Coordinates in Astronomy.owl

[Ed Shaya]

I promised to discuss coordinates in an OWL ontology but I got deterred 
big time.  Here is my OWL description of Coordinates.

The syntax is triplets of:
Class Property ObjectOfProperty

We begin with some class inheritances:

EquatorialCoordinates isa Coordinates isa Position  isa Location
Same for GalacticCoordinates, SuperGalacticCoordinates, 
EclipticCoordinates, etc

RA isa CoordinateValue isa Component isa Quantity
Same for DEC, GLON, GLAT, SGL, SGB, X, Y, Z, etc
RightAscension isEquivalent RA
Declination isEquivalent DEC

RAhat isa UnitAngle isa Angle
UnitAngle isa AngularCoordinate isa Coordinate
Same for DEhat, Lonhat, etc

Xhat isa UnitDirection isa CartesianCoordinate isa Direction
UnitDirection isa Coordinate
CylindricalRadius isa Radius isa Coordinate
Same for Yhat, SGXhat, SGYhat, SGZhat, etc

EquatorialCoordinateSystem isa CoordinateSystem
Same for GalacticCoordinateSystem etc.

Origin isa Point isa Position
=====================
Now, some properties of relevant classes:

Coordinates hasQuantity CoordinateValue (multiple)
	"       inCoordinateSystem CoordinateSystem (single)

CoordinateValue (has properties of Quantity: values, error, min, max, etc)
    "            of (owl:SomeValueFrom Coordinate) (single)
So, for instance:
X of Xhat means any X-value is in the Xhat direction.

RAhatJ2000 could be the Instance of RAhat that describes RAhat for the 
particular EquatorialSystem J2000. And, RAJ2000 could be the subclass of 
RA quantities with "of RAhatJ2000".

CoordinateSystem hasOrigin Origin (single)
         "        hasCoordinate Coordinate (multiple)

EquatorialCoordinateSystem hasEquinox Date

Direction fromPoint Point
     "     toPoint Point

Angle fromDirection Direction
   "   toDirection Direction

Point hasCoordinates Coordinates (single)
[Note Point isa Position but it hasCoordinates, even though Coordinates 
isa Position as well]
++++++++++++++++++
I beleive this lets one set up a complete definition for any coordinate 
system.  For instance the GeocentricEquatorialSystem can be described as 
an origin at the center of the Earth.

Have I forgotten about the issue of time?  Position is a function of 
time.  No.  Position in a coordinate system is the position in the 
coordinate system.  Objects may move through the coordinate system, so 
time should be handled by the properties of PhysicalObjects.

PhysicalObject hasLocation PositionAtTime
PositionAtTime hasLocation Position (cardinality 1)
			   atTime Time (cardinality 1)

If you are interested in examples of RDF headers added to the VO_Catalog
to implement this for search, let me know and I will send them.  Not for 
the faint of heart.

Ed