Abstract
Computational models are the predominant instruments for studying water-related phenomena in hydraulic engineering. Over time, these models have become more and more opaque, making it more difficult for modelers to grasp their functioning. As a result, both model developers and users straddle discovery and manipulation, since they may not be able or willing to reflect on how computational models shape their understanding of the world. Hydraulic engineers often engage models in a reflective manner that is aimed at understanding a model’s underlying design. However, models in the form of software travel easily to domains outside of hydraulic engineering, where they are not used in a reflective manner. In order to prevent that model users become subject to the will-o’-the-wisp of opaque computational models, the aforementioned reflective approach to modeling warrants adoption by model developers as well as model users.