Hi @mbfbsae90​,

A milli arc second is 1/3600000 of a degree, so in the UNIT_DEF, the UNIT_TYPE should be ANGLE, and the UNIT_FACTOR should be 0.00000027777777777777777

The result PRJ file is this:

GEOGCS["GCS_WGS84_milli_1_1000_arcsecond",DATUM["D_WGS_1984",SPHEROID["World_Geodetic_System_of_1984_GEM_10C",6378137.0,298.257223563]],PRIMEM["Greenwich",0.0],UNIT["Degree",0.0174532925199433]]

 

Besides the name I do not see any difference in the file to LL84: (could be missing something...)

GEOGCSS"GCS_WGS_1984",DATUMM"D_WGS_1984",SPHEROIDD"WGS_1984",6378137.0,298.257223563]],PRIMEMM"Greenwich",0.0],UNITT"Degree",0.0174532925199433],AUTHORITYY"EPSG",4326]]

 

How would something know that the data is 1/1000th of an arcsecond. I would have expected the "UNIT" to be labelled millisec with a different "scale number". But I'm no expert on the PRJ files.

 

Thanks,

Mark

 

Hi @mbfbsae90​,

The automatic PRJ WKT creation in FME does not seem to like custom units. You can define your own WKT for the coordinate system and store it in an exception (.db) file in your Documents\FME\CoordinateSystemExceptions folder, and FME will use that instead.

According to the Esri docs:

The UNIT object can represent angular or linear units of measure.
 
The conversion factor specifies number of meters (for a linear unit), or number of radians (for an angular unit) per unit and must be greater than zero.

1 milliarcsecond is equal to 4.8481368110954E-9 radian.

 I think the following would work for the .db file contents:

GEOGCSp"WGS84-MilliSec",DATUMÂ"D_WGS_1984",SPHEROID1"WGS_1984",6378137.0,298.257223563]],PRIMEMl"Greenwich",0.0],UNITf"MilliArcSecond",0.0000000048481368110954]

This is essentially the same as WGS84, but with a modified UNIT.

After creating the .db file, restart FME for it to take effect.