project

Methods for decision-making under deep uncertainty

Progressing climate change, technological developments and population growth put our society and also the water sector in a situation of deep uncertainty. Deep uncertainty is a technical term for the unpredictability of complex systems beyond statistical uncertainty. This situation is common in the drinking water sector, but is still little recognized, acknowledged and treated as such. This project aims to provide BTO partners with a better understanding of the occurrence of deep uncertainty and what its relevance is, including prospects for dealing with deep uncertainty.

Deep uncertainty

Deep uncertainty is a technical term for the unpredictability of complex systems due to lack of agreement on assumptions. There is also no agreement on the most appropriate predictive model for such situations and (social) interactions influence the outcome. With deep uncertainty, there is no agreement on which models are useful and the possible outcomes van diverge widely. Thus, this is quite different from statistical uncertainty.

A good example of deep uncertainty is climate change. Model predictions indicate the possible range of average temperatures for the coming decades, but this is influenced by how much we manage to reduce our CO2 emissions. Moreover, models do not include all the physical processes that may be relevant in such a complex process as climate change.

Closer to daily water practices, predictions of what water demand will be in two decades are also deeply uncertain. This is due to a combination of unknown factors, such as population size, including possible (climate) migration and our behavior. But also because of how we can influence these factors. And especially by predictions of water availability.

Also in the drinking water industry

Deep uncertainty is common in the drinking water sector, but is still little recognized, acknowledged and dealt with as such. There are several techniques for dealing with deep uncertainty, such as Dynamic Adaptive (Policy) Pathways, Flexible Design (engineering options, real options) and Robust Decision Making. In this project, we map where deep uncertainty occurs in the drinking water sector and examine which (combinations of) techniques are most suitable for reaching decisions in these situations (decision making under deep uncertainty, DMDU). The focus is on common methods and the most recent developments in this field. We then apply a selected method in a case study to investigate and illustrate how it can work in a drinking water application.

Relevance and prospects

The aim of this project is to provide a better understanding of the occurrence of deep uncertainty in the drinking water sector and its relevance. An action perspective in relation to deep uncertainty is outlined and illustrated using a case study with a selected technique for decision making.

 

Image 1. Decision-making under deep uncertainty, according to Dall-E.