Ling, MengJohnson, Jeffrey A.Feng, ZhiquanChen, JianKarsten Rink and Dirk Zeckzer and Roxana Bujack and Stefan Jänicke2018-06-022018-06-022018978-3-03868-063-5https://doi.org/10.2312/envirvis.20181136https://diglib.eg.org:443/handle/10.2312/envirvis20181136Exploratory data visualization, an idea proposed by Tukey in 1977, is meant to output various types of visualization in order to make the data more understandable. While visualization researchers have generated many novel solutions to critical and complex environmental problems, to everyday environmental consultants, some practical considerations have to be made in the visualization analysis to help stakeholders generate and test hypotheses that would not be possible otherwise. We present two environmental case studies of using visualization to communicate key findings: constructing stratigraphic units (layered) and generating groundwater contaminant plumes (volumetric). These real-world cases show that many times visualization alone may not give us correct answers; often what works is the combination of visualization, domain experts' knowledge, and interpretation of the visualization solutions. The lack of any of them may lead to faulty conclusions. The first case study illustrates together how domain experts, visualization, and contamination conditions assisted in interpreting limited and ambiguous lithologic data. The second case study emphasizes conceptual and technical understanding and discusses some common factors affecting 3D interpolation, which again suggests that we must incorporate domain experts' knowledge as well as analytics into visualization for defensible decision making.10.2312/envirvis.2018113641-46