For example, and I’m sure Michael here has experienced this:
Surveyor A has been contracted to provide photogrammetric control (50 targets) for a 40km^2 new space port. In his research or lack thereof, no true NAVD88 benchmarks have been found anywhere near the site. He establishes GNSS position and geoidal heights for the site project targets. When providing the data for the photo targets to the photogrammetrist, his statement is all targets tied to NSRS.
After the site has been designed by civil engineering firms based on Surveyor A data, mass clearing of the site begins … and as always all control is wiped out !
No problem. Site referenced to NSRS.
Surveyor B is hired to re-establish control for the project site. In checking for remaining control and probably other photo targets, he finds there is an approx. 0.1 meter difference in Surveyor A’s vertical control. Surveyor B has done his due diligence and low and behold, there are true NAVD88 local benchmarks surrounding the site. He ties his vertical control to the newly established control and is contracted to stake all newly designed surfaces based on the original design site. He also mentioned to the engineering firm the difference in heights for the project.
For whatever reason, no adjustment is made by the engineering firm on the design surface.
Contractor is hired for mass grading and Surveyor B notifies contractor, contractor contacts engineering firm… “no problem, we forgot to adjust, don’t worry”. Site was designed for all excavation to be balanced onsite.
Surveyor B begins layout constantly reminding of the problem.
Oh oh !!! Contractor has suddenly had stroke !!! He either has to import or export dirt (4,000,000) cubic meters of dirt !!!
Just an extreme example… of two different surfaces
Think of the methodology that was done in the early part of the 20th century to establish horizontal and vertical control. Triangulation and leveling (some of the leveling was crap done by USGS and contracted agencies, not to be confused with formerly known United States Coast and Geodetic Survey, i.e NGS).
When I have extra time, one of my hobbies is occupying old triangulation marks and NAD27 benchmarks (that’s another issue in of itself, NAVD88/NGVD27 heights). Most of triangulation stations will amaze you in the spatial differences between marks, many times < 1-2 cm !!
This is literally every big excavation project. Very rarely are with within 10% of contracted excavation/embank figures as derived from the existing contours provided at the time of estimating. The project I just mentioned above was estimated at a 110,000cy import, per the drone and the BM on the plan the engineer’s existing contours were 2 inches high and now it is a 131,000cy. Engineer A no longer works on the project and we have demobilized because the site is getting regraded.
So let me see if I got this straight, from a local stand-point, where it seems that almost all surveyors (I am obviously not a such) are using a VRS-solution based on a network of 40 stations, that are ties to ETRS89.
Using the VRS and a Geoid-model (so not just undulation), you would still run the loop to a local official monument?
Would you run more than 1?
I made a dense-cloud from photogrammetry for a client, where the primary purpose of the job was “soil-accounting” (directly translated from Danish ) over a 1 km^2 area. I had a surveyor company place 10 checkpoints on the ground, and I used one of these for my base, when laying down my GCP’s. I also checked in on a few of the Checkpionts for validation.
So here, the level-loop should be run by the Surveyor, but it would have been a good idea for me do run a level loop to a local monument, just to check the surveyors results?
I have since processed my base-position against CORS stations using ETRS89, and got the same results as the survey. I am guessing that would also suffice?
But… these are estimated orthometric heights (close to actual ground).
It may be in some areas (and have been in the past with us) this is all you can have to estimate ground values to use as vertical control. It needs to be stated for the project if used.
Here’s the quality report for the County of Lancaster SC in the USA for the 2008 LIDAR data-set, good reading. The whole state was flown for LIDAR by Dewberry Associates, a leading national firm in aerial products. We perform ground verification everyday during our surveys whether intended or not as we always carry elevations via terrestrial traverse or GNSS (either leveled NAVD88 datum or GNSS derived geoid-orthometric). Since availability, we’ve never seen more than 0.5’ variation in the published data. This was performed for 2’ ground contour intervals. Great example of what we’ve discussed.
So I did another collection on Nantucket Island Massachusetts EPSG: 2250, NAD83 and the elevation points are all showing -67 or something close. Is this because I didn’t use NAVD88 for the vertical?
I tried a quick transform with VDatum using the center of the golf course I could see on Nantucket and the elevation you provided. It gives a NAVD88 elevation of around 31’ which is aligned with topographic data I can see, contour lines of 30’ in the golf course area.
I don’t think it would since the NTRIP service you’re using probably broadcasts ellipsoidal elevations. Someone correct me if I’m wrong, but I think that’s why RV3 gives a warning when you choose another vertical datum than WGS84 to make sure your base uses the same one.
At least, it wouldn’t be too difficult to use a tool like the one I linked in the above post to transform all your GCP elevations by loading a CSV.
To transform your elevation, I used the first WGS84 (EGM2008) option for the Source but you’d have to match the model used by your correction service provider. For the Target, I used NAVD88 but didn’t check the geoid box. I think it just applied a default geoid which is the latest one, GEOID18. I think you check this box only if you want to use a specific model.