ausblenden:
Schlagwörter:
-
Zusammenfassung:
We increasingly rely on complex environmental models to understand the functioning and the fate of our planet. Models now inform policy makers and the general public, and therefore have impacts way beyond the scientific realm. Yet, while our models have grown in complexity, scale and societal relevance, there has not been a concurrent development of model evaluation. For example, model evaluation still commonly relies on comparison with historical data at a few geographically biased sites, which is insufficient if models are used for scenario exploration.
In this contribution, we want to reflect on the evolution of model evaluation and discuss the requirements that evaluation strategies for complex environmental models should meet. We do so by presenting an alternative strategy applied to global water models, namely evaluation using functional relationships between model forcings and model outputs. We further discuss the more general need for a “theory of evaluation” by looking at approaches from related disciplines, such as the use of emergent constraints in climate modelling or qualitative and participatory approaches to evaluation in sustainability science.
