Uncertainty visualization: why might it fail?

There is a gulf between the rhetoric in visualization about the importance of uncertainty, and the practice of visualization in which uncertainty is rarely seen other than as a laboratory exercise. We reflect on why something viewed as fundamental in science and engineering is rarely if ever adopted in visualization practice. Our analysis is informed both by research progress and by our own experience in an ongoing industrial case study on modelling and mapping underground assets, where it would appear that uncertainty plays a major role. In this case study, we try to identify promoting and limiting factors. We conclude that the value of uncertainty visualization is severely limited by the quality and scope of uncertainty data, by the limited confidence in the data itself, and by the perceptual and cognitive confusion that the depiction of this data can generate. We hope to broaden the discussion on the utility of uncertainty in visualization from the purely technical and perceptual issues to social and organizational factors.