Over 10 weeks this summer, collaborators in the Data to Discovery (D2D) program explored new ways to represent data from computational models of Earth's climate systems. At the end of the program, the team presented prototypes for interactive tools-named CliMAScope and CLOVE-which enable researchers to gain new scientific insights from their data while also serving as a springboard for artistic inquiry.
A Unique Collaboration
Now in its 12th year, D2D is organized by a team of faculty and leaders in data visualization across Caltech, JPL, and ArtCenter. Each year, the organizers select two to three proposals from Caltech and JPL research groups to engage in a collaboration. They then hire three to five student interns with backgrounds in design and programming to spearhead the visualization projects over the summer.
"What we look for are projects that we know have really interesting data with strong visual exploratory potential," says Santiago Lombeyda, senior computational scientist in Caltech's Center for Data-Driven Discovery and one of D2D's co-organizers. "We want to know that there's going to be a need for some type of tool and a community that will use it."
This year, the selected proposals both sought to represent complex data generated by models of Earth's climate systems. One proposal, from the group of JPL research scientist Anthony Bloom, focused on the carbon data model framework (CARDAMOM), which models the earth's carbon cycles. The other proposal, from researchers in the Climate Modeling Alliance (CliMA) led by Tapio Schneider, the Theodore Y. Wu Professor of Environmental Science and Engineering, and Alexis Renchon, associate research software engineer, focused on CliMA's Land model, which models vegetation processes and remote-sensing data.
A hallmark of the D2D program is the methodology the team uses to develop visual approaches. Rather than planning a specific tool or solution from the get-go, the interns and organizers work to understand the existing ways researchers interact with their data and then co-design, iterate, and test prospective solutions in close collaboration with the researchers.
"I'd liken it to speaking multiple languages with others who also speak multiple languages," says intern Linh Pham on the collaborative process between designers, programmers, and researchers. "Depending on how fluent we are in each language, there may be limits on how much we can communicate. However, when everyone has a good grasp of various disciplines, the ideas we can explore together become more than the sum of their parts."
This methodology, based in radical participatory design and human-computer interaction research, is more than just a way to ensure the end users are satisfied with the tools that are developed. It also opens the door for artistic inquiry into the process of creating and interpreting meaning in visual information.
"When we're ideating, I get most interested when we see something that isn't what we would expect," says Hillary Mushkin, Research Professor of Art and Design in Engineering and Applied Science and the Humanities and Social Sciences at Caltech and co-organizer of the D2D program. "There are certain kinds of diagrams that are very familiar, that are used often in data visualization because they're tried and true. But sometimes we end up hitting upon things that are weird and radically different."
