NAD83 to NAD83 2011 transformation methods

1077
1
08-10-2023 01:21 PM
Labels (1)
GISAdminSHN
Regular Contributor

Good afternoon and thanks in advance for your help! I have a few questions about the transformation options that show up in the list in both ArcMap and Pro.

Data Frame: projected state plane coordinate system that uses the NAD 1983 2011 datum

The Data: projected state plane "vanilla" NAD 1983

I have been researching for days trying to understand the situation inside and out in order to determine for myself the best and most appropriate transformation between the two, and I've also tested them all and compared the results with the NGS NCAT tool. Here are the questions so far, @MelitaKennedy please chime in!

1) Where can I find more detailed information about the composite coordinate frame method transformation that shows up in the top of the list?

2) Is a coordinate frame transformation inherently more accurate than a grid comparison transformation? (NADCON or NADCON5) On NGS's website they seem to stress that they are "interpolative", do you have any kind of qualitative analysis of ESRI's implementation of NADCON5 transformations versus the coordinate frame method above?

3) What NADCON transformations would you do in ArcGIS to try to match the results from NGS NCAT?

1 Reply
MelitaKennedy
Esri Notable Contributor

@GISAdminSHN wrote: -- I've added my responses in-line.

 

Data Frame: projected state plane coordinate system that uses the NAD 1983 2011 datum

The Data: projected state plane "vanilla" NAD 1983

 

1) Where can I find more detailed information about the composite coordinate frame method transformation that shows up in the top of the list?

I suspect you're seeing something like NAD_1983_To_WGS_1984 + WGS_1984_(ITRF08)_To_NAD_1983_2011. (I may not have the names 100% correct).

You can find out more about our transformations (and other objects either through the PDFs that are installed with the software and are available via the documentation. See the bottom topic on this page or this ArcGIS Pro page. You can also check out the Esri projection engine's public Github page. Select the text folder and open the appropriate object type's file. You may have to expand the results. 

In this particular case, if I got the 2 transformations correct, NAD_1983_To_WGS_1984 is a null transformation, defined when before these two coordinate systems were beginning to be refined via new realizations or re-adjustments. The two systems were assumed to be equivalent. The other transformation that includes ITRF in the name was originally based on a transformation between ITRF96 and NAD 1983 CORS96/HARN, and even though it's older, should get rid of most of the 1+ meters that now exist between ITRFxx/WGS84 and NAD 1983 (2011). Even though it was defined for a different realization. 

The best way to convert between NAD 1983, assuming you really have original, 1986 version data is to use NADCON5 transformations, like are used in NCAT. You have to have the ArcGIS Coordinate Systems Data setup installed. The official steps, which convert between all the realizations, are: 

NAD 1983 - NAD 1983 HARN - NAD 1983 FBN - NAD 1983 NSRS2007 - NAD 1983 (2011)

Complicating this is that we currently allow only 2 transformation at a time, so you can't do this in ArcMap data frame or ArcGIS Pro map/scene. 

Honestly, though, I think most people do not have original NAD 1983 data but, at best, have a mix of data that might have started out as NAD 1983 but is likely a mix of several realizations if any editing or updates occurred over the years.

2) Is a coordinate frame transformation inherently more accurate than a grid comparison transformation? (NADCON or NADCON5) On NGS's website they seem to stress that they are "interpolative", do you have any kind of qualitative analysis of ESRI's implementation of NADCON5 transformations versus the coordinate frame method above?

No. Grid/file-based transformations can better model the differences over an area. An equation-based method like coordinate frame is a least squares best fit based on the control points. Grid/file-based methods take the closest points (it's like a raster) and interpolate between them to figure out the offset for a particular location. It doesn't guarantee that's what you would get if you went out and surveyed, but that's true for equation-based methods too.

3) What NADCON transformations would you do in ArcGIS to try to match the results from NGS NCAT?

Look for NADCON5 in the name or check using the Github website where the transformation parameter for the file will include nadcon5 in the file name or for the method. 

NADCON method is NADCON 2.0/2.1 and uses a different file format. It includes original NAD83 and NAD83 HARN only. There was a set of in-between transformations released for GEOCON and GEOCON11, but those have been replaced by the NADCON5 files.

 

Melita


 

0 Kudos