[OSM-talk] Coastlines

[J.D. Schmidt]

Andy Robinson skrev:
> That's not surprising as the Landsat data position is not that accurate.
> There is a shift in the Landsat. In my Birmingham area I normally have to
> slide items created off the Landsat imagery (railway lines and rivers etc)
> by about 50m to the south west.
> 
> Cheers
> 
> Andy
> 
> Andy Robinson
> Andy_J_Robinson at blueyonder.co.uk 
> 
After getting OJW to import PGS coastlines for my area here in Denmark, 
the accuracy of the coastline is pretty good, even small marina quays 
shows up.
But it is evident there's a consistent shift to the south east. My 
preliminary investigation shows that the coastline is about 30 meter 
south and 18 meters east of where it should be. Square mentioned on the 
IRC channel, that thediscrepancy of 30 meters S-N direction, and 18 
meters E-W direction is equivalent to about 1 arcsecond in both cases.


I have so far only checked it at two places (a bit north of Copenhagen 
near Hellerup, and at the end of the small island of Amager where 
Copenhagen Airport is located.) I've checked it against the municipality 
  data at the danish KMS geo-portal site, as well as making gpx tracks 
and plotted them on the high resolution Digital Globe 2006 sat images 
available on gmaps for Copenhagen. In both cases they concur with the 30 
meters south and 18 meters east shift. I'll post some screenshots on the 
wiki later.


The danish PGS coastline is not from the same set of the files where the 
IoW data was imported. AFAIR the danish data comes from the CD21 data, 
while the IoW data came from the CD19 data. But if it is the same 
discrepancy of 1 arcsecond in both cases, the import script should be 
able to change the uploaded coordinates accordingly.

J.D. "Dutch" Schmidt