Suggested Questions for Breakout #1

Owned by Renata McCoy
Oct 22, 2020
2 min read

Atmosphere, Ocean and Land breakout sessions:

  1. What are the grand challenges in atmosphere, ocean/ice, and land modeling? For example:

    1. What are the key missing or more uncertain processes that should be prioritized for model development?

    2. What processes require more improvements in scale awareness?

    3. Is there a game-changing scale for modeling processes in each component?

  2. How can we overcome these challenges and accelerate progress?

  3. What opportunities/recent advances can E3SM leverage?

  4. How can we improve E3SM’s development and evaluation process? For example:

    1. How could development better target known biases?

    2. How can we ensure good coupled model behavior while developing component models?

    3. How can development and evaluation be made more efficient?

Computation breakout session:

  1. What are the grand challenges in earth system modeling from a computational perspective?

  2. What strategies can be implemented to accelerate progress?

  3. What are the opportunities/recent advances in algorithms, performance optimization, programming models and software engineering can E3SM leverage?

  4. What 5-year goals should be prioritized for computational science effort?

    1. Can we achieve universal convergence with time step and vertical levels?

    2. Can we make verification more expected and uniform?

    3. Are any improvements needed to DevOps (git workflow, build system, test coverage, test frequency, non-BFB, dealing with test failures), or incremental maintenance is sufficient?

    4. What are opportunities for new algorithmic efforts?

    5. How can computational science experts contribute to large ensemble modeling, infrastructure, algorithms, diagnostics?

ML/AI breakout session:

  1.  What are the challenges in using ML/AI for parameterization development? What are some opportunities to address these challenges?

  2. What areas of human-earth interactions may make use of ML/AI for significant advancement?

  3. How may ML/AI be used to address or quantify uncertainty in model simulations and projections?

  4. How may ML/AI and physically-based models be used in combination to improve model fidelity or design modeling experiments?

  5. What other areas can E3SM development, simulation and analysis, computational performance and infrastructure benefit from ML/AI?

  6. What strategies are needed to make good use of ML/AI in E3SM development?