Spatial reference does not match data frame

[RobertStevens]

Can someone answer this question. It is not peculiar to any version of Arcmap desktop: I have seen it in 10.0, 10.1,10.2,10.3 and 10.4.

When I "start editing" from the editor toolbar I get a dialogue box asking which layer or workspace I wish to edit.

After I choose one I get another warning dialogue box which lists all (well, all, as far as I can tell) the layers in the frame and where each one has a message "Spatial reference does not match data frame". There are many layers in my frames because I am using Business Analyst basemaps which have multiple datasets which display at different scale values.

Why do I get this message? Both the data frame and the layers are using the same geographic coordinate system (WGS_84), although the former is projecting the data using the World Mercator projection.

The warnings don't prevent my being able to edit the data: I am only editing fields which are not geographic. But the message seems to imply that if I were digitizing new features that I would be met with failure.

[NeilAyres]

It's just a warning. Telling you that it is probably wiser to edit the spatial data in the same coordinate system as the data itself.

[DanPatterson_Retired]

This will become important if you decide to create data because you may end up with a file in Web Mercator instead of GCS WGS 84, which I presume, is the coordinate system that you would prefer to keep all you data in

[RobertStevens]

Ty for responses Neil and Dan.

Dan, are you saying that if I created new features (I am not BTW), that

those new features would change the coordinate system of

the underlying dataset?

Ie: I have features with GCS WGS 84 coordinates. I have a map

with world mercator projection. I create a new feature, and

the coordinates are stored in meters in the original dataset?

So I have umpteen features stored as a pair of angles, and

one feature stored with a pair of huge numbers?

Or is it rather that the new feature will cause all the values

in the original dataset to be converted to meters, and

the coordinate system description will become Web Mercator

(which I wouldn't want since I had a reason for having the

original dataset in CGS WGS 84)

[DanPatterson_Retired]

Nothing changes the coordinate system of something that has been started and the edited again.  I said new features as in a new feature class or shapefile.  I have seen way too many times that the difference between the data frame and desired coordinate system has led to hours of confusion.  Just be aware of the rules and the safest rule is never to mix coordinate systems.  I always go the extra mile and standardize the data to one coordinate system.  This unfortunately doesn't work for people forced to work with web-based basemaps and the like.  They have to be more vigilant.

[NeilAyres]

What is your reason for having the data frame in Web Mercator? Because it's only purpose, as far as I can see, is to be able to display web maps of the world. So it is not very good at scale or area (or anything else projection related). So your decision the have your base data in GCS_WGS84 is fine. Nothing will happen to that, even if your df is set to Web Mercator. The data will still be captured in the coord sys of the source. As Dan says, this can cause confusion, if a user starts edit a new dataset but is not aware of the coordinate system that they are looking at.

And a picky point to Rick...

Data is never IN WGS84. Because WGS84, by itself, is not a coordinate system. It is a datum, geoid & ellipsoid, but not a coordinate system.

GCS_WGS84 is a coordinate system, geographic coordinates based on the WGS84 datum.

Could have many multitudes of other coordinate systems, UTM, State plane, zillions of others, all based on the underlying WGS84 datum.

I have lots of experience with data based on various "classical" datums. Then you need a "transformation" between the two.

Remember, it's ON not IN.

[RickGeittmann]

Neil - good point on the WGS84 I should have been specific and not used a general catch all for terminology and confuse folks.

We use the Web Mercator in the data frames for any data that is used in online mapping applications and transform the data to Web Mercator into a DB set up to serve our online applications to prevent the slow down of transforming the data on the fly for the online apps or real time edits resulting in errors from an on the fly transformation.

As Dan Patterson hit the nail on the head in his response.

[RickGeittmann]

It would appear that the data is reprojecting on the fly from WGS84 to Web Mercator.

I would not edit the data in this case if my statement above is correct - editing in this manner will most likely induce errors into your data.

Remember the first data element added to a ArcMap session sets the data frame reference for the coordinate system unless someone manually set the coordinate reference first then brought the data in.

Information: Spatial reference does not match data frame—Help | ArcGIS for Desktop

[RobertStevens · Accepted solution]

Neil, Rick, and Dan

Thx for these responses: I am learning something from you guys.

Neil asks: "What is your reason for having the data frame in Web Mercator?".

I should have been more accurate and said that the frame is in WGS_1984_Web_Mercator_Auxiliary_Sphere projection (WKID 3857)

Does that change the discussion? But to answer the question: I am using basemaps supplied by Business Analyst.

Those basemaps "choose" WGS_1984_Web_Mercator_Auxiliary_Sphere for me.

(BTW: my area of interest is Portland OR of an area about 30x30 miles, and sometimes a larger area of Oregon, but not the whole state).

Business Analyst (I hope) is supposed to relieve the user of the necessity to worry about such things,

so I suppose I should reflect the question back to ESRI.

Would one of you expert people comment about WGS_1984_Web_Mercator_Auxiliary_Sphere.

And I apologize if, by not specifying the name exactly, I have caused unnecessary work.

Ty

Rob Stevens

[NeilAyres]

So is the business analyst data coming from an esri online service?

I guess so, since the frame defaults to "Web Mercator".

But for output, do you not switch to your local coordinate system, state plane or whatever it might be?