Ontology: include terms covering the spatial regions of the scenario analysis space

[stap-m]

Description of the issue

territory, ~International transport~, Area outside of territory modelled, Upstream chains outside the territory included

Ideas of solution

If you already have ideas for the solution describe them here

Workflow checklist

• [ ] I discussed the issue with someone else than me before working on a solution
• [ ] I already read the latest version of the workflow for this repository
• [ ] The goal of this ontology is clear to me

I am aware that

• [ ] every entry in the ontology should have a definition
• [ ] classes should arise from concepts rather than from words

[Vera-IER]

I think it would be nice to have the terms

• model regions
• exogeneous regions
• exchange processes (import and export)

in the OEO to cover the terms in the description of the issue.

For international transport, I use the processes international aviation and navigation, but I think it would fit better to discuss them in a transport related discussion/issue.

[l-emele]

For international transport, I use the processes international aviation and navigation, but I think it would fit better to discuss them in a transport related discussion/issue.

I agree. Could you please create a separate issue?

[stap-m]

• model regions
• exogeneous regions
• exchange processes (import and export)

Thats a good starting point. Instead of model region we could call it sth like covered region to make it applicable not only for models but also e.g. for all kinds of data sets.
Any ideas for definitions?
The regions could be classified as 2-dim spatial region and the exchange processes as process of course.

[Vera-IER]

Okay, I just created a seperate issue for the international transport process.

This sounds a bit strange to me:
A covered region is a 2-dim spatial region...

Could we use the term mapped region ore endogeneous region instead?

[Vera-IER]

Additionally a label region code would be helpful, because we never use ful region names in the models.

[l-emele]

Additionally a label region code would be helpful, because we never use ful region names in the models.

In #461 (not yet part of the dev branch or even a pull request) I've introduced a unique individual identifier (_A unique individual identifier is an identifier that is unique for one individual of a class. Unique individual identifiers follow usually a structure defined e.g. by a sector division._) We deliberately labeled and defined this rather generically so it can be also used here for region labels.

[Vera-IER]

Okay perfect.

@stap-m can't we stay with the label model region? Because a data set application could use the term 2-dim spatial region for describing the regions of the data set.

[stap-m]

The def of 2-dim spatial region is _A two-dimensional spatial region is a spatial region that is of two dimensions. (axiom label in BFO2 Reference: [039-001])_. I don't think that really fits for the descripton of data sets.
It could be useful to use synonyms for this term.
Maybe it helps to agree on a definition first and discuss the label afterwards?!

[Vera-IER]

Ok, some suggestions for def.:
_A covered/model/endogeneous region is a 2-dim spatial region that is characterised with data which is averaged over the spatial dimension of the region and that is located within the system boundaries of the model/data set._
_A exogeneous region is a 2-dim spatioal region that is only characterised by exchange processes and that is located outside of the system boundaries of the model._
_A exchange process is a process in which two or more parties give something of value to each other to satisfy perceived needs._

The last defenition comes from here and other two from me.

[l-emele]

@christian-rli : Any feedback for this discussion?

[sfluegel05]

I agree with the definitions. Only a little detail: A covered/model/endogenous region is a 2-dim spatial region that is characterised _by_ data

Maybe we can solve the naming conflict by using endogenous region as a broader term and using model region and data set region as subclasses.
Possible definitions would be: _An endogenous region is a 2-dim spatial region that is characterised by data which is averaged over the spatial dimensions of the region and that is located within the boundaries of a certain system._
_A model region is an endogenous region defined by the boundaries of a model._
_A data set region is an endogenous region defined by the boundaries of a data set._

[Vera-IER]

Maybe we can solve the naming conflict by using endogenous region as a broader term and using model region and data set region as subclasses.

This proposal sounds good to me.

[christian-rli]

I also like the combination of endogenous region as a broad term with the different region subclasses.

I am wondering though. Is an endogenous region always two-dimensional? Maybe it usually is, but I could imagine a three dimensional research region. (This is setting aside the fact that when people describe two dimensional areas, they usually also mean not just the surface of an area on earth, but also the ground and air adjacent to it.) As theoretically it may take any dimension, maybe it should be a direct subclass of spatial region?

[christian-rli]

As a side-note: I wasn't super happy with the definitions for 2- and 3-dimensional spatial regions. "A two-dimensional spatial region is a spatial region that is of two dimensions" just doesn't seem very explanatory. I wanted to suggest changing it, until I saw that it was actually a BFO definition, so I guess the term can be considered rather self-explanatory.

[stap-m]

I'd actually prefer one generic term that doesn't need to distiguish between model, data sets, measurements ... ideally it should also include potentials, which always refer to a certain spatial region.
Does "endogenous" fit here? It's not the wording I'd use, but if it is the one we can agree upon, I can go with it. Btw, we can also use synomyns (e.g. modelled region, covered region) here.

[Vera-IER]

I like the idea with the synonyms and having one generic term to simplify it. With including potentials to regions do you mean it would be good to express somehting like this example:
Germany has a potential of free space PV of 80 GW.
or
Germany has a (land area) potential of free space PV of 200 000 ha?

[sfluegel05]

I can agree with endogenous region as one generic term. Does the current definition (_a spatial region that is characterised by data which is averaged over the spatial dimensions of the region and that is located within the boundaries of a certain system_) already include potentials?

To relate regions to potentials maybe we could use has disposition.

[Ludee]

Green
Black
Red
Blue

[stap-m]

Does the current definition (a spatial region that is characterised by data which is averaged over the spatial dimensions of the region and that is located within the boundaries of a certain system) already include potentials?

What about: _Endogenous region is a spatial region that covers the geographic dimensions of a calculation, data set, model or measurement._ It's not necessarily averaged data.

[Vera-IER]

these are the definition from yesterdays dev meeting (they describe the picture above):

red: A study region is a spatial region that is under investigation and consists entirely of one or more subregions.
(condition: red = sum(blue) )
potential alternative terms: region of interest

blue: A study subregion is a subregion of a study region.

green: A considered region region is a spatial region that is used in an analysis.
(condition: green = (red + black))

black: An interacting region is a spatial region that interacts with a study region. It is part of a considered region, but not a study region.
potential alternative terms: external region

[l-emele]

From OEO dev 9 meeting:

• _A subregion is a spatial region that is a part of spatial region._
• blue: _A study subregion is a subregion of a study region._ ( = subregion of a study region)
• green: _A considered region region is a spatial region that is used in an analysis._
• black: _An interacting region is a spatial region that interacts with a red region. It is part of a considered region, but not a study region._ Alternative term: external region
• red: _A study region is a spatial region that is under investigation and consists entirely of one or more subregions._

[stap-m]

Ah, I already wondered what that figure's supposed to mean. Thanks.
The distinction seems plausible. I have these questions/comments, though:

• I have the impression that the definitions of study region and considered region are not enough differentiated yet. Without the explanatory figure, I probably wouldn't understand the difference.
• Do we really need study subregion or is subregion enough? Accouding to the def of subregion it would be applicable to all kinds of spatial regions...

[Vera-IER]

Do we really need study subregion or is subregion enough? Accouding to the def of subregion it would be applicable to all kinds of spatial regions...

We thought about it too and agreed in the end that study subregion is a typical use case for models and should be included. Instead of having a class subregion we can use use "part of some spatial region".

[Vera-IER]

I have the impression that the definitions of study region and considered region are not enough differentiated yet. Without the explanatory figure, I probably wouldn't understand the difference.

Do we really need the concept of the considered region? I would prefer to leave it out in order to not confuse people.

[l-emele]

I have the impression that the definitions of study region and considered region are not enough differentiated yet. Without the explanatory figure, I probably wouldn't understand the difference.

Do we really need the concept of the considered region? I would prefer to leave it out in order to not confuse people.

We had a long discussion about this in the OEO dev meeting. We definitely need the described concept. But if the label is confusing we can try to find a better label.

[Vera-IER]

Okay, what about:
green: _An all-integrating region is a spatial region that is used in an analysis and combines the study region and its interacting regions._

[sfluegel05]

Okay, what about:
green: _An all-integrating region is a spatial region that is used in an analysis and combines the study region and its interacting regions._

I think that would definitely clear things up.

@Ludee Do you still want to implement this issue or should someone else do it?

[stap-m]

I'll implement to get it into the next release.
In a phone call with @l-emele, we had the idea to add examples of usage to explain the different classes more detailed.