(In?)extricable links between data and visualization: preliminary results from the VISTAS project

Our initial survey of visualization tools for environmental science applications iden-tified sophisticated tools such as The Visualization and Analysis Platform for Ocean, Atmosphere, and Solar Researchers (VAPOR) [http://www.vapor.ucar.edu], and Man computer Interactive Data Access System (McIDAS)andThe Integrated Data Viewer (IDV) [http://www.unidata.ucar.edu/software]. A second survey of ours (32,279 figures in 1,298 articles published between July and December 2011 in 9 environmental science (ES) journals) suggests a gap between extant visualization tools and what scientists actually use; the vast majority of published ES visualizations are statistical graphs, presenting evidence to colleagues in respective subdisciplines. Based on informal, qualitative interviews with collaborators, and communication with scientists at conferences such as AGU and ESA, we hypothesize that visualizations of natural phenomena that differ significantly from what we found in the journals would positively impact scientists' ability to tune models, intuit testable hypotheses, and communicate results. If using more sophisticated visualizations is potentially so desirable, why don't environmental scientists use the available tools?