Do scenarios “have” uncertainty? Or scenarios “are” the uncertainty?

Yesterday I attended a webinar by the prospective LCA network on uncertainty in prospective LCA. Well I also organized that together with Stefano Cucurachi and Heath Logan from the network steering group. Reinout Heijungs (VU) and Pierre Jouannais (INRAE) were presenters.

Super nice. Minutes will be in the network archives and when I find them I might post the link here.

We didn’t talk much about prospective scenarios though. I mean those done using tools like Premise 1. Simplifying a lot: Premise takes projections from integrated assessment models and uses those to make “future” versions of the ecoinvent database. These can be used to assess the impact of a technology at a future point in time.

Me and others have been thinking about the uncertainty in this case.

What is uncertainty when we have these scenarios?

 

Based on this question found on LinkedIn

[…] there are so many SSP/RCP scenario combinations and varying IAM approaches (e.g., REMIND, IMAGE, TIAM-UCL) and choosing what scenario to choose is very confusing.

 

I am posting my answer here for documentation and later use, and reminder to myself.

In principle, the many scenarios represent the full-spectrum of the uncertainty about the future.

So, again in principle, one should run the simulation across ALL scenarios, because nobody knows which scenario will become reality (maybe none of them, or something in-between them).

If there was a probability associated with each scenario, one could random sample across scenarios to get a likelihood of results. And then investigate whether the LCA conclusions hold across the simulated scenarios, and which conditions lead to the best environmental performance. And then one could make decisions and take actions that ensure to make these conditions become true.

Because, in the end, what you need to do is to support a decision, right? So how is the scenario exercise functional to making decisions? Or, in other words, WHY are you actually running the scenarios?

The problem with background scenarios is that there is very little agency because a single technology developer will have little influence on how the electricity mix will look like in 2050…

 

  1. Sacchi, R., Terlouw, T., Siala, K., Dirnaichner, A., Bauer, C., Cox, B., Mutel, C., Daioglou, V., & Luderer, G. (2022). PRospective EnvironMental Impact asSEment (premise): A streamlined approach to producing databases for prospective life cycle assessment using integrated assessment models. Renewable and Sustainable Energy Reviews, 160, 112311. https://doi.org/10.1016/j.rser.2022.112311 ↩