I have used both a base station and the state CORS network. The crews I was working with while using the base we were only good to about 6 or so miles on a good day. I know the height of the antenna is big factor. While using the base the typical checks were practically the same as using the statewide CORS network for a base and those baselines extend upwards of 10-12 miles at times. How far is too far of a GPS baseline in your opinion? I'm sure the best thing to do in this type of deal is to run a static session on a point and bring the base closer to the job, but in the event there is no base to use and you rely on the CORS network, how far would you push it, if for recon and nothing else?
Rather than anecdotes, try some of the literature on the subject. I like "Accuracy of GPS-derived relative positions as a function of interstation distances and observation-session duration," by Mark Eckl, et al which can be viewed/downloaded from here: http://geodesy.noaa.gov/CORS/Articles/Accuracy.pdf
You did not mention the equipment nor the software you use. With a good quality dual frequency GNSS receiver rather than a limit on the length of your baseline you need to insure that you observe a long enough session to meet your accuracy standard.
It is interesting to submit data to the various automated GPS-processing tools. Some of the point-positioning tools graphically depict the time to convergence. OPUS and OPUS-RS are not the only tools available. Be aware that most of the others do not provide results in SPC and current versions of NAD83. Transformations to this systems are easily accomplished with NGS toolkit items.
HTH,
DMM
For best production with my PM3s the baselines are not over 5mi.
The 6 mile figure you state is no doubt the radio link range of your transmitter. Get a stronger radio, or put the antenna higher, or use a cellular modem, and go further. It's usually enough for most uses.
The limit of how far you can go with an RTK baseline is generally taken to be 20 km (12 miles). The fundamental condition is that the base and rover be under the same upper atmosphere conditions, so that those conditions don't have to be modelled, they simply cancel out. Of course that 20 km is not a hard and fast limit, it's just a rule of thumb.
The correction of the base station degrades in the height by about 1ppm/km.
Sent from my ZTE T83 using Tapatalk
In "GPS For Land Surveyors," Jan Van Sickle says: " Single-frequency receivers are best applied to relatively short baselines, say, under 25 km....."Dual-frequency receivers, on the other hand, have the capability to nearly eliminate the effects of ionospheric refraction, and can handle longer baselines."
I believe that statement applies to static baselines. Static baselines can be hundreds of kilometers long. My comments (that is/ my alter ego's comments) applied to RTK and Rapid Static vectors.
Do any of you have any sort of distance to baseline occupation time standard for you or your crews? Such as, X occupation time for 5 miles, Y occupation time for 10 miles etc.?
Assuming we are talking about RTK vectors, you will generally find that longer baselines take longer to initialize. But once you are initialized the occupation times don't need to vary with baseline length much, if at all, within the typical radio range limit of your vector. For quality vectors it is much more important to pay attention to PDOP values and multipath conditons than occupation times.
There is precious little documented study on RTK occupation times but there are a lot of personal opinions. Mine is that there isn't much to be gained by going over 3 minutes. In fact, if you have good PDOP and limited multipath the "curve" (of the time vs. precision graph) really flattens out after maybe 15- 30 seconds. If you don't have good PDOP and limited multipath it really never flattens out.
Not sure about any others, but Trimble RTK systems converge the solution for the duration of the measurement. They also check each individual epoch to make sure it is within tolerance. When the dual criteria for time and number of epochs are meant the solution can be stored manually or automatically. Their recommendation for control points is three minutes; this is long enough to ensure that you don't have incorrect ambiguities.
The specification on most RTK systems is 1 cm + 1 ppm horizontal, and 2 - 3 times that vertically. 1 ppm is 1 cm at 10 Km, or about 6.2 miles. So up to about that distance the ppm error should be within the noise of the solution. The ppm error will be much more significant if the atmospheric conditions are different at the base and rover or if there is a high level of solar activity. In good conditions I've seen good RTK results at as much as 30 miles but I certainly don't recommend doing that; I've also seen results at that distance that were over half a foot off.
As far as running on a network, it depends on if you're using a single base solution or a VRS (network) solution. In a true VRS your "base" is only a few feet away from your initial position.