From the Program Manager, Aug ’24: Modeling the Climate at Exascale

  • August 29, 2024
  • Feature Story,Home Page Feature
  •  

    Xujing Davis, ESMD Program Manager, DOE BER

    Xujing Davis, E3SM Program Manager, DOE BER

    Welcome to the E3SM newsletter, the last issue before E3SM’s 10th birthday in October! During the past few months, significant achievements and progress have been made. E3SM scientists have also been actively engaged in DOE-organized and international science events, communicating, contributing E3SM results and broadening scientific impacts.

    Major Achievements

    Modeling Climate at the Exascale

    A recent publication documented the SCREAMv1 work, including the successful refactoring and improvement of the SCREAM’s code, as well as running and evaluation of additional 4 season simulations at 3km global setting. Though many improvements need to be made, SCREAMv1 does well at realistically simulating many phenomena, testifying to the power of resolving the scales of motion between 10 and 100 km. The E3SM Atmosphere Model in C++ (EAMxx), evolved from SCREAM, is the first truly exascale global atmosphere model. This is a significant achievement and essential stepstone for E3SM to model climate at the exascale, via both the upcoming decade long SCREAMv1 simulations and the fully coupled E3SMv4 with EAMxx as the new atmosphere component.

    E3SMv3. LR Simulation Campaign

    Another major effort has been the E3SMv3 low resolution (E3SMv3.LR or v3.LR) simulation campaign. With intensive diagnosing and testing, the team has resolved unexpected challenges of model crash issues and the campaign is now back on track! E3SMv3 includes many significant advancements from its previous versions such as a much-improved historical temperature representation. v3.LR is the workhorse configuration of the E3SMv3 physical coupled model capable of running simulations spanning thousands of years and is expected to be broadly used for a wide range of earth and environmental scientific explorations.

    Celebrating a Decade of Earth System Modeling (continued from the previous issue)

    Ten years of model development has given the E3SM team plenty of time to learn important lessons. Some of the most valuable ones are shared by the team. Most of these, if not all, will continue to serve as important reminders for future model development activities. Working together for a decade, our team has also generated several what might be called slogans, mottos, or sayings. It is intriguing to capture some of these to get a sense of the team’s chemistry, personality, humor, and secret ingredients in project management leading to where E3SM is today.

    E3SM scientists in action, enhancing integration and broadening scientific Impacts

    DOE Sponsored or Co-Sponsored Events

    One major event has been the Earth and Environmental Systems Modeling Program (EESM) PI Meeting in August in Maryland, bringing together more than 300 scientists across EESM, to share scientific progress and vision, with the goal to foster collaboration and integration among EESM scientists to accelerate discoveries and innovations for DOE mission. I greatly enjoyed meeting our talented team of scientists and hearing about the exciting progress, the increased interest in E3SM, and future opportunities. I want to thank you all for your presentations and voices, and for the help in organizing different breakout sessions as well as your contributions to the EESM PI Meeting report.

    The Annual SciDAC-5 PI Meeting was also held in Rockville, MD in July.  The BER Breakout Session focused on the update from seven BER-ASCR SciDAC partnership projects and four topic discussion sessions centered around 1) SciDAC/E3SM coordination, 2) Collaboration between computational and physical scientists, 3) AI and ML in Earth system modeling, and 4) Computational Resources. It has been a productive and engaging meeting with enthusiastic input from scientists from both BER and ASCR-relevant domains. The major outcomes and the future opportunities identified are captured in this Summary of the 2024 SciDAC PI Meeting and will serve as an important reference for our future research activities.

    Jointly funded by DOE/LANL, DoD/ONR, and NSF/NCAR, the CICE Users Workshop and Tutorial was held by LANL in Santa Fe, NM in May. A driving motivation was to develop relationships with wave-ice experts worldwide to strengthen E3SM’s new development in this area. The workshop highlighted major modeling and observational gaps and conveyed the urgency for this area of research given the fast-changing nature of the sea ice during recent years. 

    International Activities

    As one of the major US modeling center activities, E3SM project has committed to some but not all CMIP7 and its Fast Track simulations using E3SMv3. Please check out the specific plans provided by the E3SM Project Principal Investigator Dave Bader.

    In late May, E3SM team lead of Infrastructure, Performance and Data group, Robert Jacob, was invited to present an update on E3SM and the general state of exascale readiness in the U.S during the 8th European Network for Earth System modelling (ENES) Workshop on High Performance Computing for Weather and Climate in Lecce, Italy. Please check out the update from Rob on major points of his presentation as well as the exascale strategies and status from other international modeling centers.

    The Ice Sheet Model Intercomparison Project for CMIP7 (ISMIP7) Workshop was held in June with the aim to address the need to quantify uncertainties in ice-sheet projections and to represent the full range of uncertainties inherent in model projections. The DOE E3SM science community was strongly represented in both leadership roles and as members of different Focus Groups. By participating in and contributing to this important international effort of IPCC/CMIP, DOE scientists benefited greatly from their peers’ expertise and knowledge in further advancing E3SM and future coordination.  

    In June, a group of DOE lab scientists co-organized a mini-symposium entitled “Advanced Computational Methods for Climate Modeling and Analysis” during the 9th European Seminar on Computing in Pilsen, Czech Republic, and online.  Among 40 presentations, there were 20 DOE-supported works, covering a wide range of topics from atmosphere, ocean, sea ice and land ice, land and river components as well as coupled systems of the E3SM.

    Research Highlights

    Exciting progress has been made in 1) understanding the tipping point for Antarctic icesheet melt using E3SM; 2) improving E3SM ability to quantify the Greenland Ice Sheet’s (GrIS) current and future contributions to sea-level rise ; 3) autocalibration of E3SM Atmosphere Model using a Machine Learning approach with improved computational efficiency and accuracy; 4) a new method enabling extremely efficient tracer transport to find stabilized element bases; 5) more accurate simulations of coastal ecosystems via incorporating tidal-driven biogeochemistry in a land surface model; and 6) filling a critical knowledge gap through incorporating new particle formation mechanisms to improve aerosol and cloud simulations, which has important implications for assessments of historical and future climate change.

    New Resources

    Following the big success of the first E3SM In-Person Tutorial Workshop in May, the E3SM Communication Team has assembled all the material and made it available online! If you are interested in running or developing E3SM including its individual model components, then these materials, thoroughly prepared by all team leads, would be very helpful in answering any questions you may have.  

    Around DOE

    BER has recently announced a series of new awards, among which, there are 13 University Awards from the Earth System Model Development and Analysis Funding Opportunity announcement (FOA3228) in the areas of marine and terrestrial biogeochemistry, aerosol, cloud, and/or their interactions and feedbacks as well as the novel methodologies and techniques for model initialization coupled data assimilation. I want to thank all the applicants for your high-quality proposals and your interest in DOE modeling activities, I also want to express my sincere gratitude to all the reviewers whose expertise and evaluation has been central to our decision making. Please check out the details of the new awards and several Funding Opportunity announcements. 

     

    Concluding Remarks

    Congratulations to the team for a few intensive and productive months! Thanks to all for your continued dedication and contribution. I hope you will find time to take some nice breaks to enjoy the rest of the summer!  

     

     Xujing  

    Send this to a friend