Wikidata:Property proposal/Water area

water area

Originally proposed at Wikidata:Property proposal/Place

Done: water area (P8887) (Talk and documentation)

Description	Area/surface of a geographical entity that is covered by water.
Data type	Quantity
Domain	places
Allowed values	numbers
Allowed units	square kilometers
Example 1	Dauphin Island (Q932623) → 105.344035 square kilometers
Example 2	Germany (Q183) → 8634 square kilometers
Example 3	California (Q99) → 20501 square kilometers
Planned use	For the US states, cities and Census-designated places, these data is available as CSV files for 1990 to 2010. The data for 2020 will be available next year. It is easy to import them to wikidata, as the US Census Bureau will again publish CSV and Excel files.
See also	area (P2046), water as percent of area (P2927)

Motivation

Currently, for geographic entities such as countries, states and cities, there is no good possibility to have a statement about the water surface. The proposed property is comparable to water as percent of area (P2927), but should be an absolute value (usually in sqaure kilometers) and not relative. It will be a part of area (P2046) and should always be smaller or equal to it. Yellowcard (talk) 14:35, 26 July 2020 (UTC)[reply]

Discussion

Comment this could be done with area (P2046) and a suitable qualifier, but this does seem a useful metric so we should figure out some way to get it in! ArthurPSmith (talk) 14:48, 27 July 2020 (UTC)[reply]
- @ArthurPSmith: Good point. Not sure how exactly water as percent of area (P2927) is used, I have added it as separate statement but this could also be used as qualifier for area (P2046). I personally would like to have all those three statements for separate statements and use qualifiers for date and method, but in the end we should only agree on one specific procedure. Is there maybe anything comparable? Yellowcard (talk) 21:20, 27 July 2020 (UTC)[reply]
  - water as percent of area (P2927) should apparently only be used as a main value. The issue I see here is that if a place has area (P2046)100 km² and water as percent of area (P2927)36% one can't be certain that the percentage refers to the earlier statement unless they have the same point in time (P585) and source. How is the relation between area (P2046) and the proposed property? Do you add both values to get the combined land and water area (like US Census), or is the water area part of area (P2046)? --Pyfisch (talk) 19:03, 23 August 2020 (UTC)[reply]
    @Pyfisch: I agree, water as percent of area (P2927) as well as the proposed property are supposed to be used as main values. However, I do not understand what you mean by "Do you add both values to get the combined land and water area (like US Census)", as in the US Census data, the "area" combines land and water area. This is why area (P2046) will always be greater than the statement with the proposed water area property, as long as they base on the same point of time (point in time (P585)) and the same source. Yellowcard (talk) 05:56, 25 August 2020 (UTC)[reply]
    Take a look at the Florida Places Gazetteer File. It contains an entry "Everglades city" with the columns ALAND=2370668 and AWATER=738078. According to the technical documentation ALAND is the "Land Area (square meters)" and AWATER is the "Water Area (square meters)". To get the "total area" of 3.11 km² both values are combined (e.g. w:Everglades City, Florida Infobox). You most likely knew this already, but I want to make sure there is no misunderstanding. That area (P2046) is always the total area is reasonable, I just want to be explicit about it. --Pyfisch (talk) 08:44, 25 August 2020 (UTC)[reply]
    
    Nevermind, you say at the very top "It will be a part of P2046 and should always be smaller or equal to it", sorry for the confusion. --Pyfisch (talk) 08:48, 25 August 2020 (UTC)[reply]

Comment instead of separate "water area" statement we could also introduce a qualifier "thereof water area" to area (P2046) claims. This would ensure that both numbers refer to the same point in time. But for the similar case of female population (P1539) and male population (P1540) separate items were chosen too. --Pyfisch (talk) 08:44, 25 August 2020 (UTC)[reply]

I believe, based on the structure mainly used in Wikidata, that the property should be used as main statement. Furthermore, while we are discussing the water area, I believe that we need a "land area" property as well, don't we? Absolutely same usecase as the proposed water area property. Yellowcard (talk) 21:02, 25 August 2020 (UTC)[reply]

We don't need a "land area" property as area (P2046) - water area is always the land area. Creating a land area property would be redundant. --Pyfisch (talk) 07:17, 26 August 2020 (UTC)[reply]

Not sure whether that is a valid argument. Based on the redundancy idea, we have a lot of properties that could be called redundant. Just taking your example above: Nobody would call population (P1082) redundant because you could just sum up female population (P1539) and male population (P1540) (and maybe other genders) to get that number. There are many better examples, but I just took this one as you were referring to those properties above. Yellowcard (talk) 16:57, 26 August 2020 (UTC)[reply]

It is a philosophical issue at its core. I don't see the utility of a land area property in addition to the water area. --Pyfisch (talk) 00:47, 27 August 2020 (UTC)[reply]

There are cases where we know the full population (P1082) of a jurisdiction but not female population (P1539) nor male population (P1540). population (P1082) is useful for those cases. I don't think there are many cases where where we know the land area but neither the water area nor the total area.