Expected practice for RTK boundary surveying here is observe and then 20mins plus later re-obseve from a different or same base after re setting up (isolate set up blunders etc.). This gives you results in the field you can keep working with vs having to go back to office to process static.
For suburban work the actual repeatibilty (not the spec sheet single standard deviations) can be under 10mm which is fine but often gets into the 15-20mm which even when average the obs still appears a bit sloppy when you put the total station over the top.
I'm thinking that an alternative would be to do short static obs (5+ mins, lines all <1km) which generates a more precise measurement than RTK and at same time take an RTK obs which could be used to complete fieldwork and as a check on RTK. Once back in office static can be used to tighten up the calcs.
Setting aside environmental and setup factors, would you consider a RTK and a static measurement observed at the same to be independent measurements?
The twenty minutes was a rule of thumb to change the satellite configuration. That isnt necessary anymore. You can easily change the configuration since there are so many up. The base point should always be checked against known points.
That relies on you being able to fix with GAL/GPS or BDS/GLO only on your gear though?
Also I find change in geometry still does make a difference on moderate sites, less in wide open plains or hill tops for sure
It's simpler than that. I don't know how your unit works but all that needs to be done is turn off even sats record a measurement, then turn off odd sats, turn the even ones back on and record. This is one quick way to do it, there are others, you can play with the elevation mask, turn it up to 35 degrees and block lower sats then turn it back to 15 or whatever elevation mask you normally use. The point being that there are so many sats in the sky it's not necessary to wait 20 minutes for them to move to get different geometry. When you were lucky to have 5 up then it mattered.
What you are doing is really good, running back through the points is a good idea and of course the time it takes to locate them then return and re-observe will naturally shift the geometry, observing a fast static/ppk session will also enhance the location.
But, a RTK session is using satellite information common to both base and rover and the DC computes a solution real time, a PPK session at the same time is using the same common data to compute a solution post time. With the exception of an updated ephemeris there isn't much of a difference. I tend to trust PPK more than RTK since it takes more time to fix, but others I know have had all kinds of problems using PPK.
I often locate section corners or property corners that I really don't want to travel back to. When I'm setting and stamping a point I will begin a PPK session, locate the point after it fixes and then I have a static check along with the RTK location, you should also be able to record RTK and Static concurrently, but remember it is the same common data, a check to be sure, but not all that much of one. The good thing about the static session is you can bring in a local CORS or another base for a different check.
I'm assuming here that ppk still requires an initialisation period which then allows quicker repeat obs until you loose lock vs (fast) static which is looking at processing a specific baseline?
When your in and out of the truck for each point I wouldn't have though ppk would offer much of a time saving over static. I work on 5 mins + 2 mins per km (so 7 mins minimum)
This depends on your outlook. There are acronyms for surveying tasks that were developed decades ago and had some validity then.
PPK was a way to do post processing while also moving between points as long as you didn't lose lock. It is the same thing as static with lots of short static sessions in one file.
Static is basically the same file but you simply sit on the point and shut down as you move between points. At the end of the day each file is the same.
When you fix a PPK session the reason to do it is to be sure your first point is ready to collect. With static you let it fix in place but you can do the exact same thing with PPK. You pick the time to "fix" and the old recommended time of 8 minutes before you take a shot isn't valid anymore. I've set 3 minutes if there are over 5 satellites which there always is unless the point is under a tree. I've experimented with turning that down and never encountered any problems. 30 seconds is probably OK with lots of satellites but I'll leave that up to others to experiment with.
Static, Fast static, PPK, it's all the same. If I set on a point for 30 minutes in a PPK session or 30 minutes for a Static session there isn't any difference between the data collected.
Response based on Trimble gear and configuration:
In this situation I would use base/rover RTK with infill. Meaning both base and rover are logging static data. For troublesome points I would take two 180 epoch shots on the corner with a force break in base RTK solution between observations. Then I would take a separate 15 min (at least) static observation and process that shot independently at a later time and merge/average in TBC once back home.
But if critical, just use the gun and shoot it in.
Changing your rod height enough to make a re-observation or stake-out to the stored observation is also a good strategy to reveal a multipath solution. You would need to vary the height more than the L1 wavelength. A foot or a foot and a half should do it. You may get a different mulit-path solution, but you could almost never get the same mulit-path bust with a different antenna height. Still may not tighten up things enough to give you the comfort facto you seek, but it would reveal a bad fix pretty quick, if that even happens in these four-constellations in view days.
