natural=bay may intersect

[matkoniecz]

http://osmose.openstreetmap.fr/en/error/e72c9fb6-3747-e9e8-c93a-d32dccecc024

https://www.openstreetmap.org/way/40980478#map=12/59.9751/5.8811 https://www.openstreetmap.org/way/561794024#map=12/59.9815/5.8849 seems perfectly fine

[verdy-p]

• the first way is a closed area (approximative) for a local bay portion of the fjord: it could be much better defined to follow the coastline (excludinog possible islets)
• the second way is for the fjord, and much larger but draw as a very rough line (why not an area containing the local bay and sharing outer ways with it?
• the fjord would have to be in fact larger, enclosing some large islands.

I don't know why you don't follow what has been made for riverbeds of waterways (after all, fjords/rias are riverbeds with maritime saline water climbing below fresh water of the river near the surface, initially they were glaciers flowing down with only fresh water, and there was no sea there; they were enlarged and by erosion by the glacier, and became deeper and eroded the flanks when the glacier melted and the sea came there).

This signal in Osmose is good as it detects serious work to do there: IMHO, a fjord should never exist as a linear feature (which is only meaningful for the main stream of waterway=river flowing into it.

• A fjord (or aber, or ria, these are mostly synonyms depending on countries or regions;) should be a "water=fjord" area (a closed polygon line or a multipolygon relation containing at one closed ring of "outer" ways and possibly some "inner" ways for their islands, but defined selarately from the waterway=river linear feature).
• A bay inside long fjord must use the same outer riverbank outlines, then closed by a straight segment (without any tag) through the larger water area between two capes of the fjord, or possibly along a dam or bridge enclosing it.

---

Some narrow fjords have sometimes been closed artificially by a dam and no longer have saline tidal water flowing upward; in that case they are no longer maritime and the coastal line should follow the dam (unless dams are just used to regulate the flow and maintain water lever upward and reduce the concentration of salt, because the fjord will be used as a source of fresh water for humans, for agriculure, or for navigation, or for keeping enough waters in their harbours; however with high tide, saline water may still pass over it and the dame could be temporarily opened to evacuate sediments needed downward to preserve soild level in coastal deltas and saline marshes, to to limit the erosion of coast or salinisation of coastal farmlands and residential areas: sediments allow vegetation to survive and help maintain sand and fragile rocks along the coasts; dams could have raid dramatic impacts on coasts, this has been seen on mediterranean coasts of Portugal and Spain, Adriatic coasts of Italy, Slovenia, Croatia, Atlantic coasts of southwestern France, and elsewhere in the world with dramatic effects on coasts in USA, China, India, Russia, Kazakhstan, and now as well in UK for dams on the river Thames...).

[matkoniecz]

IMHO, a fjord should never exist as a linear feature

For long and thin fjords this is 100% fine Such as https://www.openstreetmap.org/way/561794024#map=10/60.1948/6.5698

[frodrigo]

Not sure about this one.

This a more general issue than bay, or not ?

Maybe the physical feature does not intersect, but only the name?

[Marc-marc-marc]

so at the same time your are in the bay Kvinnheradsfjorden and in the bay Hardangerfjorden ? imho A in A if often wrong.
I have no knowledge of this place but I would think that either Hardangerfjorden is too long or there is a missing tag to describe a part of the bay

[matkoniecz]

imho A in A if often wrong

Nested bays (small bay in a large bay, fjord in a bay) is not something unusual, fractal structure in such case should not be assumed to be wrong