Projecting NAD83 on-the-fly to WGS84

[JonathanBailey]

I have a point feature class in a geodatabase that uses WGS84 as its SRS. A colleage of mine created points in this feature class using an external database containing lat and long values. However, the lat and long values in the external database were using NAD83 (my fault -- I knew they were, I just forgot). I thought it wouldn't be a problem -- I just created a copy of the geodatabase and changed the coordinate system used by the feature class to NAD83. Should be good, right?

I added this feature class into ArcMap. I saw that the points in the copied feature class (now using NAD83) exactly line up with the original feature class (using WGS84). That doesn't seem right. If two points, having the same lat / long values, but referenced in different coordinate systems are overlaid on the same map, one of them should be projected on the fly to the coordinate system of the data frame, no?

Now, here's the weird part -- if I project the original point feature class (using WGS84) to NAD83, and add it to the map, then the points don't line up. See, this isn't what i expected, either. I expected that they would be projected to the other coordinate system (i.e., their lat / long values would be different by virtue of being projected, but they would be aligned with the original point because they were simply projected to the new coordinate system (then projected on the fly back to the original coordinate system).

Can anyone explain the behaviour that I'm seeing? And, how do I properly ensure that the NAD83 lat and long values collected by my colleague are projected to WGS84 in my target geodatabase properly?

Thanks,

Jon.

[DavidViljoen]

Check out the ESRI WKT for WGS 1984

http://spatialreference.org/ref/epsg/wgs-84/esriwkt/

and NAD83

http://spatialreference.org/ref/epsg/4269/esriwkt/

You will find them identical except for the inverse flattening.

298.257222101 (WGS84)

298.257223563 (NAD83)

As you can see, they are identical to 5 decimal places.  At mapping scales, this translates into a difference of approximately a metre but does differ depending on where you are (refer to this link:  Difference:  WGS84 and NAD83)  The difference between WGS84 and NAD83 will grow as tectonic plates move around.

If your data does not have sub-meter precision, you will not see any difference between the NAD83 and WGS84.

I hope you find this useful.

[DanPatterson_Retired]

Some supplementary materials from the help files  and a link posted there to an NRCAN website might be useful.  I am going to mention this link to see if she has any further information or suggestions on this topic.  She is the resident expert  Melita Kennedy‌

[NeilAyres]

More picky terminological issues...

You seem to be mixing up "projecting", ie changing the coordinate system from, say geographic lat/long to UTM for example.

And a datum transformation, changing the underlying basis for the coordinate system.

If you do not have survey level data I would have thought that NAD83 & WGS84 would be functionally equivalent. But there is a bunch of available transforms between the 2. But the options are many and varied. Melita can advise if you if she knows where the data is coming from.

[SachinKanaujia]

From the ArcMap perspective, everything is rendered with reference to the Data Frame CRS. You can either manually set the data frame CRS in ArcMap or it will default to the first data which is added. Then on any data having a different  GCS (Datum) will need to use a transformation method.

If you don't use any Transformation method then ArcMap will take those X/Y values and display them as is in data frames GCS because it doesn't know how to transform them.

When you are doing Projection you must be choosing a transformation method. What transformation method are you using while doing Projection? You should use the same in ArcMap to get similar results.

[JonathanBailey]

Sachin,

That seems to be the issue -- I didn't choose a transformation when adding the feature class to ArcMap -- that will teach me to click through warning dialogs.

Does anyone have any guidance on which transformation is the correct one? ArcMap indicates that it ranks the available ones in order of appropriateness -- which makes me skeptical, unless I could find some information on how it performs this ranking.

[MelitaKennedy]

Areas-of-use and accuracies, but it's easy to fool the areas-of-use check by say having a world basemap in the data frame, etc. It's not fool-proof, but works well in some cases, and is often better than the alternative: no transformation. Two relatively close GCS like NAD 1983 and WGS 1984 (current versions differ by about a meter) are actually harder because there's less difference between the various transformations.

Sometimes I've created "equivalent" transformations for ones that are originally defined for say NAD 1983 HARN and WGS 1984 (ITRF00) but using the original NAD 1983 instead. I'm less likely to do that now, because newer collected data should be defined with the appropriate GCS like NAD 1983 (2011) or NAD 1983 (CSRS), etc.

Melita

[JonathanBailey]

Hi Melita,

Thanks for your response. Now I know why it recommended a different transformation depending on whether I had a basemap or not.

The transformation that ArcMap is recommending is WGS_1984_(ITRF00)_To_NAD_1983. Currently, my feature class contains only survey-level data at the airport in Ottawa, but eventually it will contain survey level data at various airports across Canada. Would I be better off choosing one transformation to use everywhere across Canada (and would it be WGS_1984_(ITRF00)_To_NAD_1983?), or to choose a different one depending on the specific airport (i.e., maintain a different map document for editing for each airport)?

Also, my understanding is that the client uses WGS84 and NAD83 interchangeably, so I suspect that what is being reported as NAD83 may in fact be NAD83 (CSRS) or NAD83 (2011). Short of having a definitive answer on what it actually is, is there an approach that we can take to minimize the risk of assuming one over the other?

Another possibility is that, since these surveys are conducted by different surveyors (in different provinces, which regulate surveyors), is that different reference systems (NAD83, NAD83 (CSRS), NAD83 (2011)) have been used for the different surveys, but that they all ended up in the client's database as "NAD83". In the future, we can fix this (by ensuring that surveyors specify exactly which reference system is used), but are there any approaches that we could use to mitigate the fact that we don't know exactly what the reference system is for existing data?

Thanks,

Jon.

[MelitaKennedy · Accepted solution]

Hi Jon!

The only one you should have in Canada is NAD 1983 (CSRS). 2011, HARN, NSRS2007, CORS96 are all US-based re-adjustments/realizations.

WGS 1984 only exists for survey-grade data IF you have a military-grade GPS receiver. Otherwise, if the data's being RTK or post-processed, the data's being referenced to an ITRF system or to NAD 1983 (CSRS).

There's also the possibility that the surveyors or airport could have an older Canadian NAD 1983 version are linking the new data to that.

The only two transformations between CSRS and WGS84 that I think we have are:

NAD_1983_CSRS_To_WGS_1984_1

NAD_1983_CSRS_To_WGS_1984_2

The first is a 'bookkeeping' transformation that simply gets you between NAD83 CSRS and WGS84 (parameters are zeroes). The second assumes WGS 1984 = ITRF96(1997.0) so it's pretty old. I think I've seen a newer transformation, but it might have been using an algorithm that we don't support. Anyway, there's some more information here.

Melita

[JonathanBailey]

Hi Melita,

After more discussions with the client, here's what we think is happening:

1. Data are surveyed using GPS and coordinates are tied to WGS84.
2. Surveyor publishes survey data in CAD and PDF format, tied to NAD83 (Original) UTM Zone <appropriate zone for the location; -- in the case of Ottawa, Zone 18N>; not sure what transformation is used to convert from WGS84 to NAD83, as there's no metadata carried through with the data.
3. Coordinates are converted from UTM to geographic NAD83 and carried through the rest of the client's systems.
4. I'm bringing this data into a dataset that's tied to WGS84; in the absence of any additional information, I'm using NAD_1983_To_WGS_1984_1 as the transformation, which is a zero-parameter transformation.

Assuming that the surveyor used a zero-parameter transformation to go from WGS84 to UTM, the coordinates that I end up with in my WGS84 dataset should the same as those surveyed.

Does that make sense at all?

Thanks,

Jon.