add `robust`

[stap-m]

Description of the issue

From SLE project meeting 2023-09-18 and #1643

usage:

1. "robust scenario", "robust transformation pathway" or "robust inventment decision"
2. robust with restpect to a set of solutions of modelled transformation pathways
3. robust set of technologies that is met in every scenario variation

Ideas of solution

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

[stap-m]

First idea: make robust a role or other kind of realizable entity that can be inhered by a scenario, maybe others

• similar to reference role
• classification as role because it is dependent on the interpretation of the analyst whether a scenario (or else) is robust, or not

[l-emele]

Do you have an idea for a definition? Robust can have very different meanings: https://en.wiktionary.org/wiki/robust#Adjective

[stap-m]

We discussed this in the SLE project meeting and it was difficult to find any consensus on the meaning of robustness. With respect to the wiktionary link and the above mentioned use cases, I find these two contexts helpfull:

• (systems engineering) Designed or evolved in such a way as to be resistant to total failure despite partial damage: in our case the modelled energy system
• (statistics) Not greatly influenced by errors in assumptions about the distribution of sample errors.: in our case the scenario projections / model outputs

Referring to the above mentioned use cases:

1. "robust scenario", "robust transformation pathway" or "robust inventment decision"

refers to system robustness for explorative scenarios

2. robust with restpect to a set of solutions of modelled transformation pathways

refers to statistical robustness for a set of scenario projections / model runs / sensitivity analysis

3. robust set of technologies that is met in every scenario variation

refers to system robustness for a target driven scenario

@chrwm you were involved in the discussion, too. Would you agree?

[l-emele]

Reading this issue again, it seems not a good idea to try to find a general definition for robust but rather define specific robust classes which to not refer to a general robust role or similar. Defining only the specific classes like robust scenario seems to be the way to go here.

[stap-m]

I start to make the proposals more concrete:

• system robustness: system robustness is a disposition of a system that is designed or evolved in such a way as to be resistant to total failure despite partial damage.
• robust system scenario:
  • robust system scenario is an scenario that describes an energy system that is designed to evolve in a robust way.
  • 'explorative scenario' and ('is about' some ('energy system' and 'has disposition' some 'system robustness')
  • alternative labels: robust transformation pathway or robust inventment decision
• statistical robustness: statistical robustness is a disposition of a scenario that is realized in scenario projections, model calculations or sensitivity analyses resulting in output data that is not greatly influenced by errors in assumptions about the distribution of sample errors.
• statistically robust scenario: statistically robust scenario is a scenario that is the basis for statistically robust projection data.

Espedially with the statistical robustness I am not yet happy, but maybe it is a start...

[stap-m]

some notes from todays discussion with SLE modelling experts:

• statistical robustness: besser "uncertainties in assumptions" statt errors
• system robustness: resistant also to external changes
• resilience vs (system) robustness?
• robustness always with respect to certain aspects

[stap-m]

I make some changes in my proposal based on the comments from the SLE meeting:

• system robustness: system robustness is a disposition of a system that is designed or evolved in such a way as to be resistant to total failure despite partial damage or external changes.
• robust system scenario:
  • robust system scenario is an scenario that describes an energy system that is designed to evolve in a robust way.
  • 'explorative scenario' and ('is about' some ('energy system' and 'has disposition' some 'system robustness')
  • alternative labels: robust transformation pathway or robust inventment decision
• statistical robustness: statistical robustness is a disposition of a scenario that is realized in scenario projections, model calculations or sensitivity analyses resulting in output data that is not greatly influenced by uncertainties in assumptions about the distribution of sample errors with respect to a (couple of) specific parameter(s).
• statistically robust scenario: statistically robust scenario is a scenario that is the basis for statistically robust projection data.

• robustness always with respect to certain aspects

I only added the to statistical robustness: with respect to a (couple of) specific parameter(s). I think here, it fits well.

[stap-m]

Any comments to my proposal? @l-emele @chrwm and others?

[l-emele]

I am fine with these proposal.