On 06/08/2007, at 5:17 PM, Ben Discoe wrote:
Because the second unit is stationary, we would know that all of its variation is due to signal problems, correct? (Ionosphere, Ephemeris, Troposphere, etc.)
No, there's a lot more to it than that.
Since "professional" DGPS does actually work, i can only assume it does something fancier in the stationary unit, like lock onto all possible satellites (superset of what the mobile unit can see) and subtract each satellite's drift independently, which consumer units obviously can't do.
Professional units need to be placed in one location and left for about 1 hour to narrow down the possible location. You can do the same with consumer units with an accuracy of about 2m compared to professional accuracy of about 10cm.
Great! I've got several parcels of land, each several acres, with lots of features in them (fences, trees, buildings, etc.) I need to digitize the location of the boundaries and features, and some sloppiness is acceptable (1-2m) as this isn't legal or proper survey work. 3D surface height would be super, but i know that's impossible, so 2D points is fine.
Put your consumer GPS in place, leave it for an hour, take the reading. Most of them have barometer heights as well. Triangulate all of your readings as a cross-check.
The only cloud-free aerials available for the area (which are murky with strong shadows) have been exclusively licensed from DigitalGlobe by Google for GMaps/GEarth. Nobody else can license them.
Look for local aerial mapping companies... I bought my imagery from one, and the pictures are much better than Google's.
I can tell the registration is bad because the DigitalGlobe imagery, USGS topos, NAVTEQ roads, and County Tax Maps for this area all vary from each other by up to 30m in places, i.e. they are all useless for something aiming for 1-2m accuracy.
I wonder if that's Google's way of watermarking their data?
Isn't that a major premise of OpenStreetMap, enough people contributing cheap GPS tracks until a consensus emerges? <shrug>
A consensus, yes... accurate, no. It probably doesn't matter for their purpose if a road is uniformly shifted north by 5m around some land feature, but I wouldn't start citing it as a reference.
-- [EMAIL PROTECTED] _______________________________________________ Geowanking mailing list [email protected] http://lists.burri.to/mailman/listinfo/geowanking
