close proximity to your rover reduces the chances for error. If you do GPS work in the trees you will have error. Don't make your base and your rover struggle to find a matching set of SV's.
I like the idea of a dual frequency receiver to use as a TCORS. That way I can set it up near my project and use OPUS or the CORS in Ashtech Solutions to get a precise location for the base. My recording interval of choice for Reliance work is 5 seconds.
Reliance (post processed) offers a choice of submeter, decimeter, and centimeter level processing. When my objective was one foot accuracy on water valves I took 3 minute shots. You can do control points for total station work with long point occupations without having to jump some of the hoops required by other systems. Still, its the same occupation time required as if you were doing a static survey. The really good news is that you can do static survey work with the same SuperCA you use for decimeter or centimeter level Reliance processing. A decimeter level Reliance system will do centimeter level static survey work when used with Ashtech Solutions. It's a good way to get those control points for your laser or total station work.
It is possible to combine Reliance with your laser work. The set up can be a little tricky and vertical accuracy is a big question. Your error budget will determine whether you can get away with it.
For work in the trees I have three suggestions; Redundancy, redundancy, redundancy.
Thank-you gentlemen for your thoughtful and useful advice.
We would be happy with decimeter level accuracy for the streambed profile, but we don't want to spend a lot of time especially when making transects across the channel. I don't even want to even think about drowning a bunch of expensive equipment.
I guess we will keep doing our homework and hopefully rent the demo before shelling out the big bucks.