User:Amadalvarez/ontology/protected areas
This page is a draft with a proposal for ontology of Wikidata property for authority control for protected areas (Q55978235) and its related properties when use as statement of item.
 | While we are in discussion for approval, please make your comments in the discussion page and move the changes and conclusions to this page, in order to get a final elaborate and approved document when we end |
Description[edit]
By 2019, 4 years after Wikidata:WikiProject Protected areas started, the number of properties related with natural heritage IDs has grown reaching 42, as indicated in Natural heritage properties template on 2019, Jan, 31. Most of them are related with protected area (Q473972), which is where we have focused this work.
Following a discussion on WikiProject Protected areas, we have created this analysis page to:
- Make a pseudo-classification of ID properties to differentiate the "authority control" from other kind of repositories related with nature.
- Homogenize the definition, ontology and the way to get information from those considered "natural heritage authority control" properties, as it is done with cultural heritage (Q210272) or heritage site (Q358) using heritage designation (P1435), which also is able to hold natural heritage (Q386426)
- Roadmap to deploy the changes
- Provide guidelines to make it easy to load information to WD from these natural heritage catalogs.
Section "comparative property definitions" gather a summary of the 42 properties included in Template:Natural heritage properties by end of January 2019.
The list shows some key information of properties definition to develop the analysis. The changes proposed column resumes the proposal and is the object of the discussion on which to agree.
The "natural heritage authority control"[edit]
The elements in the list are a mixture of protected areas, open and / or common data bases of natural elements and specialized lists of some feature.
The green and yellow colored elements are those that must be considered as IDs of protected areas by an authority. Some of them are lists administered directly by the authorizing entity. Some others are commons databases that are used by authority entities to manage their catalog's information. This is the case of: protected areas INPN Code (P1848), WDPA ID (P809) and Common Database on Designated Areas ID (P4762).
The blue colored properties are IDs lists of some kind of elements related to nature or environment, not elaborated by an authority control or that does not give any kind of protection.
Some of these lists point to places that already have official protection, although they present different ways of organizing information for promotional or communicative reasons (Ex.:P4154, P4762,..). Others simply contain lists of places of interest without any kind of protection (ex.:P3609, P5200,...).
Homogenize the definition, ontology[edit]
This was the initial focus of discussion that generates this analysis page.
The discussion is a proposal to include the natural heritage (Q386426) protections of the item within P1435, as is the case with cultural heritage (Q210272) protections.
Currently, properties of cultural and natural protection IDs appear along with the rest of the identifiers. But, unlike it happens with cultural heritage, there is no property with the "list of all natural protections" of the item. A list from which we can, if necessary, go to recover the specific properties with the IDs. This operation is planned and valid for P1435, which initially was only valid for cultural heritage, but was extended to be able to also accept the list of natural protections. (This point was already discussed earlier)
To summarize, it is proposed to adjust the definition and ontology of the properties of IDs of natural protection to the same ontology and functioning that already have those of cultural protection.
The specific changes proposed for each property are described in "Change proposed" of the comparison table.
- Reasons for change
- When we want to know what are all the protections (not the ID) of an item, we need to do one access for each protection ID property (30 by now), because we don't know a priori which of them the item could have. In addition, the creation of new protections means having to modify the recovery code to adapt it to the new elements.
- On the contrary, having all the protections (not the properties with the IDs) as values of the P1435 allows to recover them without having to find out what there is in each case. If we also want to know the ID of any of the protections, we can retrieve it with the Wikidata property (P1687) that the item of protection has. That is, an extension of the protection with new designations has no effect on the retrieving code.
- Additional, info related to the protection, for instance start time (P580) & end time (P582) must be qualifier of P1435 and it is not currently being collected at WD.
- This change will bring homogeneity to the structure of protection's statements. Not having a homogeneous pattern, the creation of ID properties has generated a variety of constraints that hinder the predictiveness of creation treatment as well as recovery.
See the Examples section with the changes proposed on ontology.
Roadmap to deploy changes[edit]
- Add a explanation of change in discussion property pointing to this page.
- Change the name of the authority by the name of the protection as Wikidata item of this property (P1629).
Done, for green and yellow class. - Change the property constraints
- Add the P1435 entry in all the items already created, when apply
- Identify inconsistencies between mandatory properties (defined as constraints). Try to fix them.
- Upload items of "green and yellow class" when < 50 entries now.
Upload natural heritage catalogs[edit]
- Rivers.gov protected area ID (P4190): https://www.rivers.gov/map.php + en:List of National Wild and Scenic Rivers
- ZNIEFF ID (P3498): https://inpn.mnhn.fr/synthese/statistiques-znieff
- BirdLife International IBA ID (P6070): http://datazone.birdlife.org/site/search
- Australian Wetlands Code (P2584): http://www.environment.gov.au/cgi-bin/wetlands/search.pl?smode=DOIW
- Australian Ramsar site ID (P2516): http://www.environment.gov.au/water/wetlands/australian-wetlands-database/australian-ramsar-wetlands
- IDA place ID (P4977): https://www.darksky.org/our-work/conservation/idsp/finder/
- Global Geoparks Network ID (former scheme) (P2467): http://www.globalgeopark.org/aboutGGN/list/index.htm
- Danish protected area ID (P2763): https://www.fredninger.dk/
Comparative property definitions[edit]
This analysis table has the 42 properties with
⟨ Wikidata property for authority control for protected areas (Q55978235) 
⟩and P5200 & P5215 defined as
It's the status by end of January 2019, before the change was done.
Colored column on the right side has the proposed changes.
Table[edit]
Notes[edit]
P31 considerations[edit]
Property P31 should show "what the item is", but sometimes shows "characteristics" of the item. In the table, we can see that several properties have information talking about the protection of the item in addition (or instead) of item's instance.
- It's correct ONLY when item is about "the protected area" itself. Ex.: Montseny Natural Park (Q22678731), that is the biosphere reserve (Q158454)
- It's wrong when the item is about and area/park/geographical object,... that HAS the protection. Ex.: Montseny Natural Park (Q1401508) is the physical area that grant the Q158454 protection.
When the P1435-based method we propose were deployed, the new way to indicate that there is a protection will make some of the constraints concerning P31 unnecessary.
P814 considerations[edit]
Property P814 doesn't run as other "protection ID properties". Instead of have the ID able to point to a unique entry with the information of one area, the P814 has the level/class of protection. So, the protection that item has is not literally IUCN protected areas category (P814), but IUCN category Ia: Strict Nature Reserve (Q14545608) or IUCN category II: National Park (Q14545628),... for instance.
In the proposal we consider that items may have as protection, either the generic concept of "protected area category" as well as any of the specific values defined by the property itself (see list in Q21510859 values of Property:P814#P2302).
In addition, as the P814 doesn't have the protection ID of item. Therefore, we'll use the correspondent WDPA ID (P809) where any protection has a unique code (see #P809 considerations)
P3425 considerations[edit]
Likely it happens in P814, the property P3425 has sub-category classification: Special Area of Conservation (Q1191622), Special Protection Area (Q2463705), site of community importance (Q796174) or proposed site of community importance (Q60534895) (see list in Q15069452#P527). Nowadays, the items just have the value of the information entry, but no the real class of protection assigned by the "Natura 2000" program. One item may have more than one of these protections, but only one ID for all of them, because the Natura 2000 database has one entry by area with all the protections granted by item.
P809 considerations[edit]
The WDPA ID (P809) property, as well as Common Database on Designated Areas ID (P4762), are two common databases for protected areas description and information, but they are not protection authority. P809 has an international scope and P4762 is just oriented to some EU protections. For instance, P809 includes Ramsar area and P4762, doesn't. The P4762 is, in fact, a subset of P809 and is one of the feeders of P809. Both use the same ID code (the SIDE CODE).
Unlike the P3425, that has one ID code by each physical area with all the natura 2000 protection in the same entry, the P809 has one entry (and one ID code) for each protection of each area. For instance, the concept "Yellowstone" has 7 entries, 3 of them are protections of the Yellowstone National Park (Q351) and other 4 are protections of specific parts within the park.
The present definition of P809 force to have distinct-values constraint (Q21502410), which is common on ID properties. It's a bad constrain for many items. However, although we could have more than one value at the P809, we would need a procedure to qualify each of them.
So, the proposal is do not use P809 as one more protection, but to include the correspondent WDPA ID as a qualifier of P1435 protection that have not their own ID, as happens in P814.
P1848 considerations[edit]
Initially protected areas INPN Code (P1848) is not a protection authority, but a database for protected area of France (Q2828309). However, it runs as the official place to handle the protection system of France natural heritage. Now, the related items have a P31 with the protection level and a P1848 with the id of protection, which allows to link to the entry in INPN.
The list of possible protections (see subclasses of protection area in France) should be those protections able to be used as a P1435 value. Probably it coincides with the P31 present value. However, it must be reviewed. Present values (12 feb 2019) of P31 are:
- geopark (Q1324355)
- regional nature reserve (Q15089606)
- regional natural park (Q1818761)
- national nature reserve (Q19656847)
- réserve naturelle volontaire (Q19698111)
- protected area of France (Q2828309)
- prefectoral decree for biotope protection (Q2864343)
- marine nature park (Q3364603)
- Corsican nature reserve (Q3457472)
- dark-sky preserve (Q3457162)
- biological reserve (Q445467)
- national park (Q943017)
Properties P3974, P5965 & P6230 considerations[edit]
P3974, P5965 & P6230 are ID for protect Naturschutzgebiet (Q759421) (nature reserve in Germany) for North Rhine-Westphalia (Q1198), Baden-Württemberg (Q985) & Bavaria (Q980) separately from the rest of Germany (Q183). There are no specific ID to identify nature reserve for the other lands of Germany, they used WDPA ID (P809) as ID.
Property P4029 considerations[edit]
The P4029's URL is not self-built by ID. It's hold on P4001. No other protection ID property has this ontology solution.
Examples[edit]
- Simple example with Pointe au Sel (Q3393570)
| Situation with present guidelines | Comments |
|---|---|
| present value on item | |
| mandatory by P1848 | |
| Not present in this case. It is mandatory by P809, but it's not necessary as already has protected area of France (Q2828309) (a subclass of Q473972). | |
⟨ Pointe au Sel (Q3393570) ⟩ Conservatoire du littoral ID (P3009) ⟨ 539/28-la-pointe-au-sel-974_reunion ⟩ |
present value on item |
| present value on item | |
| present value on item | |
| Not present in this case. It is mandatory by P809 |
| Situation after proposed guidelines | Comments | Chg |
|---|---|---|
| same as now | = | |
The Inventaire national du patrimoine naturel (Q3153864) in P1435 does the same function. |
– | |
| new. Before it was deducted by the existence of P3009 | + | |
| new. Replace the P814 = Q14545639 and use Q14545639 directly as a protection in the P1435. As IUCN categories do not have any IDs, but they have a specific entry within the WDPA container, we indicate this ID as a qualifier. |
+ | |
| new. Before it was deducted by the P31=Q2828309 | + | |
⟨ Pointe au Sel (Q3393570) ⟩ Conservatoire du littoral ID (P3009) ⟨ 539/28-la-pointe-au-sel-974_reunion ⟩ |
same as now | = |
Proposed for remove it, but It may be kept for compatibility. P809 is a container of protection descriptions with separate entries for each protection. It means we may have several IDs for the same "object/site/area", and we need to point the entry of WDPA from a specific protection described in P1435. |
– | |
| same as now | = | |
|
|
Proposed for remove it, but It may be kept for compatibility. The P814 is not necessary to define the class of UICN protection because it now appears in P1435. |
– |
- Complex example with Aigüestortes i Estany de Sant Maurici National Park (Q1470066)
| Situation with present guidelines | Comments |
|---|---|
| present value on item | |
| present value on item | |
| It is mandatory by P809. | |
| present value on item | |
| present value on item | |
| present value on item | |
| present value on item | |
| present value on item |
| Situation after proposed guidelines | Comments | Chg |
|---|---|---|
| same as now | = | |
| Proposed for remove it. Being National park, is already a park. |
– | |
| Proposed for remove it. | – | |
| Not necessary, but may be kept for compatibility. | – | |
| same as now | = | |
| Specific protection within Natura 2000 program | + | |
| Specific protection within Natura 2000 program | + | |
| Specific protection within Natura 2000 program | + | |
| value of IUCN category As IUCN has no ID, we use the correspondent P809 as qualifier, instead being property |
+ | |
| same as now | = | |
| same as now, but used by 3 protections. See P809 considerations | = |