Jump to content

 
Photo

NAD 83 vs NAD 27

- - - - -

  • Please log in to reply
7 replies to this topic

#1
Martin Gamache

Martin Gamache

    Ultimate Contributor

  • Associate Admin
  • PipPipPipPipPipPip
  • 980 posts
  • Gender:Male
  • Location:Washington DC
  • Interests:History of Topographic Cartography
    Topographic Mapping
    History of Relief Depiction
    Thematic Cartography
    Demographic Cartography
    Cartographic techniques, methods, and tools
    Orienteering
    Panoramic & Kite Photography
  • United States

I'm about to embark on a map project in Wyoming and my data is coming from a variety of agencies and includes legacy USGS and forest service products in a UTM projection and NAD 27 Datum. My DEM also from the USGS, and most of my imagery and some of the vectors are in NAD 83 (exceptions are some satellite imagery in WGS 84 and DRG mosaic in NAD27). Doing the conversions is trivial but time consumming and I am afraid may introduce some error. Some of the data has already been converted from NAD 27 to NAD 83 once (NED DEM for example) so I am reluctant to run it through NADCON once more.

Does anyone have any insight into this issue. From a user perspective does it matter much? NAD83 is more accurate.... My preference would be to use NAD 83 but since most of my base vectors come from a NAD27 source I am uncertain if I shouldn't stick to that. Are vectors more succeptible to errors in the conversion than rasters? They are certainly easier to convert (by that I mean faster). I'll be using the NADCON transformations within Manifold 6.5 to do the transformation unless someone can give me a good reason not to. I know Manifold use to be a little bit less robust in these conversions but they seem to have improved this with the last release.

Any thoughts?

#2
Rob

Rob

    Legendary Contributor

  • Moderator
  • PipPipPipPipPip
  • 418 posts
  • Gender:Male
  • Location:Kailua, Hawaii
  • Interests:anything outside.
  • United States

this statement:

"The accuracy of transformations between NAD 27 and NAD 83 (1986) are typically 12-18 cm and 5-6 cm between NAD 83 (1986) and HPGN."

from NGS here:

http://www.ngs.noaa....con/Nadcon.html

don't know about the differences between data types.

rj

#3
Pete Y.

Pete Y.

    Contributor

  • Validated Member
  • PipPip
  • 23 posts
  • Location:San Bernardino, CA
  • United States

If you'll be producing data that someone might later use in conjunction with GPS you would save them headaches by doing the transformation now. I've run into issues in which a not-so-savvy end user starts mixing the WGS84 points with NAD27 base data, not knowing anything about datums, and ends up wondering where the 100-meter "error" came from (I'm in southern Cal).

Will this project result in new data via analysis or are you simply doing cartography? It would seem to me that the error introduced by NADCON would be dwarfed by the error that is likely to be in those older data sources.

Not having anything concrete to back me up on this, I would think a vector dataset would be a "safer" transformation, b/c once the coordinates are transformed, there won't be any further interpolation as there would be on a raster dataset. The sequence of transformation-then-interpolation would seem to introduce a greater possibility of error. But that's just a semi-educated guess. Good luck!

#4
DaveB

DaveB

    Hall of Fame

  • Validated Member
  • PipPipPipPipPipPipPip
  • 1,053 posts
  • Gender:Not Telling
  • Location:Redlands, CA
  • United States

I asked our projections expert and this is what she said:
"Converting to NAD83 won't make the NAD27 data any more accurate.
Although rasters aren't quite the same (because of the resampling that must occur when projecting), the NADCON transformation is lossless. You can convert between NAD27/NAD83 multiple times without introducing more error.

While I don't want to encourage use of NAD27 for new data, if most of your existing data is NAD27, it is faster just to use NAD27."

This answer may be more specific to ArcMap. I don't know whether the same applies to Manifold.
Dave Barnes
Esri
Product Engineer
Map Geek

#5
gorillamap

gorillamap

    Contributor

  • Validated Member
  • PipPip
  • 11 posts
  • Location:Lancaster, California
  • United States

If you are going to be using the mixed datum layers together in a GIS, they'll be projected on the fly (I'm an ArcGIS guy, so I don't know if Manifold treats this differently). From what I understand, the algorithm used for this is not as solid as an actual conversion like NADCON, so you may just be better off converting everything into one datum. Although there will be some error, it shouldn't be too horrible unless you're making a map at a really large scale. Most statewide layers are generalized to the point where it probably won't matter.

As far as picking one datum over another, it really boils down to the end user's preference. NAD 83 is great, but a lot of agencies have simply never made the switch from NAD 27.
Mike Pesses
Antelope Valley College
http://avconline.avc.edu/mpesses

#6
Martin Gamache

Martin Gamache

    Ultimate Contributor

  • Associate Admin
  • PipPipPipPipPipPip
  • 980 posts
  • Gender:Male
  • Location:Washington DC
  • Interests:History of Topographic Cartography
    Topographic Mapping
    History of Relief Depiction
    Thematic Cartography
    Demographic Cartography
    Cartographic techniques, methods, and tools
    Orienteering
    Panoramic & Kite Photography
  • United States

Because this will be a print project and for performance issues, I need to pick one datum and operate in it. Reprojection on the fly is great for a few layers but not really productive when dealing with dozens of layers including large rasters, especially in Manifold. It is a case in which I need to pick one, reprojection on the fly is also easier when the datum is the same. Becasue of the resampling issue with rasters and the fact that it is only vectors that are in NAD27 it will be NAD83.

#7
A F

A F

    Newbie

  • New Member
  • Pip
  • 1 posts
  • United States

Definitely NAD83, it is more precise. S
See charts in the original NOAA NADCON publication at http://www.ngs.noaa....S_LIB/NGS50.pdf
The difference between NAD27 and NAD38 varies across US territory - west cost is most significant up to 100m.

As to concerning transformation package, if you on Linux – search Google – there many free libraries (open source).

For .NET – http://www.apsalin.c...nadcon-dll.aspx.
They have loose license terms – basically download once, use wherever you want, single dll file. But has only straight NAD27 NAD83 transformation.

For more sophisticated conversion http://www.bluemarbl...ducts/gcalc.php

#8
Melita Kennedy

Melita Kennedy

    Key Contributor

  • Validated Member
  • PipPipPip
  • 69 posts
  • Gender:Not Telling
  • Location:Redlands, CA
  • United States

If you are going to be using the mixed datum layers together in a GIS, they'll be projected on the fly (I'm an ArcGIS guy, so I don't know if Manifold treats this differently). From what I understand, the algorithm used for this is not as solid as an actual conversion like NADCON, so you may just be better off converting everything into one datum. Although there will be some error, it shouldn't be too horrible unless you're making a map at a really large scale. Most statewide layers are generalized to the point where it probably won't matter.

As far as picking one datum over another, it really boils down to the end user's preference. NAD 83 is great, but a lot of agencies have simply never made the switch from NAD 27.


Just an FYI that in ArcGIS, at least, the algorithms used to project data in ArcMap are the same the tools that make permanent datasets that are projected (Project Tool and Project Raster Tool). If you set a transformation in ArcMap and it says the method is NADCON, you're going to get the same results as using the NADCON software itself.

Melita Kennedy
ESRI product specialist




0 user(s) are reading this topic

0 members, 0 guests, 0 anonymous users

-->