On Tue, Feb 2, 2010 at 4:59 AM, John Smith <[email protected]>wrote:
> On 1 February 2010 13:31, Anthony <[email protected]> wrote: > > that altitude error will always be worse than lat/lon error. But I think > > that's a good argument for not recording absolute elevation but rather > > recording some sort of relative elevation. > > The reason for absolute elevation is simply because most GPS software > isn't capable of differentiating between relative layers, although now > that I think about it some more, from a programming point of view this > seems it should only be an issue until the GPS figured out which layer > you are currently driving on. > I'm not sure what GPS software you're talking about. For recording of altitudes, I'm not sure GPS is not accurate enough to be very useful. Also, there's the issue of what vertical datum is being used. For most GPSes it is easy to set the datum for the lat/lon to WGS84, but the vertical datums vary quite a bit, they're often hard to determine, and in many cases they can't be changed. On the other hand, it's pretty easy to find out the height of an overpass, and/or estimate it (by photographs, by memory, or even by triangulation using free apps available on some smart phones). Therefore, I believe these two measurements should be stored separately. We know that overpass A passes 6 meters over road B, within say 1/3 of a meter. Whereas we know that road B is at 10 meters above sea level, with respect to NAVD83, within say 4 meters (which is a reasonable vertical GPS error bound). If we had measured the roads separately and given them absolute elevations, we might be off by 4 meters on overpass A, and 4 meters in the opposite direction on road B, and not even know which one is above the other. And then when we drive on the road our GPS readings might be off by another 4 meters.
_______________________________________________ Tagging mailing list [email protected] http://lists.openstreetmap.org/listinfo/tagging
