The Northern high-latitude (NHL) ecosystems above permafrost are considered to be most vulnerable to climate change. With permafrost warming, the dynamics of carbon fluxes in these regions is likely to have tremendous impacts for the future global climate through both positive and negative feedbacks involving physical and ecological systems, and through human dimensions such as land use change. In combination, these changes are thought to imply a strong net positive feedback to increased climate warming through increased CO₂ emissions, decreased albedo, and intensification of the hydrology cycle. However, our current understanding of the full suite of processes and their responses to recent warming in terrestrial NHLs are far from complete. While continued research on development of more detailed earth system models (ESMs) are essential to understand these interactions and feedbacks, major challenges are the treatment of these biogeophysical and biogeochemical processes in ESMs. While proper inclusion of these feedback mechanisms is likely to significantly improve the accuracy of model-based projections of future climate responses to greenhouse gas forcings, the complex interactive processes will make it difficult to rapidly integrate and test new algorithms and mechanisms required to improve our prediction of climate change using more detailed ESMs.

This proposal responds to these combined scientific and computational challenges in climate modeling by expanding our understanding of the biogeophysical-biogeochemical processes and their interactions in the NHLs using an ESM approach, and by adopting an adaptive parallel runtime system in an ESM to achieve efficient and scalable climate simulations through improved load balancing algorithms. We propose to address these challenges through two integrated areas:

• Extending the capabilities of the Integrated Science Assessment Model (ISAM), a process-based land surface model, by coupling with NCAR's Community Earth System Mode 1 (CESM), and using the coupled ISAM-CESM framework to evaluate key interactions amongst Earth's climate and terrestrial processes in the NHL regions.
• Applying novel load-balancing mechanisms to the ISAM land surface model, and subsequently extending these load-balancing capabilities into other components in the CESM1 framework (atmosphere, ocean and sea ice components) to achieve dynamic and improved modeling scalability.

To address the climate science and computer science challenges, this project synergistically combines major strengths from two research groups at Illinois. From the climate science side, the Atmospheric Science specialists has undertaken a series of coupled ISAM-CESM modeling studies to better understand how the interactions among the climate, biogeochemical and biogeophysical systems can quicken or slow the pace of climate change. From another side, the Computer Science specialists have been developing the Charm++ adaptive runtime system, which has empowered various scientific applications to scale to thousands of processors, in particular through the use of dynamic load balancing. This combined effort has the required ingredients to significantly accelerate the execution of state-of-the-art climate simulations on modern supercomputers. We will evaluate the ESM performance and the computational techniques derived in our investigation on large systems, including the Blue Waters machine, which will be deployed at Illinois in late 2011. One of the goals of this proposal is to engender a broader collaboration between computer scientists and earth science scientists, and train young scientists to become the next generation of researchers in this important field. The PIs will be working with The National Center for Supercomputing Applications at University of Illinois at Urbana-Champaign on a broad theme to bridging the gap between the atmospheric and computer scientists and technologists, to create cross-disciplinary opportunities for graduate students from multiple domains to build suitable tools graduate students from multiple domains to build suitable tools.