Run it in with the level.
We generally do all of our topo's on NAD83/NAVD88. If an NGS point is not available within a reasonable distance we usually hit an NGS point with RTK. Calibrate to it and check another point. We then run all of our onsite control via level and pick up and FFE's or other necessary points. Always works well. Lets face it, control onsite generally needs to only be relative for design, unless your dealing with floodplains and such, and even then I think RTK is good. The time you spend with static work and post processing could be spent more efficiently. Static to me is for projects requiring a certain high degree of accuracy (1st or 2nd order). RTK is for everything else.
With that said, holding one point and moving the GEOID up and down is the way to go over calibrating 4 points.
We had a discussion on this subject last year on the other site, and I posted a detailed comment on March 13, 2010. I tried to import it, but all I can see is the first line of my post. If someone more tech savy than I am could bring it over here I think you would enjoy the read.
In brief, after reading some comments last year about how inaccurate GPS is, I took my unit over to the local calibrated baseline, and got the usual accurate results.
We started using a lower elevation mask as well. When post-processing, you can cut out bad data but you can't add data you don't have.
Thoughts On Vertical Control Via GPS
First off, assuming Sioux Falls, SD.
There are 3 CORS around Sioux Falls so OPUS is not a problem. OPUS requires a 2 hour minimum so your 1 hour is a definite problem. The additional CORS farther out around Sioux Falls are not favorable for OPUS-RS but it should work. ECSD CORS has a 15 second collection rate so I would use a 5 second rate on my receiver to use as much available CORS data as possible in my post processing. In fact I would interpolate ECSD to 5 second data and post process myself.
A 5° mask is useless for elevation work anyway. Even a 15-20° mask does not deteriorate GPS elevations. Collect using your defaults and work out a greater mask if necessary to clean up your post processing.
If I had only 1 L1/L2 receiver available I would put it on my most favorable location and use 2 or 3 L1 receivers on the other points.
For good Vertical GPS you want to double your occupation time for Horizontal. That would be 30 minute minimum for OPUS-RS.
"Control" means tight. In RTK you have to live with what you get, good or bad. Therefore "control" means "static". You can get close to "control" by meaning out a minimum of 3 RTK sets, which pretty much equals static time.
Paul in PA
Thoughts On Vertical Control Via GPS
Thank you for the info! Having to transfer classroom theory to what is practical is fun to do. This is up for debate, but for fairly level ground how far would you all go to bring in vertical control to a site via level run?
So much knowledge on here so thanks for sharing.
I Doubt If I Would Go 500 Feet To Bring In An Elevation
I set up GPS and let it run, while I do other work on site.
It takes a minimum of 2 to do an efficient level run. The last time I used a level I had 2 level rods with bipods since I work alone most of the time. For accuracy's sake you want to do it twice with the rods swapped. The only reason I used the level was that the vertical benchmark could not be occupied with GPS.
Paul in PA