Think about it like this. Where is grid in relation to ground? Below or above? How far above or below? The change in physical difference either grows or shrinks depending on the distance between any two points. The difference in two miles will be different from the difference at 10 feet.
I don't select Grid to Ground on my DC, so it sets its own SF. If I'm mixing RTK coordinates with total station coordinates, don't I want everything on the grid?
Then it makes sense to compensate for the difference in elevation between control points.
I'll leave that discussion to those who work in the mountains. None of those anywhere near me.
this thread is why LDP is the answer.....
I don't know a thing about LDP, but I found some good information about it.
I don't select Grid to Ground on my DC, so it sets its own SF. If I'm mixing RTK coordinates with total station coordinates, don't I want everything on the grid?
Depends on what you're doing and what the project requires. Recorded surveys need to show ground distances, but often our clients are working in grid (referencing in files from other sources).
But sometimes everything is in grid because we're working on a small site and the combined scale factor is so close to 1 that there is no discernible difference between grid and ground.
In that case we are technically working in grid, but also ground, and that is effectively what an LDP does - fit the grid to the ground surface so that we don't need to worry about whether we are in grid, or whether our terrestrial measurements will match our GNSS measurements.
Any decisions about what to run with in the field should be made during the post-processing or primary control phase, and after that there should be a master template with the system, projection, and scale factor already defined, and with adjusted control embedded.
If you wish to survey in State Coordinates your DC should do all those calculations for you as you collect data. This has been so since the early 1980's and the HP48 being used as the DC. Prior to that each state plane distance needed to be calculated by hand before the forward coordinate could be determined.
In fact, if you use an LDP it doesn't change anything about the process, the LDP simply shifts the surface closer to ground by changing the grid scale factor and moves bearings closer to true north by shifting the central meridian. Instead of surveying 15,000' underground with a 4degree rotation (west and east central Montana) you're now "almost" true north and "almost" along the surface of the earth. The issue with LDP's is that you need lots of them to cover much of an area.
A township size area is about perfect for an LDP, there will be about 2' of rotation near the E-W limits and depending on the elevation changes less than 10PPM for distances.
Wow - why must we make things so complicated?
Would I have Star*net output a "Ground Coordinate" file, and put those into the DC?
No! Your control coordinates should control. There's a new concept.
If you wish to survey in State Coordinates your DC should do all those calculations for you as you collect data.
All of our jobs need to use SPCs. I guess it's because we're a county public works and our GIS personnel like it that way.
SPCs are good for cataloging various projects over a large area. Their utility greatly diminishes when used for any other reasons.
Are you having trouble making your TS match up with your GNSS?
But sometimes everything is in grid because we're working on a small site and the combined scale factor is so close to 1 that there is no discernible difference between grid and ground.
Most of our job sites are less than a mile long, maybe even half a mile long. Sometimes a simple fence location adjacent to a drainage easement turns into a boundary survey if the fence encroaches on county property. It bothers me when we get a 0.10' distance error between RTK and total station on the same control points. I think about how that affects property corner location and how that error is compounded throughout the job.
I don't know what field software you are using, but many of them will allow for computing scale factors for total station setups/observations on the fly, to bring those ground observations to grid.
It bothers me when we get a 0.10' distance error between RTK and total station on the same control points.
That could easily be the difference in precision between RTK and total station observations, and might have nothing to do with projection distortion.
If two RTK points are observed, each with a nominal horizontal precision of 0.05' at one-sigma, the combined error in the inverse between them is 0.07'. Multiply by two to get 95% confidence, and you're already at 0.14 feet for that line. So it's absolutely possible to see a tenth in that backsight check when you set up a total station on those points.
(Ignoring the fact that total station setups also have random errors that produce an observed line that contains error, albeit a lot less than RTK in general.)
It's a rough calculation, but good to remember. I've seen quite a few crews burn valuable time trying to figure out why their backsight check was "bad", ignoring the fact that it was perfectly acceptable given their RTK observations.
You should see .01-.03' for any error between GNSS and TS when checking control points. Any bigger and we will be looking at the control. Even .03' might be a concern to figure out. 0.1' is unacceptable.
I checked Florida east and west zones, you will need to either be at a huge elevation or measuring a large distance to see .1'. even at the central meridian it's only .06' in 1000' near sea level, as you move east-west from the meridian it gets closer to 0' in 1000'. I would use state plane under those conditions.
@rover83 If I ever saw a 0.1' difference in RTK and a total station between control points I would immediately stop work and find out why. RTK in the open ought to be better than that. I sometimes see 0.02', rarely 0.03'. Last time I kept track on a big apartment site the average difference was 0.009'. Again, in the open between control points where I wanted accuracy. Shots on parking stripes, not so much.
Thanks for checking out the SPC zones. We use the Florida SPC east zone. We typically set control points two hundred to five hundred feet apart.
For sure, it depends on the circumstances.
We might be running base-rover RTK, dual-base RTK, network RTK (GPS+GLO only), or network RTK (full constellation) depending on the state, equipment available and project specs. Networks vary quite a bit too; station density is usually good near population centers but not so great in the boondocks, and cell latency might not be great out there either. Conditions vary from wide-open lower latitudes to heavy canopy at higher latitudes, not to mention a lot of urban areas too.
Baselines might be anywhere from a few hundred feet to 5+ miles (or 20 for NRTK). Occupation time and time of observation are big, big factors too.
We see a pretty wide variation as a result. A tenth is pretty common; so is three hundredths.
