Example data visualization of the temperature field (blue-red colors) and composition (green-yellow colors) in Earth's mantle. This calculation assimilates Earth's plate motion history for the past 410 million years.
Analysis of large 3D datasets is difficult on traditional desktop software, as the limited perspective often clutters the view or hides important detail. Previous experience has demonstrated that these visualization challenges are overcome with interactive and immersive virtual reality (VR) tools. Building off of the research done at the UC Davis KeckCAVES, the DataLab will extend the capability of our current 3D data visualization software 3D Visualizer to read and visualize emerging hierarchical mesh data structures such as those used by the community-developed ASPECT simulation code.
As our understanding of mantle-tectonic systems has grown, so too has the need for the increased model resolution these new data formats provide. With these new upgrades, users will be able to collaboratively explore geologic models to the sub-kilometer scale, allowing them to capture all relevant physical processes.
This project will focus on two components: 1) create a module inside 3D Visualizer to represent and visualize hierarchical meshes produced by the ASPECT code; and 2) create a reader to load ASPECT simulation results from files in HDF5 format. The HDF5 format is a container format optimized for high-performance parallel reads and writes. Reading ASPECT data from HDF5 files will decrease software resource requirements, and be more compatible with existing workflows. The standard format also means researchers won't have to re-format their existing data before loading them into 3D Visualizer.
Two sample data sets will be used to enable development and testing. The first data set is for a regional model of subduction in a Cartesian geometry. The second data is a global model of mantle convection in a spherical geometry (Example shown).