tags:

views:

61

answers:

1

I have been poking around and have been unable to find a standard for what ellipsoid or geoid corresponds to "0" elevation in GPX files. Is there a standard? Do I just assume the WGS84 ellipsoid? Or if different standards are used, is there a way of figuring out from the file which one was used? Or making an educated guess?

And what is the source for your answer? I would love to have an authoritative reference.

+3  A: 

I clipped the following from the beginning of the the official GPX 1.1 Schema

<xsd:schema targetNamespace="http://www.topografix.com/GPX/1/1" elementFormDefault="qualified">
    <xsd:annotation>
        <xsd:documentation>
          GPX schema version 1.1 - For more information on GPX and this schema, visit http://www.topografix.com/gpx.asp  GPX uses the following conventions: all coordinates are relative to the WGS84 datum.  All measurements are in metric units.
        </xsd:documentation>
   </xsd:annotation>

which tends to support your assumption. Later in the schema explicit reference is made to the vertical datum to be used.

High Performance Mark
Thanks for the citation. I find where the schema says <ele> is "The elevation (in meters) of the point," but I do not see where it specifies the vertical datum. I am assuming that it is the WGS84 ellipsoid that is to be used, and not the geoid, as I doubt if many GPS units would dedicate the memory space required to record the geoid.
Dan Menes
@Dan: I read the schema quickly and misunderstood the definition of the element "geoidheight"
High Performance Mark
Interesting. The help for <geoidheight> references the "NMEA GGA message." This site (http://www.gpsinformation.org/dale/nmea.htm#GGA), which reports on someone's efforts to reverse-engineer the NMEA protocol, indicates that heights are relative to "mean sea level," i.e., geoid, and that the height of the geoid relative to the ellipsoid should be supplied (but isn't in the data I'm dealing with, of course). I don't have the quarter-kilobuck to buy the actual NMEA standard right now, so further clarification may have to wait.
Dan Menes