Coordinate System differences between ArcMap 10.2 and 10.3

[RebeccaBasquez1]

Hello,

I originally created an MXD in ArcMap 10.2 using a custom Projection of NAD 1983 UTM Zone 14N in US Foot. A couple of months later our group upgraded to 10.3.1.

When running a Select by Location on points that fall on a line in the original 10.2 MXD I get back 148 points. When I bring in the same data in a new 10.3.1 mxd, I run the same Select by Location and only get 50 points. If I add a buffer of 0.01 feet in the search distance it returns 148 features.

I exported the projection file from the 10.2 mxd and imported it into the 10.3 mxd. Then I ran the Select by Location again with no buffer and then finally got 148 features selected.

Has anyone else experienced this? Is there a workaround to this without having to import the projection file everytime I created a new MXD file?

Thanks,

Rebecca

[DanPatterson_Retired]

post the projection files (*.prj if possible) to see if the key difference can be noted.. I presume there was no upgrade in the Datum between the two versions?

[RebeccaBasquez1]

Hi Dan,

I attached both .prj files, but when I import either one into an empty 10.3.1 MXD I get the 148 features selected.

I'm starting to wonder if it's an issue with our SDE, which is in 10.2, but if you can take a look I'd appreciate it.

Thanks,

Rebecca

[DanPatterson_Retired]

I see no difference... SDE... there is that thing again rearing its head... work local

PROJCS["NAD_1983_UTM_Zone_14N",
GEOGCS["GCS_North_American_1983",
DATUM["D_North_American_1983",
SPHEROID["GRS_1980",6378137.0,298.257222101]],
PRIMEM["Greenwich",0.0],
UNIT["Degree",0.0174532925199433]],
PROJECTION["Transverse_Mercator"],
PARAMETER["False_Easting",1640416.666666667],
PARAMETER["False_Northing",0.0],
PARAMETER["Central_Meridian",-99.0],
PARAMETER["Scale_Factor",0.9996],
PARAMETER["Latitude_Of_Origin",0.0],
UNIT["Foot_US",0.3048006096012192]]

#------------------------------------------
PROJCS["NAD_1983_UTM_Zone_14N",
GEOGCS["GCS_North_American_1983",
DATUM["D_North_American_1983",
SPHEROID["GRS_1980",6378137.0,298.257222101]],
PRIMEM["Greenwich",0.0],
UNIT["Degree",0.0174532925199433]],
PROJECTION["Transverse_Mercator"],
PARAMETER["False_Easting",1640416.666666667],
PARAMETER["False_Northing",0.0],
PARAMETER["Central_Meridian",-99.0],
PARAMETER["Scale_Factor",0.9996],
PARAMETER["Latitude_Of_Origin",0.0],
UNIT["Foot_US",0.3048006096012192]]‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍

[MelitaKennedy]

We do have NAD83 "UTM" on US feet. It's called a BLM rather than UTM zone. It was added in 9.3, so it was available at 10.2. 

The only thing I can think of is that the ArcSDE data has an interesting (AKA nonstandard) xy tolerance value. When you import the prj file to set the data frame, the software is likely calculating the default xy tolerance value which should be 0.00328083. If that's bigger than what the ArcSDE data has, that would explain why Select by Location finds more points.

Melita