The Visualization for Terrestrial and Aquatic Systems project helps environmental scientists produce visualizations for their own research and for presentation to other scientists and stakeholders including decision makers. A critical finding of work to date is the extent to which scientists use visualizations not only to explore data in new ways and present results, but also to work with stakeholders to jointly produce information that can be used during decision-making processes. The role of scientific visualization in the co-production of knowledge is as yet untested, even though this involvement could be critical in creating acceptable solutions, information, or technology. This proposal recasts VISTAS to co-produce visualization tools, i.e., exploring how negotiations between the users? needs and technological capacity shape the type of visualizations used and tools implemented, change or modify the research questions posed by scientists, and impact how results are interpreted so communities can respond to critical ecological challenges, including climate change. This is a unique experiment and collaboration among social-, computer-, and environmental scientists, with non-scientist stakeholders, to co-produce data visualizations for use in decision making. Social science methods will be used to explore knowledge co-production coupled with technology innovations that lead to community decision making to solve problems of climate change adaptation. The extent to which distinctions between scientific visualization for scientists and non-scientists need to be made will be determined, and unique visualizations will be developed jointly with project collaborators. The goal is to determine the influence of visualization on the co-production of knowledge among scientists and stakeholders on critical decisions related to climate adaptation. This project involves both computer scientists and social scientists. Computer science: VISTAS, a C++ scientific visualization application with significant GPU processing, helps environmental scientists produce images that allow them to ?see? the effects of topography on ecological phenomena. For this award, new visualization techniques will be developed, visualization and visual analytics research that enables effective presentations to decision makers will be conducted, and technical support for environmental- and social scientists will be provided. If time and funds permit extensions to the current software that render it both more usable by primary and secondary users, and more maintainable and extensible directly by primary users will be provided: VISTAS engineers will proceed with a longer term strategy of migrating from C++ to Python, which will enable more effective and flexible user interface development, end user programming of data or visualization plug-ins, and use of emerging and existing Python and R libraries for visual analytics. The social science inquiry will help determine how the co-production enables usable software that answers the needs of both environmental scientists who generate large difficult to interpret data sets as well as decision-makers who must balance multiple demands as they make important choices. Case studies with three collaborators will be conducted as they work with stakeholders to co-develop usable information; these are structured through a comparative pre/post-test design with three phases to explore changes in how participants view and communicate scientific results before and after involvement in visualization development. In the baseline phase VISTAS social scientists will work with participants to document their current understanding of their data, expectations for the visualization and analytic products, and ability and tools used to communicate science to others including non-scientists. During the development phase case participants will be observed as they work together to create the visualization and analytic products. The post-assessment phase seeks to determine changes in understanding of data and ability to communicate science as a result of participation in visualization development. The usability of different types of visualizations and analytic tools, identifying the characteristics that contribute to or distract from usefulness, will also be explored. Information will be collected primarily through semi-structured interviews with participants (collaborators and stakeholders). Existing scales measuring environmental attitudes and preferences for science in decision-making and general attitudes toward science will be used so comparisons with larger national and international samples can be made. In addition, scoping and development meetings will be observed to determine how shared understanding of user needs is developed and then framed as a visualization problem.