[talk-ph] Does WMS/Landsat have an offset?

[Ed Garcia]

Yup!  Always WGS84...  anyway, from my experience, the PRS92 vs WGS84 offset
is nearly 200 meters (not just 50m)

hmmm... I have just tried the position adjustment in wms in josm.  Great!  I
did not know it was possible to move the image relative to the other
layers.  But now I also take it that the landsat position as not always
precisely positioned.

thanks

On Sat, Apr 11, 2009 at 12:26 PM, Maning Sambale <emmanuel.sambale at gmail.com
> wrote:

> Some pointers:
> 1) Are you sure you've set you GPS unit projection and datum to Geog
> lon/lat WGS-84?  I found gps units set to PRS92 with the same offset.
>
> 2) You can actually adjust the position of the wms layer in josm.  There
> is an icon with arrows in four cardinal directions.  I think you can
> also do that in potlatch.
>
> 3) Landsat has various offset errors depending on the location
> (sometimes by just 10 meters or so)
>
>
> On Fri, 2009-04-10 at 21:50 +0800, D Tucny wrote:
> > 2009/4/10 Ed Garcia <eppgarcia at gmail.com>
> >         Related my earlier reply post on coastlines, here is an
> >         example showing that the landsat image is probably offset or
> >         "not calibrated"?
> >
> >         here is a screen capture of JOSM showing Aslom Island south of
> >         Mindoro:
> >
> >         http://www.waypoints.ph/images/aslomisland.jpg
> >
> >         The blue trace is what OSM/Mapnik is currently showing as
> >         Aslom Island (quite a big error),  And, the green raster image
> >         is the landsat image showing Aslom Island.  Now, the yellow
> >         trace is an actual GPS track on a boat that was actually made
> >         to go around the island with its outrigger (katig) just about
> >         5 meters from the land edge to trace the shape of the island.
> >         The GPS has the 12 satellites in view (which means very very
> >         low position error).
> >
> >         As can be seen, the offset is about 50 meters between the
> >         landsat image and the GPS tracks.  This is about the same
> >         offset observed when we tracked the mangrove river in Sariaya
> >         that I mentioned in my earlier post.
> >
> >         Does this indicate that the landsat image is generally offset
> >         or not calibrated?
> >
> > Landsat is generally 'good enough' at lower zoom levels, but, at that
> > level it's pretty much at it's limit... It's not often perfect, but,
> > it does vary between being pretty close and pretty far off the mark...
> > Around here the offset looks to be about 40m to the north east based
> > on offset of bridges and riverbank compared to GPS tracks and yahoo
> > (which matches the GPS tracks pretty well, at least for roads that
> > existed back then...)
> >
> > d
> >
> > _______________________________________________
> > talk-ph mailing list
> > talk-ph at openstreetmap.org
> > http://lists.openstreetmap.org/listinfo/talk-ph
>
>


-- 
website administrator:
- www.waypoints.ph
- reeflife.eppgarcia.com

PADI Divemaster #491048
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstreetmap.org/pipermail/talk-ph/attachments/20090411/5603efea/attachment.html>