Re-Imagining the Center for Space Environment Modeling to Address Challenges in Modeling

Re-Imagining the Center for Space Environment Modeling to Address Challenges in Modeling Dan Welling & the CSEM Team Developing Heliophysics Standards and Cross-science Collaborations Workshop August 11, 2025 History of CSEM • Founded by Tamas Gombosi in early 2000s to support the development of the Space Weather Modeling Framework. • Main goal: organize numerical modelers within CLaSP (né AOSS) and promote inter-departmental collaboration. • Major Accomplishments: • Brought software engineering principles into heliophysics at a time where such practices were uncommon. • Worked with CFD specialists in Aerospace Engineering to create new solvers for the nacent BATS-R-US MHD model. • Developed the SWMF and many numerical models: BATS-R-US, AWSoM, etc. • As goals were achieved, center activity ramped down. • New director named in 2024 with objective to re-imagine CSEM. Challenges with Growth & Success 1. Responding to evolving landscape of software standards • Navigating open-source software licenses with heterogeneous model portfolio • Learning and embracing new tools & paradigms (e.g., Github vs. CVS) 2. Formalizing an educational mission to meet our reputation • Employers have come to expect certain capabilities from our students • How do we ensure and certify our student’s capabilities? 3. Centralization and stratification of knowledge • Limited amount of deep knowledge about models creates information chokepoints. • Information on model use for specific scenarios is relegated to pockets or sub-groups. • Knowledge bases are not well advertised; hidden from those in need. 4. Expanding use in science • Need to inform non-users about capabilities and connect with power users. Power User Power User Power User Power User Power User User User User User User User User User User UserUser User User User User User User User User User Over-Centralization of Information • Developer – Power User – Regular User landscape. • Required Information Flow: • Questions about use • Help with development • Bugs & issues • Dev requests • Growth of model use can over-tax developers • Infrequent peer-to-peer discussions. • New insights not shared broadly. • Lack of communication channels outside email. Core Dev Core Dev Core Dev Stratification of Information Tutorial material exists! •SWMF user manual •Aaron’ Ridley’s engineering & software development YouTube channel •My own Python syntax primers …but people don’t know where it is or how to get it. •Most common response: “wow, I didn’t know this existed!” Everyone is creating their own material in a vacuum. •We’re not building off each other’s material, just recreating. Stratification of Information A scenario: How to configure and use CIMI within the SWMF at Geospace. • CIMI used less frequently than RCM. • CIMI is more sensitive to simulation configuration than RCM – it’s harder to use. • Example input files exist but are out-of-date given updates to CIMI. • CIMI expertise lies more strongly with former Umich students. • New simulations require finding out who knows what and communicating this knowledge throughout the team. Use-case knowledge and known solutions are also types of “stratified knowledge” • Easily lost through group attrition! • Who-knows-what is not well communicated throughout the group. • Some lessons-learned happen many times over. • Applies to code use and code development scenarios! Connecting Science Models and Scientists A case study: including the plasmasphere in the SWMF • Multifluid BATS-R-US since 2009 • Incorporation of the DGCPM into the SWMF done in 2013 • Connecting the two yields new science on the impact of plumes – and lots more to come. Neat things like this come from connecting non-modelers to the models. • Devs always looking for low-hanging fruit model improvements • Non-devs don’t know what it takes to accomplish these things. • Non-modelers are not aware of existing capabilities. CSEM’s Vision & Mission VISION (the “what” of CSEM): The Center for Space Environment Modeling supports and grows excellence in software, science, and education within the field of heliophysics modeling. MISSION (the “how” of CSEM): CSEM promotes excellence in heliophysics modeling by empowering our members, endorsing exemplary models, connecting model developers and researchers, and hosting programs to improve the capabilities of the community. Statement of CLIMATE & INCLUSION: CSEM is an inclusive and diverse space that values and elevates the voice of all members using best practices outlined in our CODE OF CONDUCT. CSEM’s Expanded Vision & Mission 1. EXCELLENCE IN SOFTWARE: Foster innovative model development through a genuine software-engineering approach. • Host and support software repositories open to the broader heliophysics community. • Engage in a campaign to learn, train, and adopt modern programming practices and tools • Integrate knowledge/resources from computer science and information science into heliophysics. 2. EXCELLENCE IN SCIENCE: Grow application of models to accelerate heliophysics research. • Educate the broader community concerning existing and developing modeling capabilities. • Support development and planning of proposals that integrate numerical modeling. • Work with obs. and theory-based teams to identify new opportunities to use numerical models. 3. EXCELLENCE IN EDUCATION: Train the next generation of model developers and “power users”. • Host online help and discussion forums so that existing users can assist novice users. • Share sets of example simulations, input files, and thorough documentation that can help new users design numerical experiments. • Support CLASP’s teaching mission by helping to develop curriculum related to software development and numerical methods.