Integrating Problem Structuring Methods and Agent-Based modelling to deal with ambiguity: the case of the Glinščica river.

Abstract

In the Glinščica catchment the obstacle to agreement was not missing data but competing readings of the same data. Problem Structuring Methods are designed for exactly this situation: they make the framings held by different actors explicit and comparable before any quantitative model is built. This paper reports their integration with Agent-Based modelling, so that the qualitative structure elicited from stakeholders determines what the simulation represents and who acts in it.

The sequence runs from individual elicitation of problem understanding, through the construction of a shared causal structure, to an agent-based representation in which the decision-makers identified during the participatory phase behave according to the rules and priorities they themselves described. Ambiguity is not resolved by averaging the positions; it is carried into the model, where the consequences of each framing can be explored and compared.

The case is the Glinščica river in Slovenia, one of the nine NAIAD demonstration sites, where flooding, biodiversity loss, poor water quality and groundwater depletion overlap in a small peri-urban catchment.