WGS 1984 for GoogleEarth and Bing

3053
3
06-14-2010 03:37 PM
GISDept
New Contributor
Hello and thank you in advance.

I am still, unfortunately, a little confused on the projection for posting data in Google Earth:

I understand that GE and MS apps use a spheroid only model, which is the projection ESRI calls:
Geographic Coordinate System, (spheroid-based)  "GCS_WGS 1984 Major Auxiliary Sphere" or
Projected Coordinate System (World) "WGS_1984_Web_Mercator"

Whereas ESRI uses a spheroid or ellipsoid model, which is the projection ESRI calls:
Geographic Coordinate System (World)  "GCS_WGS_1984" or
Projected Coordinate System (World) "WGS_1984_Web_Mercator (Auxiliary Sphere)"

The instructions in Technical Article ID: 34749 say when using 9.3 SP1 to use the:
Predefined Projected Coordinate System (World) "WGS_1984_Web_Mercator"

HERE IS MY PROBLEM:
When projecting from NAD83 to Projected WGS_1984_Web_Mercator in the toolbox,
there is a message that this is an undefined projection, and I can only do it with a
double: NAD_1983_To_WGS_1984_5 AND WGS_1984_Major _Auxiliary_Sphere_To_WGS_1984

The same is true (obviously) if I try to project from NAD83 to Geographic GCS_WGS_1984_Major_Auxiliary_Sphere

Isn't this exactly the same as just going from NAD_1983_To_GCS_WGS_1984, which is the ESRI model,
not the Google Earth or MS (Bing) model, right?

In the end, projecting to GCS_WGS_1984 is equivalent to
projecting to PCS_WGS_1984_Web_Mercator which is equivalent to
projecting to GCS_WGS_1984_Major_Auxiliary_Sphere

The only difference is that the denominator of the flattenting ratio for the PCS_WGS_1984_Web_Mercator is infinity
instead of 298.257224.   No matter what, the locations lie directly on top of one another and NOT where I can
see them (my well points) on the Google Earth image.

So, how do folks line up their very locatlized point locations on a Google Earth visualization for their projects/clients?

In addition, using the double transform to get to
Projected Coordinate System (World) "WGS_1984_Web_Mercator"
yields the same projected or local bounding coordinates as
Projected Coordinate System (World) "WGS_1984_Web_Mercator (Auxiliary Sphere)"
HOWEVER, Google Earth will not import the file "WGS_1984_Web_Mercator (Auxiliary Sphere)"  (says zero features)
Why does this happen?

HELP!!

Nicki Cook
Tags (2)
0 Kudos
3 Replies
MelitaKennedy
Esri Notable Contributor
Hello and thank you in advance.

I am still, unfortunately, a little confused on the projection for posting data in Google Earth:

I understand that GE and MS apps use a spheroid only model, which is the projection ESRI calls:
Geographic Coordinate System, (spheroid-based)  "GCS_WGS 1984 Major Auxiliary Sphere" or
Projected Coordinate System (World) "WGS_1984_Web_Mercator"

Whereas ESRI uses a spheroid or ellipsoid model, which is the projection ESRI calls:
Geographic Coordinate System (World)  "GCS_WGS_1984" or
Projected Coordinate System (World) "WGS_1984_Web_Mercator (Auxiliary Sphere)"

The instructions in Technical Article ID: 34749 say when using 9.3 SP1 to use the:
Predefined Projected Coordinate System (World) "WGS_1984_Web_Mercator"

HERE IS MY PROBLEM:
When projecting from NAD83 to Projected WGS_1984_Web_Mercator in the toolbox,
there is a message that this is an undefined projection, and I can only do it with a
double: NAD_1983_To_WGS_1984_5 AND WGS_1984_Major _Auxiliary_Sphere_To_WGS_1984

The same is true (obviously) if I try to project from NAD83 to Geographic GCS_WGS_1984_Major_Auxiliary_Sphere

Isn't this exactly the same as just going from NAD_1983_To_GCS_WGS_1984, which is the ESRI model,
not the Google Earth or MS (Bing) model, right?

In the end, projecting to GCS_WGS_1984 is equivalent to
projecting to PCS_WGS_1984_Web_Mercator which is equivalent to
projecting to GCS_WGS_1984_Major_Auxiliary_Sphere

The only difference is that the denominator of the flattenting ratio for the PCS_WGS_1984_Web_Mercator is infinity
instead of 298.257224.   No matter what, the locations lie directly on top of one another and NOT where I can
see them (my well points) on the Google Earth image.

So, how do folks line up their very locatlized point locations on a Google Earth visualization for their projects/clients?

In addition, using the double transform to get to
Projected Coordinate System (World) "WGS_1984_Web_Mercator"
yields the same projected or local bounding coordinates as
Projected Coordinate System (World) "WGS_1984_Web_Mercator (Auxiliary Sphere)"
HOWEVER, Google Earth will not import the file "WGS_1984_Web_Mercator (Auxiliary Sphere)"  (says zero features)
Why does this happen?

HELP!!

Nicki Cook


Ni Nicki,

You're caught in between our transitioning to "WGS_1984_Web_Mercator (Auxiliary Sphere)". Both it and "WGS_1984_Web_Mercator"  will have exactly the same coordinate values for the same location on the Earth. They're just different ways of emulating the coordinate system used by Bing and Google maps when they cache their data.

If you access the coordinates as latitude-longitude values, they are WGS84. But what's cached is in a Mercator projection where the math supports a sphere only. For the sphere's radius, they use the semimajor axis (equatorial radius) of WGS84: 6378137.0 meters.

So why are there two versions?

When we originally made a definition for "Web Mercator", we had only one implementation of Mercator. It contains equations to use with a sphere or a spheroid/ellipsoid. The only way to trigger the sphere equations that are needed is to use a geographic coordinate system (GCS) based on a sphere. We built a custom WGS 1984 Major Auxiliary Sphere for it and defined "WGS_1984_Web_Mercator". Later, we added a new implementation of Mercator that supports a parameter (Auxiliary Sphere Type), which describes how to calculate a sphere's radius from a normal spheroid/ellipsoid-based GCS and use it with the projection. The projection's name is Mercator_Auxiliary_Sphere.

Both versions do the same thing, force spherical math. The earlier version, using the custom GCS causes problems because you have to do a geographic (datum) transformation to it, so it's awkward. The 'new' version uses the standard WGS84 definition, so it simplifies the workflow.

The version you're using hasn't been updated to support caching your data with the newer "WGS_1984_Web_Mercator (Auxiliary Sphere)".

Melita
0 Kudos
GISDept
New Contributor
Thank you Melita:

So, really I should just let it transform on the fly or apply the closest fit for now, and wait for 10?

Nicki
0 Kudos
GISDept
New Contributor
Hello Melita:
We just noticed that since our last discussion, newly surveyed points (NAD1983) that previously
best fit was to let it convert on the fly (essentially NAD_1983_To_GCS_WGS_1984_1) was the closest fit - now are about five feet south of the visible locations on Google Earth - consistently.

Has the transition to WGS_1984_Web_Merc_Aux_Sphere been completed? (We are still using 9.3.1). You were in process back in June of this year. 

Has there been any other change or update in the process for
aligning GIS data in the GE environment?

Thank you for your help!
0 Kudos