[OSM-talk] Coastline Update

[Paul Norman]

> -----Original Message-----
> From: Andrew Errington [mailto:a.errington at lancaster.ac.uk]
> Sent: Saturday, March 31, 2012 6:01 PM
> To: talk at openstreetmap.org
> Subject: Re: [OSM-talk] Coastline Update
> 
> On Sun, 01 Apr 2012 09:31:53 Clifford Snow wrote:
> > On Fri, Mar 30, 2012 at 8:26 PM, Paul Norman <penorman at mac.com> wrote:
> > > I have completed another coastline generation and it has uploaded.
> > > This version respects odbl=clean.
> > >
> > > The shapefiles are in their normal place at
> > > http://pnorman.dev.openstreetmap.org/coastlines/
> > >
> > > Included is a .osm file with all the error points.
> > >
> > > There are no significant multi-square flooded or dry areas. The
> > > following areas have significant number of error points:
> > >
> > > Pudget Sound in Washington State
> > >
> > > I just pulled the osm file into josm.  I did a check of a couple of
> > > error
> >
> > points in the osm file in Puget Sound (Edmunds WA.)  They appear to be
> > about 10+ meters from the shoreline according to the bing image.
> > Since I'm new at this, but since I live in the Puget Sound and would
> > like to help fix it, what is an acceptable error distance?  Also - the
> > bing image doesn't give a date and time to determine the tide level.
> > Or am I missing something?
> 
> Are you sure the Bing image is correctly aligned?  There is no guarantee
> that it is.  Find a bunch of GPS traces nearby for a road or other very
> visible feature and ensure they line up with that feature on the Bing
> image.  If they do, you're fine- use the Bing image to correct the
> coastline.  If they don't, then use the tools in Potlatch or JOSM to
> move the Bing layer into alignment
> *then* correct the coastline.
> 
> The coastline should follow the line of high tide.  Use your best
> judgement to determine this.  You're not missing something.

The MapQuest Open Aerial imagery should be pretty well aligned.

My procedure is as follows:

Pick an area that needs work and download it
Select natural=coastline | child natural=coastline in JOSM and run the
validator on it
Pick somewhere to start and delete dirty coastline and then start tracing. 

Just be careful to not accidentally upload the .osm that indicate the
problems.



A new set of files is up, with data from about 18 hours ago I believe.
processedc_p.tar.bz2 is still uploading. I have another fresh set generating
too.