Ultrascale Visualization of Climate Data

Fueled by exponential increases in the computational and storage capabilities of high-performance computing platforms, climate simulations are evolving toward higher numerical fidelity, complexity, volume, and dimensionality. To meet the challenges and exploit these opportunities, a consortium of four national laboratories, two universities, a government agency, and two private companies formed to explore the next wave in climate science. Working in close collaboration with domain experts, the Ultrascale Visualization Climate Data Analysis Tools (UV-CDAT) project aims to provide high-level solutions to a variety of climate data analysis and visualization problems. The project integrates computational and domain scientists from all sectors in tackling these problems, with the ultimate goal of building an ultrascale data analysis and visualization system. UV-CDAT brings to bear many capabilities that directly address climate scientists needs, including parallel streaming statistics, optimized parallel IO, remote interactive execution, workflow capabilities, and the automatic capture of data provenance. Already, the UV-CDAT architecture is enabling new capabilities in visualization, regridding, and statistical analysis.