The Facet position is about 3ft off in Google Maps. I found a Monument in my area and it is 3 foot off.
I am using RTK2go for correction but it is 29 miles away. Also tried Polaris correction and it is 39 miles away.
I bought the Facet to help me find survey stakes on my property. I have a good platt but positions seem to off.
Any thoughts on what is causing this?
Google Maps Imagery is an OrthoMosaic and not very useful for judging GNSS accuracy.
Have a look at this guide, for more details:
CEO’s like Google Maps/Earth, very visceral, but yes a metre of offset is to be expected, you’d also perceive the Earth as being flat. You can usually select which imagery you want over the span of decades on Google Earth.
Check if you can see the antenna for the CORS / NTRIP commercial base station, to perhaps gauge how bad the registration is locally.
Survey things like parking lot corners, or fire hydrants, stuff that’s constant and immovable to establish localized registration, these can be easier to situate than survey-markers placed by the state/municipality etc. And also less distorted due to physical height.
That’s not going to improve Google Earth, nor distortions from the center-line of the satellite pass. But you’d know what you’re working with. Not sure they provide a fudge-factor adjustment, but you could keep your measurements in an Excel sheet and export fixed ones to other columns or a KML file.
Not a fan of observations from beyond the horizon. You really want data from within 6-7 km
There is North Carolina GNSS CORS 1 1/2 miles away. One time fee of $500. Its Datum is NAD83. The Facet uses WGS84, how much difference will that make.?
Basically, a GNSS device that receives RTK corrections will output coordinates in the reference frame of the Base Station that sends the corrections.
So if you use the NC State CORS network for corrections, your Facet will output NAD83 coordinates to your data collection software.
Does your Survey Plat have coordinates labeled at any of the points?
If so, they are likely State Plane Coordinates.
You will also need to decide if you have Grid or Ground Distances on your Plat.
Example: If you measure 1,000 feet across the Earth and located the endpoints with GNSS, the resulting GNSS coordinates will not be 1,000 feet apart.
This Video does a great job of explaining the concept.
The receiver will shift by whatever offset it solves via the ECEF coordinates provided by the Base. The ellipsoids being slightly different results in a distortion depending on how far you are from the notional center of the US in Kansas.
The Army Corp of Engineers has conversion/translation software, as I recollect you could push lat/lon back to ECEF via one ellipsoid, and then pull back lat/lon in the other ellipsoid. I did this for Chicago many years back, sub metre as I recall. A lot of the police dispatch systems were using NAD83 based mapping.
I’d suspect Surveyor orientated apps would be able to adjust for the mapping/coordinate systems in frequent use. It might be better to find a forum, or a local surveyor about resources, and local norms, given that some of the plat maps probably date back to the 1800’s
YES, I found NC Grid Coord. (NAD 83) on the platt.
I went to the Monument and took a measurement . Its a little over 3 feet offset
I the video explains alot , maybe to much. LOL
Why is there 3ft difference from the RTK Facet. I can understand inches not feet.
Well definitely quite a rabbit-hole, and data in a lot of formats, with different drivers to leave it alone, or revise with new methods and tools.
For the old chart plotters they used to have bit map scans of maps/charts, and you’d plug in coordinates for the corners so your position could be extrapolated onto the available imagery.
Now there’s a lot of reliance on various internet sources, and means of integration, and imagery from assorted satellites or oblique aerial photography sources.
A lot of counties/municipalities engage with services to do local oblique imagery both to maintain their own GIS/mapping, and for tax assessment and enforcement.
If I am on top of the monument should the Facet have the same position.
The NGS Data Sheet
1 National Geodetic Survey, Retrieval Date = OCTOBER 3, 2024 15:11:44 EDT
FY4170 ***********************************************************************
FY4170 DESIGNATION - D WILSON
FY4170 PID - FY4170
FY4170 STATE/COUNTY- NC/ROCKINGHAM
FY4170 COUNTRY - US
FY4170 USGS QUAD - SOUTHEAST EDEN (2019)
FY4170
FY4170 CURRENT SURVEY CONTROL
FY4170 ______________________________________________________________________
FY4170 NAD 83(2001) POSITION- 36 22 34.12733(N) 079 39 22.74017(W) ADJUSTED
FY4170* NAVD 88 ORTHO HEIGHT - 240.151 (meters) 787.90 (feet) ADJUSTED
FY4170 ______________________________________________________________________
FY4170 GEOID HEIGHT - -31.928 (meters) GEOID18
FY4170 LAPLACE CORR - 4.62 (seconds) DEFLEC18
FY4170 DYNAMIC HEIGHT - 239.945 (meters) 787.22 (feet) COMP
FY4170 MODELED GRAVITY - 979,769.7 (mgal) NAVD 88
SWmaps
In short, no. That’s why we wrote this. Different reference frames and plate shift have to be factored out.
@sparky and I both linked to the same SparkFun Guide, it’s very useful.
The only thing that I would add to the Guide is to initially select a control point that’s First Order Horizontal (h1).
I see two (h1) to your Southwest on the NGS Map Explorer
Adjust the published coordinates for Time, per the guide and get to ECEF.
They even included a link to an excel spreadsheet you can use for the final evaluation.
Note for NTRIP users: many CORS stations do not adjust their positions with respect to time but the tectonic plates keep moving every year 
To most people, that’s not a big deal - but something to consider when you’re “chasing” down accuracy.
Hmm, it could be the distance from the correction station.29 miles is quite far, especially for RTK2go, where accuracy drops with distance.Try finding a closer base or using a local NTRIP Also, double-check your Facet settings – especially the antenna.Second thing: plats might be old or outdated, so always check the accuracy of physical monuments on-site.You might want to calibrate the device right around your property too. 
There is a State Base little over a mile away
Antenna Reference Point(ARP): REIDSVILLE CORS ARP |
| ------------------------------------------------- |
| PID = DM3525 |
| |
| |
| IGS08 POSITION (EPOCH 2005.0) |
| Computed in Aug 2011 using data through gpswk 1631. |
| X = 922515.101 m latitude = 36 21 42.74193 N |
| Y = -5059060.180 m longitude = 079 39 56.55600 W |
| Z = 3760740.823 m ellipsoid height = 233.474 m
$500 to get the NC network. |
|
Uses the NGS tool Horizontal Time-Pependent Position , Worked as you said, Thanks
North Carolina has similar annual plate movement to my home State, Mississippi.
I’ve performed the Accuracy Verification from the SparkFun Guide and the results were exactly as you would expect.
When comparing the Published Coordinates from NGS for a First Order Control Point, my RTK solution closely matched when using NTRIP from a CORS…but they were wrong, as the CORS was still using it’s 10-20 year-old Base position from when it was commissioned. But my RTK Rover has no way of knowing if the Base Position is good or not (in relative terms).
When I updated the NGS Control Point’s Published Coordinates to account for plate movement (Time Dependent), then the PointPerfect RTK Corrected Solution closely matched the True Current Position, as it should. That’s because the PointPerfect terrestrial network is updated annually, unlike most CORS.
We’d expect the best results by operating our own local Base with it’s position established by a few days of observation and post processing to commission the local Base Station.
The good news is all this wont matter too much for what you are wanting to accomplish.
No matter how you operate your Facet, your RTK positions should get you close enough to recover the existing property corners. Generally speaking, it doesn’t matter if a Property Corner matches the distance in the legal description or not… the iron pin (or other monument on the plat) is the factual Property Corner, and the source of ground-truth for that location. GNSS get’s you to where it’s “supposed” to be, and then a metal detector shows you where the Corner actually is in most cases.
I have found most of the Stakes, there are around 30 . There is one run that is 1400ft through the woods up a large hill and ending in a creek. Metal detector has not found it. I have been using a program called GPSExpect to draw and connect all the platts around me. It works very well, all platts connect to each other. I can export a KML file to SWmaps. All stakes are 3-5 feet off . I can measure stake to stake with Facet, the distance is on the money. The direction I is always off a little.
The facet is very good tool, will worth the money.
