[Talk-ca] Coastline rendering problems, Richelieu River

[David Groom]

> ----- Original Message ----- 
> From: "Pierre Béland" <infosbelas-gps at yahoo.fr>
> To: "talk-ca" <talk-ca at openstreetmap.org>
> Sent: Sunday, June 10, 2012 11:27 PM
> Subject: [Talk-ca] Coastline rendering problems, Richelieu River
>
>
> There are also Coastline rendering problems on the Richelieu river around
> Saint-Jean-sur-Richelieu and Chambly. The OSM history changesets list 
> shows
> many Non-ct data remove and recreate operations in the area done by user
> PurpleMustang recently.
>
> In Saint-Jean-sur-Richelieu, to correct the situation, I redrawed part of
> the riverbank who was simply deleted.
>
> In Carignan, Ile Ste-Marie is still flooded. There is no riverbank 
> relation
> and I dont know if it has simply been deleted.
>
> In Chambly, the river enlarges. There are two river that connect to the
> basin and island around extremity of these two rivers. There is a relation
> that was probably containing all the riverbanks roles (outer and inner). 
> But
> the relation 305614 only contains to inner roles representing two islands.
>
> In south-east part of Chambly Basin, there is a point and islands on the
> east riverbank facing Fort de Chambly. The two islands included in the
> relation are rendered as lake and the water as land.
> see
> http://www.openstreetmap.org/?lat=45.45798&lon=-73.26392&zoom=15&layers=M&relation=305614
>
> Just
> on the right, there is the Huron river coming in. The polygon
> representing the part of the riverbank that join the Richelieu River is
> lost. And there is way waterway=river for this river. This means that
> the flow of water in the river is not indicated an no river name can be
> showed on the map.
>
> http://www.openstreetmap.org/?lat=45.46277&lon=-73.25257&zoom=16&layers=M&relation=1775822
>
> Part of the Huron riverbank joining into the Richelieu near the point as
> simply vanished.
> see
> http://www.openstreetmap.org/?lat=45.45798&lon=-73.26392&zoom=15&layers=M&way=120716933
>
> The riverbank around the basin is not included in the relation. This way 
> is
> also improperly traced on the point facing Fort Chambly. It should be 
> closer
> to the shore and not going around the tw islands.
> see
> http://www.openstreetmap.org/?lat=45.45798&lon=-73.26392&zoom=15&layers=M&way=164251774
>
> On this last image of the Basin, we see on top four islands for wich I
> cannot find the coastlines.Ile Goyer, Ile Demers, Ile-aux-Lièvres,
> Ile-aux-foins riverbanks have disappeared.North of these islands, the
> Riviere L'Acadie is joining Richelieu River. The junction of the two 
> rivers
> is improperly traced. Polygons 24346195 (Riviere L'Acadie riverbank) and
> 164254015 (Riviere Richelieu) simply do not join.
> http://www.openstreetmap.org/?lat=45.47204&lon=-73.2869&zoom=15&layers=M&relation=1775822
>
> I have to conclude again that these rendering problems are caused by
> improper definition of coastlines and riverbanks.
>

I'm not sure coastlines are the current issue here. As you yourself have 
pointed out there are missing ways, and ways drawn which do not meet the 
full extent of the water area, and ways which are not included in the 
relation.

I've tried to fix as much as I can

David

>
> In JOSM, we can see what the riverbank should look like using Geobase
> Imagery.
> wms:http://ows.geobase.ca/wms/geobase_en?service=wms&request=GetMap&version=1.1.1&SRS=EPSG:4326&style=&format=image/png&transparent=true&layers=nhn:hydrography,nhn:network&WIDTH={width}&height={height}&BBOX={bbox}&
>
>
>
> Pierre