AXICCS 2016 Workshop Cultivates Big Ideas in Computational Climate Science
To explore new directions in climate modeling and simulation, a forward-thinking workshop on “Advancing Cross-Cutting Ideas for Computational Climate Science” (known as AXICCS) invited the scientific and research community to address future computational challenges in climate science. Climate scientists, computer scientists, and applied mathematicians joined in a community-driven effort to advance global climate models through cross-disciplinary collaboration.
Sponsored by the U.S. Department of Energy’s Office of Biological and Environmental Research (BER) and Office of Advanced Scientific Computing Research (ASCR), the workshop convened on September 12-13, 2016, in Rockville, Maryland. Kate Evans, Oak Ridge National Laboratory (ORNL), and Esmond Ng, Lawrence Berkeley National Laboratory, organized the event.
“Our goal was to move beyond the next experiment and model update, think more deeply about what we’ll want our models to be able to do in the next decade, and strategize on how we should best engage computer scientists and mathematicians to help us get there,” said Kate Evans, group leader for Computational Earth Sciences and member of the Climate Change Science Institute at ORNL
Earth system models (ESMs) that can account for the full range of dynamic interactions among the physical, chemical, and biological processes that shape climate are essential tools to improve understanding of global and local climate changes and predict future impacts. However, the increasing performance expectations for multiscale, global, coupled ESMs—to integrate more processes, improve response time, and refine spatial and temporal resolution—result in significant computational challenges for current petascale and future exascale systems.
Responding to this charge, AXICCS engaged experts from national laboratories, industry, and research institutions in a unique brainstorming environment. Workshop participants were asked to submit “ideas” papers to pursue future developments in research and modeling capabilities. The workshop’s program drew from the inspiration and enthusiasm of these ideas, fueling a dialogue for discovery. Plenary talks and panel discussions were structured around collaboration. Climate scientists presented on the current state of coupled climate models and the critical bottlenecks for increased complexity in ESMs. Applied mathematicians and computer scientists responded, opening a cross-disciplinary exchange of solutions and strategies to meet the expanding computational requirements in climate science.
A final report on the 2016 AXICCS workshop (available here) summarizes the outcome of the proceedings. Contributed papers are included in the appendices.
By Ashley Huff