Datum Pairing

I have used this paper to help people understand the relationship between specific hybrid geoid models and realizations of NAD83. One thing to note is that both GEOID12B and 18 were derived from NAD83 2010.00 ellipsoid heights, but they don't always yield the same orthometric heights, depending on location. In the areas out west that I have investigated, I have found as much as 4 cm of difference. Therefore, my advice is that if you started a project using 12B, do not switch to 18 in the middle of the project (without making checks), simply because it is the "newest" and potentially more accurate. You could induce a geoid bias by doing so.

?ÿ

@spmpls

Interesting paper. Thanks for posting. Note to the OP, Geoid18 (which didn't yet exist at the time of the 2014 paper) is the current best geoid model to use with NAD83(2011).?ÿ

You can use this tool to compute the geoid height (N) at a specific location using 12B:

https://geodesy.noaa.gov/GEOID/GEOID12B/computation.html

Then do the same for 18 at the same location to see what the variation is, if any.

https://geodesy.noaa.gov/GEOID/GEOID18/computation.html

If you're just getting started with datums a note about WGS 84.?ÿ First of all be very cautious when someone says they have positional information better than about 2-3 m in WGS 84.?ÿ The reference frame is defined and maintained by the National Geospatial-Intelligence Agency (NGA) and they DO NOT provide any datum realization (passive or active control stations) that would provide high accuracy coordinates to those of us outside the Defense Department.?ÿ If somehow you do think you have relatively high accuracy coordinates then they should be referenced by the specific WGS 84 iteration they were define in (e.g., WGS 84 (original), WGS 84 (G730), WGS 84 (G873), WGS 84 (G1150), WGS 84 (1674) or the current WGS 84 (G1762).?ÿ I would highly recommend that you download the official NGS publication "WGS 84 It's Definition and Relationship with Local Geodetic Systems". Look at Table 2.1 on page 2-4 for their outline of these datum iterations.?ÿ In addition, if you quote WGS 84 positions you should also provide the epoch of those values (e.g. Jan 26, 2021 = 2021.071).?ÿ Any attempt to accurately transform WGS 84 to any other datum would require that metadata.?ÿ If have a legitimate requirement for WGS 84 for a client you can use the ITRF values provided by the NGS CORS network and then just label the data as the appropriate WGS 84 as noted above.?ÿ My long experience with many GPS users is that they often relabel NAD 83 as WGS 84 because they have been told they are the same -- THEY ARE NOT.

Wow. These two just dropped some good stuff for you. Read up.

You may also want to spend a few minutes with several YouTube videos on datums I did for the Florida GNSS Users Group a few years ago, all designed for those relatively new to geodetic surfaces and datums. -- https://www.youtube.com/channel/UCG69vYuN1Q61fWKiXffzo9A/videos

@base9geodesy Excellent stuff here. Thank you for the service you have done and for sharing all of your knowledge. These are great videos!

Posted by: @base9geodesy

My long experience with many GPS users is that they often relabel NAD 83 as WGS 84 because they have been told they are the same -- THEY ARE NOT.

That's been my experience as well. Surveyors do it more than they should, but IMO environmental/resources folks tend to do it the most.

Beware the geodetic dataset labeled as simply "NAD83" or "WGS84" with no realization or epoch tag, and no metadata concerning collection or processing methodology.

Thank you StlSurveyor - it's been an amazing ride.?ÿ I completely agree with Rover83's comment.?ÿ As the use of GPS/GNSS has exploded (and will continue even more so) and requirements for higher degrees of geospatial integrity become the norm, far too many users have received their training in the fundamentals of geodesy/datums from whomever they bought the equipment from and they seldom know much on those topics and/or how it's realized as part of the U.S. National Spatial Reference System.

@base9geodesy

That's fantastic material. I was at your datums presentation at last year's LSAW conference, and it's great to be able to have a recording for review.

Going to send this link to a bunch of my colleagues...and review it again myself.

Adding to the good information already posted,I add the following screen captures. They show the geoid-ellipsoid separation using the GEOID18, USGG2012 and EGM08 geoid models. I also include a geoid height calculation using the tool at the UNAVCO site which should not be used.?ÿ

Natural Resources Canadaƒ??s site is:

https://webapp.geod.nrcan.gc.ca/geod/tools-outils/gpsh.php?locale=en#

?ÿ

These tools yield heights (meters)of:

GEOID18 = -27.468 NGS

USGG12 ?ÿ= -27.555 NGS

EGM08 ?ÿ ?ÿ= -27.544 NRC (Canada)

UNAVCO ?ÿ= -14.371 said to be based on NGAƒ??s EGM96

?ÿ

Another hopefully useful graphic from Penn State taken from GPS for Land Surveyors:

Due to increasingly characteristic carelessness on my part, I screwed up the computation using the UNAVCO tool. I used West rather than East longitudes.

I should have realized the error. No more excuses.

I would not use the UNAVCO tool as it uses EGM96 (order 360) with the current EGM08 being order 2159 and including more and better data and modeling.

Ellipsoid-geoid separations are:

GEOID 18 ?ÿ= -27.468 for use with NAD 83 ellipsoid heights

Below are for use with ITRF ellipsoid heights

USGG 12. ?ÿ= -27.555

UNAVCO. ?ÿ = -27.780 using EGM 96

NRC. ?ÿ ?ÿ ?ÿ ?ÿ = -27.544 using EGM 08

Corrected computation

@base9geodesy,

Thank you for your reading recommendation.?ÿ I have poured over it pretty good but have a few needed clarifications if you do not mind commenting.

I understand from an academic perspective, not an actual problem, the following is true (please comment if incorrect):

1)?ÿ if I do not have orthometric heights, then I use the Geoid12b geoid to determine the height above the WGS84 ellipsoid to determine elevation.

2)?ÿ If I have orthometric heights then I should use the EGM2008 to work with WGS(1674).

3)?ÿ To work with WGS(G1762) I should use the ITRF2008.

4) Assuming the above lines 1-3 are correct, I am unclear about the older WGS84 realizations, such as G730, G873, G1150, G1674; do I just use the EGM's that fall within a close time frame of the GPS week number, such as EGM96, EGM96, EGM2008, EGM2008 respectively?

Thanks again for your time and consideration.

?ÿ

@spmpls,

Thanks for the paper, it was very helpful.?ÿ My understanding from the reading is that the following geoids can be paired for NAD83 transformations as follows:

NAD83(1996) - Geoid03

NAD83(NSRS2007) - Geoid09

NAD83(2011) - Geoid 12a

NAD83(2011) - Geoid 12b

NAD83(2011) - Geoid18

If there is anything incorrect with the above, please comment.?ÿ I will keep in mind for the 12b to 18 variations.?ÿ Is there a similar paper for NAD27 transformations?

?ÿ

@cturlington

I am a little confused about the chronology of the posting to this thread. WRT your posting, I would argue that the most recent NGS hybrid geoid model can be used with ANY NAD83 ellipsoid height to derive a NAVD88 compatible height.?ÿ

The most recent model improvements and characteristics are described in the screen capture below:

@base9geodesy

my eyes roll into the back of my head when someone says WGS84

Had one project that they kept asking for that, it turns out that what they were really working in was web mercator...they wanted to be able to relate the coordinates with what was clicked in Google Earth.

They assured me that they had a nifty converter...so I gave it to them in ITRF, carefully documented what they had and begged them to make sure to include datum and all metadata with any information transfer.

Posted by: @cturlington

Thanks for the paper, it was very helpful.?ÿ My understanding from the reading is that the following geoids can be paired for NAD83 transformations as follows:

NAD83(1996) - Geoid03

NAD83(NSRS2007) - Geoid09

NAD83(2011) - Geoid 12a

NAD83(2011) - Geoid 12b

NAD83(2011) - Geoid18

If there is anything incorrect with the above, please comment.?ÿ I will keep in mind for the 12b to 18 variations.?ÿ Is there a similar paper for NAD27 transformations?

That's the idea. Notwithstanding base9geodesy's comments, which, on matters of geodesy, will always supercede any opinion I might have.

But if, in the year 2021, you are trying to determine an elevation using, say, NAD83(2007) and Geoid09, I have to wonder what the heck you are doing that would make that a thing. Far better and more practical would be to use NAD83(2011) and Geoid18 and apply an offset as necessary to match old project values. So the question would seem to me to be academic - no practical application.?ÿ ?ÿ

@norman-oklahoma

You could be working in California. With a contract specified datum and reference frame.

I had this in 2018:

NAD83 1991.10(possibly different, I'm trying to forget about it....)