Show simple item record

dc.contributor.authorKappe, Christopheren_US
dc.contributor.authorBöttinger, Michaelen_US
dc.contributor.authorLeitte, Heikeen_US
dc.contributor.editorGleicher, Michael and Viola, Ivan and Leitte, Heikeen_US
dc.description.abstractIn order to gain probabilistic results, ensemble simulation techniques are increasingly applied in the weather and climate sciences (as well as in various other scientific disciplines). In many cases, however, only mean results or other abstracted quantities such as percentiles are used for further analyses and dissemination of the data. In this work, we aim at a more detailed visualization of the temporal development of the whole ensemble that takes the variability of all single members into account. We propose a visual analytics tool that allows an effective analysis process based on a hierarchical clustering of the time-dependent scalar fields. The system includes a flow chart that shows the ensemble members' cluster affiliation over time, reflecting the whole cluster hierarchy. The latter one can be dynamically explored using a visualization derived from a dendrogram. As an aid in linking the different views, we have developed an adaptive coloring scheme that takes into account cluster similarity and the containment relationships. Finally, standard visualizations of the involved field data (cluster means, ground truth data, etc.) are also incorporated. We include results of our work on real-world datasets to showcase the utility of our approach.en_US
dc.publisherThe Eurographics Association and John Wiley & Sons Ltd.en_US
dc.titleAnalysis of Decadal Climate Predictions with User-guided Hierarchical Ensemble Clusteringen_US
dc.description.seriesinformationComputer Graphics Forum
dc.description.sectionheadersSpatial Data Applications

Files in this item


This item appears in the following Collection(s)

  • 38-Issue 3
    EuroVis 2019 - Conference Proceedings

Show simple item record