2009 Opportunities and Challenges in Computational Geophysics

Proposed Schedule for the CIG Workshop on Opportunities and Challenges in Computational Geophysics

March 30-31, 2009, on the Caltech Campus, Pasadena CA

Day 1: March 30 (Monday) SHARP LECTURE HALL AM SESSION/BENIOFF ROOM PM SESSION

8:00 am-- SHARP LECTURE HALL Welcome and Overview: the goals of the workshop & schedule, including a brief summary of the history (NSF funding and initial goals of CIG-I), the CIG-II proposal writing committee and the deadlines for that proposal.

SESSION 1: CURRENT AND FUTURE COMPUTATIONAL GEOSCIENCE DIRECTIONS: SUCCESSES & FAILURES

Representative from each of these working groups summarizes the major computational activities of CIG and uses one-pagers & workshop application feedback to summarize the scientific activities supported by CIG. 20 minute talks with 10 minutes for questions

8:30 am

Short-term Tectonics
Speaker: Mark Simons

9:00 am

Long-term Tectonics
Speaker: Dennis Harry

9:30 am

Seismology
Speaker: Jeroen Tromp

10:00 am

Coffee Break (Outside Sharp Lecture Hall)

10:30 am

Magma Dynamics
Speaker: Laurent Montesi

11:00 am

Mantle Convection
Speaker: Shijie Zhong

11:30 am - 12:30 pm

Break-out Sessions: Groups self-organize by Scientific Application (MC, ST, LT, MD, SEIS) Discussion of the success and failures of CIG-I regarding its role in enabling innovative computational geoscience, supporting code development for new scientific directions, and energizing computational geoscience community. What should CIG keep doing, do differently, add, change...?

Discussion Leaders (responsible for giving 5-minute summary after lunch):

Short-term Tectonics: Rowena Lohman

Long-term Tectonics: Todd Ehlers

Seismology: Jeroen Tromp

Magma Dynamics: John Hernlund

Mantle Convection: Clinton Conrad

12:30 pm

Lunch - Athenaeum, Main Lounge

1:30 pm

BENIOFF ROOM Short Reports of Breakout 1: including 1-slide, 5 minutes per group, 5 groups

SESSION 2: INNOVATIONS IN & OPPORTUNITIES FOR COMMUNITY SOFTWARE DEVELOPMENT

Speakers discussing the role specific software packages or architectures can and/or have played in CIG-supported software, and strategies for integrating new computational methods into new or existing software, and distributing these products. 15 minute talks with 5 minutes for questions

2:00 pm

AMR & Deal-II What is it? How is it being used for CIG application (example)? How can it be better utilized in the future? Speaker: Carsten Burstedde

2:20 pm

Code Complexity: Current & Future Infrastructure: background and experiences with current CIG and other tools, and where we are headed with multi-scale and multiphysics problems in the future. Speaker: Steve Quenette, VPAC.

2:40 pm

Hybrid Finite Element Methods: What is it? How is it being used for CIG application (example)? How can it be better utilized in the future? Speaker: Louis Moresi

3:00 pm

Automatic Code Generation and the FEniCS Project. Speaker: Anders Logg

3:20 pm

Coffee Break (Benioff Room)

3:50 pm

Data Assimilation: Examples and strategies for effective and efficient data assimilation in large-scale computation. Speaker: Gary Egbert

4:10 pm

Seismology: Seismological Grand Challenges in Understanding Earth’s Dynamic Systems. Speaker: Jeroen Ritsema, Univ. of Michigan

4:30 pm

Software Design: Geodynamics Software Development in Australia - Lessons learned and future directions. Speaker: Bill Appelbe, VPAC

4:50 pm

Micro-to-Macro-Scale Modeling: Predictive Modeling and Simulation of High-Energy Density Dynamic Response of Materials (PSAAP, Caltech) Speaker: Michael Aivazis

5:15-6:15 pm

Breakout Session 2: Groups self-organize by interest in Multi-scale/Multi-physics Modeling, Solver-Coupling/Mesh Generation/Scalability, and Data Assimilation. Discussion of how we can better take advantage of opportunities/innovations from computer science. What has worked well or not (examples)? What do our scientific goals demand from computational methods in the future?

Discussion Leaders (responsible for giving 5 minute summary Tuesday Morning):

Multi-scale/Multi-physics Modeling: TBD

Solver Coupling/Mesh Generation/Scalability: TBD

Data Assimilation: TBD 

6:30 pm

Dinner: Athenaeum, Hall of Associates

8:00-10:30 pm

Poster Session, Athenaeum, Main Lounge

Day 2: March 31 (Tuesday) SHARP LECTURE HALL

8:00 am

Overview of Day 2, Comments, etc.

8:05 am

Short Reports of Breakout 2: including 1-slide, 5 minutes per group, 4 groups

SESSION 3: EDUCATION & TRAINING OF THE COMPUTATIONAL GEOSCIENCE COMMUNITY

CIG success depends on training the current and next generation scientists to use and understand the software developed and made available by CIG. Education and training can take many forms, from learning the nuts-and-bolts of running a code, to the theory behind the methods used, to the science achieved with the software.

8:30 - 9:30 am

E&T Activities of CIG-I: What worked, what didn’t, what have we learned? Summary from each group based on workshop surveys and other information. 10 minutes per group 10 minutes for general questions/comments at the end.

Short-term Tectonics: Brad Aagaard

Long-term Tectonics: Todd Ehlers

Seismology: TBD

Magma Dynamics: TBD

Mantle Convection: TBD

9:30 am

SPICE: Introduction to Education and Training Approaches used by SPICE, a seismology education training group in Europe. Advice, pitfalls, what works... Speaker: Heiner Ingel

9:50 am

CIDER: Opportunities for CIG software education and training. Speaker: Barbara Romanowicz

10:10 am

Coffee Break (Outside Sharp)

10:40 am

Strategies For Creating Accessible Software. Developing intuitive GUIs, real-time visualization of running models, and useful documentation. Speaker: Dave Stegman

11:00 am

NSERC: Strategies for integrating CIG research, training and education goals. Speaker: Dave Mogk.

11:30 am - 12:30 pm

Break-out Session 3: Led by Dave Mogk. Discussion on: What are the best methods and venues to train people to use and understand the geoscience software? What different education strategies should CIG support/implement (online tutorials, workshop training sessions, dedicated summer training sessions, levels of training)? How do we motivate individual scientists to be more involved in the education and training activities (both as teachers and students)?

12:30-1:30 pm

Lunch, Athenaeum

1:30 pm

Short Reports of Breakout 3: including 1-slide, 10 minutes per group, 2 groups, 10 minutes for questions/comments.

SESSION 4: STRATEGIES FOR CONTINUING SUCCESS IN COMPUTATIONAL GEOSCIENCE

How can CIG help to create opportunities in computational Geosciences, take advantage of innovations in computer science, and enable cutting-edge research by overcoming computational challenges.

2:00 pm

CIG-II: Brief talk summarizing workshop so far and introduction of the draft CIG II proposal. Speaker: Louise Kellogg

2:30-3:30 pm

Break-out Session 4: Participants self-organize into 4 groups based on the major aspects of the CIG-II Draft Proposal and discuss the future directions for scientific, software and community development in Computational Geosciences.

Discussion Leaders (responsible for giving 5 minute summary after coffee break):

1. TBD
2. TBD
3. TBD
4. TBD

3:30 pm

Coffee Break (Outside Sharp)

4:00 pm  

Short Reports of Breakout 3: including 1-slide, 5 minutes per group, 4 groups

4:30 pm

Time for open discussion

4:55 pm

Closing Remarks