[OSM-talk] coastline accuracy?

mle info at dynoyo.plus.com
Thu Apr 30 00:34:52 BST 2009


Hi All

Thanks for the replies, I re-examined my tracks, existing roads and the 
Yahoo layer - if only we had an extra zoom level or two it would be more 
obvious ...

The island in question is Menorca.

The road that I travelled every day, due to a one way traffic system, is 
along the edge of an east-facing limestone cliff with a sharp vertical 
drop to sea level.  So high water and low water are the practically the 
same from a satellite view.

The road-to-cliff edge distance is mostly about 1-2 times the GPS stated 
accuracy. The GPS tracks were consistently plotted further west than 
expected relative to the PGS data coastline, which at that point matches 
the Yahoo layer pretty well.  Repeat journeys matched very well.

Looking at other thin promontories, this was consistent along the 
coastline. Some inland roads also appear displaced.

I then loaded the same GPS tracks into another well known satellite 
viewer which allows higher zoom levels.  They all seemed to have a more 
accurate relation to the coast. I checked this at various locations 
around the island and a consistent shift is seen. So local coastal 
adjustments won't be sufficient.

My naive conclusion is that both the whole PGS and Yahoo layers need to 
be shifted, together, in the south west direction to get everything 
lined up with the road data.

Is this possible and acceptable?  If so, any hints as to how to do this 
would be welcome.

regards,
Martyn




More information about the talk mailing list