Raith, FelixBlecha, ChristianRink, KarstenWang, WenqingKolditz, OlafShao, HuaScheuermann, GerikDutta, Soumya and Feige, Kathrin and Rink, Karsten and Zeckzer, Dirk2020-05-242020-05-242020978-3-03868-115-1https://doi.org/10.2312/envirvis.20201093https://diglib.eg.org:443/handle/10.2312/envirvis20201093In the Underground Rock Laboratory Mont Terri, research has been conducted for over 20 years into the storage of radioactive waste in Opalinus Clay. The fitness for such storage depends on the prevailing geological material. Experiments and multiphysics simulations investigate the long-term changes in the Opalinus Clay. The resulting data are highly multivariate, and environmental scientists visually analyze the data using predefined color lookup tables. The fiber surfaces of Raith et al. offer the researchers a new approach for visual analysis. However, the existing algorithm for the calculation is subject to certain limitations due to special cases that lead to no or incomplete fiber surfaces. In this paper, we improve the fiber surface algorithm of Raith et al., which reduces numerical errors and accelerates the existing algorithm. This improvement also makes it possible that the interactor no longer needs to be closed and convex. We then use the Full-Scale Emplacement Experiment to show how the improved algorithm can help in the visual analysis of multivariate data.Attribution 4.0 International LicenseHuman centered computingVisual analyticsScientific visualization10.2312/envirvis.202010939-16