NTR: 'carbon emission'

[mark-jensen]

SDGIO has requested this term for work on indicators. It is appears in 9.4.1

Carbon emission per unit of value added

ENVO has a class carbon-bearing gas emission process, but I believe the indicator refers to the output of the process. It is possible that the term 'carbon emission' used in the indicator may refer to a proportion, concentration, or other calculated value. Even if it does, that value will ultimately refer to the output gas, not the process.

carbon emission can be a defined class:

environmental material and (is_output_of some carbon-bearing gas emission process)

[pbuttigieg]

It seems to be the volume of carbon emitted. Strange class, but as we have the process branch we can express it.

environmental material and (is_output_of some carbon-bearing gas emission process)

I think environmental material has to be constrained to carbon (or possibly a list of gases like carbon dioxide and methane) for this to make sense. The carbon-bearing gas emission process could have outputs that have nothing to do with carbon efficiency.

[cmungall]

Note: 'is output of' is not a functional relation. This means that a query for material-and-has-part-some-carbon will not return carbon emission unless seemingly redundant axioms are added. Formally this is because the axioms do not exclude models in which CBGEP has multiple outputs, with one of these being carbon-containing, and the other not.

[pbuttigieg]

Thanks @cmungall
Also, aren't most carbon molecules on the surface of the planet emitted at some point?
I think what the SDGs want should be better defined - are we talking only about anthropogenic processes that emit carbon? Or do we include carbon emissions by ecosystems such as rice paddies and rainforests? Do we differentiate between types of carbon-based molecule emitted? Methane is considerably more damaging than carbon dioxide in most settings.

[cmungall]

Yes I think we really want to examine how this class will be used. Can someone point to a definition of the relevant indicator?

Is the use cases about net emissions or gross?

[mark-jensen]

Just found out now, that the indicator 9.4.1 has changed:

Originally:

Carbon emission per unit of value added

Now :

CO2 emission per unit of value added

[mark-jensen]

Somewhat helpful doc for understanding "per unit value".

And this from World Bank, very relevant, and perhaps with energy intensity too.

Remember the goals here is to give UNEP a term relevant to the indicator for tagging purposes, not completely model the indicator.

[pbuttigieg]

From the World Bank definition:

Carbon dioxide emissions are those stemming from the burning of fossil fuels and the manufacture of cement. They include carbon dioxide produced during consumption of solid, liquid, and gas fuels and gas flaring.

So should we specify this narrow range of processes for the initial indicator? We can revise as we get requests.

• fossil fuel burning process
• cement manufacturing process
  Both would have output of carbon dioxide. SDGIO should probably have the class that represents carbon dioxide generated by these two processes, as this is quite specific.

[cmungall]

Also a def of 9.4.1 from:
http://unstats.un.org/sdgs/files/metadata-compilation/Metadata-Goal-9.pdf

"CO2 emission per unit of value added is a ratio indicator between the carbon emission and value added. Carbon emission is estimated from the data on energy consumption"

"value added" is outside the scope of ENVO, but thanks for the useful links @mark-jensen it helps us understand the context (although we really need input from the statisticians, environmental accountants and so on, otherwise we will end up precisely defining something that is very different to what they want).

My simplistic view is that the indicator is about tons of CO2 (or CO2e, as per @pbuttigieg's comment) released by the various commercial and industrial processes within a country (I have no idea what I mean by 'within' here. For example, all the CO2 from ships registered under a Liberian flag of convenience? But that's an SDGIO problem not an ENVO one).

I think a good ENVO class would be CO2 amount, in the envo-p hierarchy. We can also have CO2e amount (related syn: greenhouse gas level). These would be pure physical quantities, measured in units like tonnes. We could have parallel classes for level which would be ratios (ppm etc). These could be measured in the whole atmosphere, or in portions.

These classes can then be conveniently used in expressions, e..g CO2 amount produced-by some (process and occurs_in some thing). SDGIO would use the relevant expression.

[pbuttigieg]

I think a good ENVO class would be CO2 amount, in the envo-p hierarchy.

Perhaps ENVO environmental material is more natural, but otherwise I agree for amounts. Levels would be more a ratio issue (ppm)

+1 for more interaction with environmental accountants or UNStats.

[pbuttigieg]

I think a good ENVO class would be CO2 amount, in the envo-p hierarchy.

Perhaps ENVO environmental material is more natural, but otherwise I agree for amounts. Levels would be more a ratio issue (ppm)

Note: @cmungall suggests this as a quality rather than a material.

[pbuttigieg]

Today's release should have the classes you need to start.

@cmungall
"CO2 amount" as a quality has yet to be explored. I'm unfamiliar with this representation, so will need a bit more information on creating it. Right now, we have "carbon dioxide emission" as a material with some axioms linking it to a "carbon dioxide emission process" (to be populated by inference), CHEBI:'carbon dioxide' and PATO:'quality of a gas'.

[mark-jensen]

@pbuttigieg Ok to close this issue now?

[pbuttigieg]

@pbuttigieg Ok to close this issue now?

@cmungall - do you think we should add the quality of 'carbon dioxide amount' and get that pattern established before closing?

[cmungall]

+1

[pbuttigieg]

Great, perhaps we can discuss this tomorrow and line up a new release.