With averaging, if the random errors dominate the systematic errors, you will in fact have a better idea where the road is. IIRC, during the course of several hours ionospheric delays would cause systematic errors, but if you instead drive/ride the same road on different days, it's going to be relatively random.<br>
<br>Around Vancouver, I know of at least one place where Geobase is less precise than data I have from my GPS (my 20+ tracks can not be all wrong and offseted _in the same direction_). Needless to day, I moved the GeoBase road to match the averaged tracks in that place.<br>
<br>Michael.<br><br><div class="gmail_quote">On Mon, Nov 16, 2009 at 7:26 PM, James Ewen <span dir="ltr"><<a href="mailto:ve6srv@gmail.com">ve6srv@gmail.com</a>></span> wrote:<br><blockquote class="gmail_quote" style="border-left: 1px solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;">
<div class="im">On Mon, Nov 16, 2009 at 6:48 PM, john whelan <<a href="mailto:jwhelan0112@gmail.com">jwhelan0112@gmail.com</a>> wrote:<br>
<br>
> Multiple approximations can often be very accurate when averaged.<br>
<br>
</div>I agree... put a commercial grade GPS on a spot, and record the<br>
results over a long enough period, you can come up with a close<br>
approximation of where they device was located.<br>
<br>
However, driving up and down a road 5 or 6 times is not going to give<br>
you a superior representation of where the real road is located. You<br>
are not going to get the exact same locations recorded on each<br>
subsequent trip down the road, so each location has to be looked at as<br>
a point of dubious quality.<br>
<br>
The OSM project is always going to be of "dubious quality" because of<br>
the equipment used to capture the data, and the abilities of the<br>
person doing the data entry.<br>
<br>
Those needing super accurate data need to be looking elsewhere. This<br>
project will never be able to provide sub-metre accuracy.<br>
<br>
>From my observations, the GPS accuracy that I regularly observe puts<br>
me within 5 to 15 feet or 2 to 5 meters of reality. This would be in<br>
line with the GeoBase stated accuracy factor as well.<br>
<br>
James<br>
VE6SRV<br>
<div><div></div><div class="h5"><br>
_______________________________________________<br>
Talk-ca mailing list<br>
<a href="mailto:Talk-ca@openstreetmap.org">Talk-ca@openstreetmap.org</a><br>
<a href="http://lists.openstreetmap.org/listinfo/talk-ca" target="_blank">http://lists.openstreetmap.org/listinfo/talk-ca</a><br>
</div></div></blockquote></div><br>