Dr. Yixin Berry Wen

Contact

Email yixin.wen@ufl.edu

Office Turlington Hall 3203

Focus Areas

Research Statement

My research seeks to harness artificial intelligence to tackle fundamental challenges in Earth Science, ensuring AI is guided by domain knowledge and societal needs. For AI to save lives from natural hazards and mitigate climate risks, it must move beyond black-box prediction toward transparent, trustworthy science. My work confronts three core challenges of AI in Earth science: the uniqueness of Earth science data, the opacity of “black box” models, and the communication gap between AI and the scientists and communities who need it. As the inaugural editor of JGR: Machine Learning and Computation, I am helping to lead this transformation not just through research, but through community service.

My research is organized around overcoming these three core challenges.

From Messy Data to Foundational Knowledge. To address the data challenge, my research transforms incomplete observations into high-fidelity, physically consistent datasets. My group has established benchmark methods for cross-validating and fusing multi-platform remote sensing products from the atmosphere and cryosphere, ensuring that ‘big data’ are also ‘good data’. We have recently pioneered the use of AI diffusion models to generate physically plausible, high-resolution (1-km) precipitation fields, creating a new paradigm for filling observational gaps over data-sparse regions like oceans and mountains. This work, complemented by my leadership as chair of NASA’s GES DISC User Working Group, is building a more reliable data foundation for all Earth sciences.
From Black Boxes to Trustworthy Science. To build trust in AI, I develop methods that fuse the predictive power of machine learning with the explanatory power of physical law. My primary tool is symbolic regression (SR), a technique that distills complex models into compact, human-readable equations. We have used SR to discover new, interpretable equations governing rainfall microphysics from radar data and to generate more accurate tornado probability models. This approach transforms AI from an opaque oracle into a partner in scientific discovery, generating trustworthy and operational knowledge that connects atmospheric processes to hydrological impacts.
From Opaque Outputs to Collaborative Communication. To bridge the communication gap, I am building collaborative AI agents that democratize access to complex Earth data. These agents use natural language to run models, interpret climate projections, and deliver tailored reports, empowering both scientists and community decision-makers. Crucially, this work is deeply rooted in community partnership. Through co-design with Native tribes in Oklahoma and Florida, we are ensuring these tools integrate traditional Indigenous knowledge with modern AI, creating a more holistic and equitable approach to building climate resilience at the intersection of the hydrosphere and the anthroposphere.

Recent Funded Projects

NSF CAREER: Elucidating Florida Sea-Breeze Convection using High-Resolution Remote Sensing and Interpretable AI (PI, 9/2026-08/2031)

NASA: The Integrated Multi-satellite Retrievals for Global Precipitation Measurement (GPM) mission (IMERG) based Climate Data Record (PI, $531,933, 2026-2029)

NASA: Precipitation Efficiency (PE) quantification for tropical and subtropical storms using combined GOES-IR and sub-pixel TROPICS observations (Co-PI, $247,766, 2026-2029)

NASA: Using NASA Earth observations to support drinking water quality management in the U.S.-Mexico border region (Co-PI, 74,933, 2026-2029)

NSF 25-530 Collaborative Research: CAIG: Reliable Uncertainty-Aware Generative Downscaling for Geoscience Data (PI, $353,178, 2025-2028)

NASA, Delineating and Characterizing Dry Spell Events in the Horn of Africa (PI, $100,000, 2025-2027)

NASA: A.47 Earth Action: Community Action for Equity and Environmental Justice, Building Co-Design and Co-Learning Digital Twins against Floods on Tribal Lands in support of Indigenous Communities (Co-I, $750,374, 2024-2027)

NASA Jet Propulsion Laboratory, California Institute of Technology A19-0075-001, Analysis of the performance of the Atmospheric Infrared Sounder (AIRS) retrieval system (PI, $140,000, 2018-2021)

Recent Publications

Yi, C., Yu, M., Qian, W., Wen, Y., & Yang, H. (2026). Efficient kilometer‐scale precipitation downscaling with conditional wavelet diffusion. Journal of Geophysical Research: Machine Learning and Computation, 3(2), e2025JH000941.

Zeraati, M., Farahmand, A., Seager, R., Fowler, H. J., Madani, N., Parazoo, N., ... & AghaKouchak, A. (2026). Assessing flash drought development and propagation across the contiguous United States using remote sensing. Earth's Future, 14(3), e2025EF007037.

Kisembe, J., Wen, Y., Wainwright, C. M., Funk, C., Odongo, R. I., & Qian, W. (2026). Contrasting Changes in Rainy Season Length, Rainfall Frequency, and Intensity across Eastern Africa. Journal of Hydrometeorology, 27(3), 325-342.

Song, J., Yong, B., Zhang, H., Zhang, F., Ahmed, Z., Chan, N. W., ... & Wen, Y. (2026). Advancing Spaceborne Precipitation Radar Monitoring: Systematic Comparison Between PMR and DPR Observations. IEEE Transactions on Geoscience and Remote Sensing, 64, 1-18.

Wang, Y., Yong, B., Qi, W., & Wen, Y. (2025). Recent oceanic performance of GPM multisatellite precipitation estimates benchmarked by passive aquatic listeners. Journal of Hydrology, 134475.

Liu, Z., & Wen, Y. (2025). Accelerating Earth Science to Action. Bulletin of the American Meteorological Society, 106(10), E2043-E2051.

Rahaman, M., Southworth, J., Wen, Y., & Keellings, D. (2025). Assessing Model Trade-Offs in Agricultural Remote Sensing: A Review of Machine Learning and Deep Learning Approaches Using Almond Crop Mapping. Remote Sensing, 17(15), 2670.

Mei, J., Yong, B., Lyu, Y., Qi, W., Wen, Y., Wang, G., & Zhang, J. (2025). Runoff evolution responses to climate change: A case study in the headwater area of Yellow River, China. Journal of Environmental Management, 384, 125512.

Qian, W., Wen, Y., Gao, S., Li, Z., Kisembe, J., & Jing, H. (2025). Evaluation of near‐surface specific humidity and air temperature from Atmospheric Infrared Sounder (AIRS) over oceans. Earth and Space Science, 12(4), e2024EA003856.

Zhu, S., Li, Z., Chen, M., Wen, Y., Liu, Z., Huffman, G. J., ... & Hong, Y. (2025). Evaluation of IMERG climate trends over land in the TRMM and GPM eras. Environmental Research Letters, 20(1), 014064.

Recent courses

GEO 6119 Proposal Writing

MET4410 Radar and Satellite Meteorology

MET4224C Machine Learning in Meteorology

MET1010 Intro to Weather and Climate

Educational Background