One of the fundamental challenges in the study of past climate change in Earth history is how to reconstruct and compare past ocean circulation (as derived from sparse geochemical data collected from fossils in deep-sea sedimentary cores) with modern ocean circulation (as constructed from modern oceanographic observations and computer simulations) to yield insight into the differences and processes governing ocean circulation throughout the last glacial cycle from 150,000 years ago to the present. Similarly, a major challenge in computer science is how to extract information from sparse datasets, and how to effectively combine computational thinking with automated data analysis, to extract new knowledge about features and processes. Our multidisciplinary and multi-institutional project brings together a team of computer scientists, physical oceanographers, paleoceanographers, and computational geophysicists to address these challenges. We are developing an innovative suite of computational tools to explore past changes in global ocean circulation by allowing researchers to interact with data in a 3-dimensional environment across a geologically relevant time domain (the 4th dimension). This project will merge analysis of ocean flow with 40 years of previously-collected geochemical data from deep-sea sedimentary cores in order to gain new insights into past ocean circulation change. Our research will take advantage of the unique analytical resources and interdisciplinary collaborative environment provided by the UC Davis KeckCAVES (W.M. Keck Center for Active Visualization in the Earth Sciences). The KeckCAVES provides immersive interactive visualization technologies that help scientists identify meaningful patterns in complex datasets. In this unique collaborative environment, we are developing methods to improve data interpolation, to extract ocean flow patterns, and to characterize ocean flow over time for potential change detection, along with interactive means of visualizing and interacting with those large and time-dependent datasets. The computational methods developed in this project are adaptable to a wide range of visualization and data-analysis applications for analysis of flow fields. The scientific visualization methods we develop will be used to develop innovative, provocative, and intuitive instructional materials for undergraduate and graduate education; for making other researchers familiar with new scientific methodology (visually-aided analysis); for bringing science and scientific results to the public; and for bringing interactive scientific visualization technology to all of the partner institutions in this collaborative project.
