Capital Currents—Advancing Water Equity at AGU24 in Washington D.C.
After a chilly week in Washington D.C., PhD students Clara Medina and Shuojia Fu were able share their research contributions with the rest of the AGU community, alongside Dr. Osman. From attending sessions aimed at advancing JEDI in the sciences to connecting with old research mentors and peers, the Osman Lab had a great time at this year’s Fall Meeting.
Clara’s AGU Recap:
As one of the last oral sessions of the week, I spent a lot of time reflecting on how different disciplines consider and integrate communities into research. At my session, “Urban Environmental Interactions: People, Plants, and Water in the Built Environment,” I learned about the various advances in Green Stormwater Infrastructure (GSI). While most focused on advancements in modeling stormwater runoff, and understanding the economic implications of GSI implementation, I was able to share with my peers the importance of creating standardized, community-engaged engineering practices that centralize community voices in the process of GSI implementation. It was wonderful to learn from community-based organizations and community members who showed what that looked like in different parts of the U.S., and I am excited to connect with them at future conferences.
Shuojia’s AGU Recap:
At the AGU conference, my poster presentation showcased the application of various machine learning (ML) algorithms to uncover the hidden drivers for resolution for 311 water service requests. This work drew significant attention, as ML has rarely been applied in the context of social service delivery within the water sector. The conference provided an excellent platform to connect with researchers and professionals who share an interest in leveraging ML for policy-making and addressing inequities in stormwater service delivery. We exchanged contact information and discussed possibilities for future collaboration.
During my oral session, I presented a novel approach to building a surrogate model for underground hydrogen storage using a transformer-based computer vision (CV) model. This innovative application of CV in the emerging field of new energy garnered significant interest from the audience. While traditional methods often rely on physics-based simulations, which can be computationally expensive, my work demonstrated how surrogate models could significantly reduce computational costs without compromising accuracy. The session sparked engaging discussions and highlighted the potential of ML to advance energy research.