5

I have done a ton of research on the most accurate transformation to use in mainland Alaska to convert between the NAD83 and WGS84 datums, but I just can't seem to find a definitive answer.

ArcMap gives 3 options for my area of interest:

  1. WGS_1984_(ITRF00)_To_NAD_1983
    • This one I am really unsure of and I have not found much information on when it should be used. I know it has something to do with epochs or realizations of datums, but I am in over my head on this one. Can someone please enlighten me?
  2. NAD_1983_To_WGS_1984_5
    • This is the one I think might be most applicable, but I am unable to find any definitive information to say whether or not it should be used for Alaska.
    • Esri says it is applicable for Alaska, but I can't find confirmation of this anywhere else
    • EPSG doesn't list this transformation when I search with coordinates in Alaska (or at all for that matter)
  3. NAD_1983_To_WGS_1984_1
    • I am fairly certain that this is a "null" transformation that doesn't actually do anything, and would therefore be least accurate. Especially in Alaska where everything shifts with time.

I should also note that we are using NAD83 (NSRS 2007) in our geodatabase and converting GPS survey data from WGS84 to NAD83 (NSRS 2007) with a custom transformation developed by a geodesist for our area of interest which shifts everything by a meter or so in x and y (the Esri transformation for this is a null transformation that doesn't actually do anything). What I am trying to do is load legacy/reference data that is currently in NAD83 by first transforming to WGS84 (my question here) and then transforming to NAD83 (NSRS 2007) via our custom transformation.

I need to define a definitive procedure for the project team to follow and it will be used on potentially hundreds of layers so I need to get it right.

5

WGS 1984 has had several "releases". I'm not sure whether or not to call them re-adjustments. WGS 1984 is loosely tied to the International Terrestrial Reference Frame (ITRF), maintained by IERS.

The first transformation, WGS_1984_(ITRF00)_To_NAD_1983, assumes that WGS 1984 is the one tied to ITRF00 and the NAD 1983 realization is CORS96 or similar. That version is also known as WGS 1984 (G1150). The current realization is WGS 1984 (G1674) which is related to ITRF08 instead. This transformation uses the non-time-related parameters from the NGS CORS website. They've since published a new transformation relating ITRF08 and NAD 1983 (2011), which Esri has as WGS_1984_(ITRF08)_To_NAD_1983_2011.

NGS has released, as beta, grid file-based transformations between "HARN" and NSRS2007, GEOCON, and between NSRS2007 and NAD 1983 (2011), GEOCON11. They do include Alaska. Esri has not incorporated them yet into the software. The GEOCON/GEOCON11 file format and interpolation method differs from previous NGS file formats and interpolation methods. Their beta version only accepts Blue Book-formatted data as well.

The 2nd transformation you list assumes WGS 1984 is ITRF96, so it's quite old compared to the other two. EPSG has either deprecated it, or lists the extent as excluding Alaska, but Esri has customers there using it which is why Esri lists it as covering Alaska too.

So for now, I"m not sure what to tell you. I think I would recommend WGS_1984_(ITRF00)_To_NAD_1983 as closest to NSRS2007.

Disclosure: I work for Esri.

  • Thanks for the information. How do you suggest handling public base map type data (roads, parcels, etc.)? These almost always come in the "plain" version of NAD83 or WGS84 and I am trying to figure out if using our custom transformation or even WGS_1984_(ITRF00)_To_NAD_1983 to shift the data to NAD83 (NSRS 2007) will help or hurt me both with respect to accuracy and usability (i.e. if I transform one layer I have to transform ALL layers including background imagery for it to be useful). Since it is not survey data I'm not sure how these transformations apply. – jmpreiks Apr 2 '14 at 20:28
  • Also I understand base map data isn't usually accurate enough to notice this type of shift, but I just don't want to inadvertently introduce further error. – jmpreiks Apr 2 '14 at 20:31
2

I'm an Alaskan GIS user. Familiar with datums up here.

Roads from AKDOT, derived from 1:63,360 scale maps, treat as "NAD83" and continue to do so, accepting NONE for transforming to other forms of NAD83 in Arc when asked. Thats because +/ 40 meter data is not at same scale as modern transforms.

As for 2007, forget about it. Don't worry about that temporary Datum. NGS never liked it, and they have moved on.

You mention parcels, so thats a different bucket of data altogether, but ask BLM what the source is and go from there.

But for general, 1950's based data, assign it "NAD83" which is 1986 vintage and apply none in all data frames.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.