Describe an emission calculation ontologically

[l-emele]

Description of the issue

An important part of greenhouse gas statistics compilation and model calculations is the (sub-)process of calculating emissions. Currently, we do not depict this in the OEO, but I think we should.

An important concept for calculating both historic and projected greenhouse gas emissions is the so called activity data. This is an abstract concept and could be many different things.

The IPCC describes this calculation process and the involving other concept as:

As with the 1996 Guidelines and IPCC Good Practice Guidance the most common simple methodological approach is to combine information on the extent to which a human activity takes place (called activity data or AD) with coefficients which quantify the emissions or removals per unit activity. These are called emission factors (EF). The basic equation is therefore:
Emissions = AD • EF
For example, in the energy sector fuel consumption would constitute activity data, and mass of carbon dioxide
emitted per unit of fuel consumed would be an emission factor. The basic equation can in some circumstances be
modified to include other estimation parameters than emission factors. Where time lags are involved, due for
example to the time it takes for material to decompose in a landfill or leakage of refrigerants from cooling
devices, other methods are provided, for example first order decay methods. The 2006 Guidelines also allow for
more complex modelling approaches, particularly at higher tiers.

Source: IPCC, 2006 IPCC Guidelines for National Greenhouse Gas Inventories, Volume 1, Chapter 1.3

So there is not the emission calculation method, but a family of different methods, which can be depicted as class with potential subclasses.

Ideas of solution

Add classes:

• emission calculation:
  • Definition: An emission calculation is a process that applies an emission calculation method to calculate emissions.
  • Axiom: 'emission calculation' SubClassOf: 'has participant' some 'emission calculation method'
• emission calculation method:
  • Definition: An emission calculation method is a methodology that describes how to calculate emissions.
  • Axiom: 'emission calculation method' SubClassOf: 'is about' some 'emission calculation'
• basic emission calculation method:
  • Definition: A basic emission calculation algorithm is a emission calculation method that uses only activity data and emission factors.
  • Axiom: 'basic emission calculation method' SubClassOf: uses some 'activity data' and uses some 'emission factor'
  • Annotation mathematical expression: emission value = activity data \times emission factor
• activity data role:
  • Definition: An activity data role is a role of a X that describes human activity.
  • Axiom: 'activity data role' SubClassOf: 'has bearer' some X
• activity data:
  • Definition: Activity data is an X that has an activity data role.
  • Axiom: 'activity data' EquivalentTo: X and ('has role' some 'activity data role')

To be decided whether X is a data item or a quantity value or both.

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

[han-f]

I agree with the suggestions, but would opt for a change here:

emission calculation method:
Definition: An emission calculation method is a method that describes how to calculate emissions.

On a more broader scale: is activity in all the emission calculation contexts always necessarily a human activity?

[l-emele]

emission calculation method:
Definition: An emission calculation method is a method that describes how to calculate emissions.

From a linguistic point I would agree that method seems more appropriate. I open a separate issue to discuss this.

On a more broader scale: is activity in all the emission calculation contexts always necessarily a human activity?

If you think of CO2 emission from volcanos there the activity is non-human, but instead volcanic eruptions per year or something like that. However, I think non-human influenced emissions are beyond the scope of the OEO.

[l-emele]

@stap-m: Do you have any thoughts on this issue? (Despite the method vs. methodology that is discussed in #1787.)

[stap-m]

Can emission calculation be understood as subclass of model calculation? I guess there is kind of a numerical model used?

I don't really understand the activity data (role) classes. Can you explain that further and give an example?

[l-emele]

Can emission calculation be understood as subclass of model calculation?

That depends on how exactly we understand model calculation. That is currently defined as: A model calculation is a process of solving mathematical equations of a model. However, there is also the axiom 'model calculation' 'has participant' some 'computer model'. This is inconsistent. There are two options how to solve this inconsistency and this affects how to define emission calculcation.

• If we agree to the definition of model calculation, then yes, emission calculation could be defined as: An emission calculation is a process model calculation that applies an emission calculation method to calculate emissions. But then we have to change the existing axiom to 'model calculation' 'has participant' some 'mathematical model'
• If we stick to the axiom 'model calculation' 'has participant' some 'computer model' then we should redefined model calculation to: A model calculation is a process of solving mathematical equations of a numerical computer model. In this case emission calculation cannot be a subclass of model calculation as as not necessarily a computer is involved.¹

I don't really understand the activity data (role) classes. Can you explain that further and give an example?

• When calculating the CO2 emission from coal combustion, then the activity data is the amount of coal burnt.
• When calculating the fugitive methane emissions from the natural gas grid, then the the activity data can be e.g. the amount of gas transported or the length of the gas grid.
• When calculating the methane emissions from enteric fermentation of cows, then the activity data is the count of cows.²

So, there are different types of data (often time series) that are combined with emission factors to calculate emission data. If such data is used, then it forms the AD of that simple equation Emissions = AD • EF. That is why I thought of introducing the activity data role (or function?). It would be very helpful to have activity data and activity data role: Using these, we can better annotate data on the OEP, e.g. describing that data from a specific column was used as activity data for some emission data provided in another column.

Footnotes

1. I often do simple emission calculations with pen and paper or even in my head. ↩

2. Cows are considered human activity as they are used to produce milk and meat. ↩

[l-emele]

This issue is stale for a while. I have a use case where I would reference the OEO and the following two concepts are important for that:

• activity data role:
  • Definition: An activity data role is a role of a X that describes human activity.
  • Axiom: 'activity data role' SubClassOf: 'has bearer' some X
• activity data:
  • Definition: Activity data is an X that has an activity data role.
  • Axiom: 'activity data' EquivalentTo: X and ('has role' some 'activity data role')

@stap-m : Did the explanation in my last comment resolve your questions? If so, do you agree on my definition proposals?