wasserkarte

Data model

Every water body is one JSON record in data/waters/*.json. Records form a tree: parent points to the water that receives this one (its Vorfluter); the three seas are the roots.

Nordsee ← Rhein ← Mosel ← Saar ← Blies

The Ordnung (order) of a water is its depth in that tree: 1 = drains directly into a sea or coastal water, 2 = into an order-1 water, and so on (the dataset reaches order 5).

Files

file	contents
meere.json	the three roots: Nordsee, Ostsee, Schwarzes Meer
rhein.json, elbe.json, donau.json, weser.json, ems.json, oder.json, maas.json	one river system each, including its lakes and canals
nordsee_weitere.json	Deutsche Bucht (Wattenmeer, Jade, Dollart …), Eider, Vechte, Wiedau and their tributaries
ostsee_weitere.json	Baltic bays, Förden, Bodden, Haffs and the coastal rivers draining into them

Placement is computed from the ancestry — add a record to any file and run make fix. Records are stored one per line so a diff shows exactly which water changed.

Fields

field	required	type	meaning
id	✓	string ^[a-z0-9_]+$	unique slug, umlauts transcribed (fraenkische_saale); homonyms get a suffix (wuerm_amper)
name	✓	string	German name with umlauts, ss instead of ß
type	✓	river lake canal coastal sea
parent	✓	id or null	receiving water; null only for seas
side	✓	"L" "R" null	bank of the parent where the mouth lies, looking downstream along the parent
length_km	✓	number or null	total length; null for lakes, coastal waters, seas, and when undocumented
length_de_km		number	length of the German section, if different
area_km2		number	lakes and coastal waters
source	✓	[lat, lon] or null	source · lake centre · canal start · inner point of a coastal water
mouth	✓	[lat, lon]	mouth · lake outflow · canal end · opening to the sea
mouth_place	✓	string	e.g. "Koblenz, Deutsches Eck"
states	✓	list	German states as ISO 3166-2:DE suffix (BY, BE = Berlin), foreign countries as ISO 3166-1 alpha-3 (AUT, BEL, NLD …)
historic		true	waterway out of service (Ludwigskanal, Eiderkanal)
note	✓	string	one concise fact, ideally ≤ 95 characters

Example:

{"id": "mosel", "name": "Mosel", "type": "river", "parent": "rhein", "side": "L", "length_km": 544.0, "length_de_km": 231, "source": [47.901, 6.885], "mouth": [50.364, 7.606], "mouth_place": "Koblenz, Deutsches Eck", "states": ["RP", "SL", "FRA", "LUX"], "note": "Grösster Nebenfluss des Rheins, Quelle in den Vogesen am Col de Bussang"}

JSON Schema: schema/waters.schema.json (editors can validate while typing — VS Code picks it up from .vscode/settings.json). Cross-record rules are checked by scripts/validate.py.

Rules of the tree

• Scope. Waters with a German connection: flowing through or along Germany, plus the foreign upper or lower course of a German river, so that every chain reaches the sea.
• Lakes hang on the water they drain into, even if a short outflow is not modelled (Königssee → Salzach is recorded as parent: "inn" with a note).
• Canals hang on the water their end point merges into. The other connection is named in note.
• Coastal waters nest inside each other (Kleines Haff → Stettiner Haff → Pommersche Bucht → Ostsee) and receive the coastal rivers (Peene → Peenestrom).
• Source-river pairs (Brigach/Breg, Weisser/Roter Main, Fränkische/Schwäbische Rezat) have side: null.

Sources for facts

Lengths, mouth locations, banks and coordinates come from the infoboxes of the German Wikipedia. The canal records were individually checked for length, end points, receiving water and operating status. When adding a water, cite nothing in the record itself but verify against the article — and prefer null over a guess.

This site is open source. Improve this page.