2025 LEAP SUMMER MOMENTUM FELLOWSHIP
Overview
LEAP’s Summer Momentum Fellowship welcomes doctoral students in data science interested in having a summer research immersion in climate data science, with the opportunity to apply their data science/machine learning skills in climate modeling and develop research interests in climate data science. Each fellow receives a summer stipend, travel support, and access to LEAP resources, such as LEAP Pangeo and workspace at Columbia University’s Innovation Hub.
Momentum Fellowship project leads work closely with Fellows throughout the summer on a well-defined, yet open-ended, machine learning research problem in climate data science. Project leads will also guide their Fellows to present their summer research at future LEAP events and other workshops/conferences.
Momentum Fellows are responsible for mentoring up to two (2) undergraduate students in the LEAP Research Experiences for Undergraduates (REU)program. Each project will host up to two (2) undergraduate students paired with one (1) Fellow.
2025 Summer Momentum Fellows
Based on his summer research as a LEAP Momentum Fellow, Sudhanshu presented his poster, “Leveraging machine learning and MCMC to optimize Community Land Model parameters for accurate gross primary productivity estimates,” at the 2025 AGU Annual Meeting (New Orleans, LA).
SUDHANSHU KUMAR
Project 3: Developing Machine Learning Emulators to Constrain Parametric Uncertainty in Land Surface Modeling and Carbon Cycling (Hawkins)
Sudhanshu is a Ph.D. student in Earth System Science at Auburn University with a Master’s in Data Engineering. His research focuses on integrating machine learning, remote sensing, and climate data to improve the prediction and understanding of hydroclimatic extremes. He is passionate about applying data-driven approaches to Earth system processes and to enhance climate-informed decision-making. Sudhanshu’s long-term goal is to bridge hydrology, climatology, and data science to address critical water and food sustainability challenges under a changing climate. He has published and peer-reviewed articles on droughts, heatwaves, and climate impacts, and is actively engaged in interdisciplinary research collaborations. One of his proudest achievements is developing a deep learning-based emulator for soil moisture and drought monitoring that performs well across multiple U.S. regions. Sudhanshu is committed to advancing equitable, impactful research that contributes to resilient and sustainable environmental systems.
Based on his summer research as a LEAP Momentum Fellow, Kris presented his poster, “The opportunity cost of 2023 Canadian Wildfires,” at the 2025 AGU Annual Meeting (New Orleans, LA).
SHENGJIE KRIS LIU
Project 2: Tropical Forest Dynamics in a Changing Climate (Weng)
Shengjie (he/him) is a PhD candidate at the Spatial Sciences Institute, University of Southern California. His work focuses on extracting actionable information from Earth data through training machine learning models, addressing real-world challenges including imperfect data, resource constraints, and open-world scenarios. He is currently developing data-driven, physics-guided deep learning models to generate temperature data at high spatiotemporal resolution. A first-generation college graduate who didn’t know about research before college and now enjoys it, Shengjie is pursuing a career as a professor or research scientist working with Earth data and AI to provide actionable information in a changing world. Outside of work, you can find Shengjie in the stands at a soccer stadium—he’s a season ticket member of LA Galaxy.
Based on her summer research as a LEAP Momentum Fellow, Giorgia presented her poster, “Inferring thermodynamic histories from in situ ice crystal imagery via conditional diffusion models,” at the 2025 AGU Annual Meeting (New Orleans, LA).
GIORGIA NICOLAOU
Project 5: Using Generative AI to Improve Modeling of Cirrus Cloud Processes (Lamb)
Giorgia is a data science PhD student passionate about applying explainable machine learning, causal discovery, and deep learning to climate science. Her research focuses on uncovering the drivers of complex atmospheric processes to improve climate models and predictions. Professionally, Giorgia aspires to become a teaching professor, combining her love for education with her commitment to impactful research. One of her proudest achievements is her long-standing commitment to supporting children with cancer, through volunteering, fundraising, and advocacy efforts that continue to inspire her work and values.
Based on his summer research as a LEAP Momentum Fellow, Mostafa presented his poster, “Spatial clustering of heat wave regimes through ConvAE latent space analysis and teleconnection linkages,” at the 2025 AGU Annual Meeting (New Orleans, LA).
MOSTAFA REZAALI
Project 1: Probabilistic Data Assimilation Enhances the Prediction of Observed Temperature Extremes (Li / Gentine)
Mostafa is a Ph.D. student at the University of Florida, specializing in AI-based modeling of heat waves and flash droughts. My academic journey began with a B.Sc. and M.Sc. in Civil and Environmental Engineering in Iran, where I graduated among the top students in my cohort. Over the past six years, I’ve worked extensively with large climate datasets—such as WRF-ARW and ERA5—and developed machine learning models to enhance climate prediction and environmental forecasting. My core research interests lie in extreme heat, droughts, data-driven modeling, and AI in climate sciences. I’ve authored several peer-reviewed publications in journals like the Journal of Hydrology and Environmental Science and Pollution Research, and I serve as a reviewer for Science, Springer Nature Applied Sciences, and Journal of Hydrology. I am also a member of the Iranian National Elites Foundation. I’m passionate about using artificial intelligence to solve real-world environmental challenges and excited to collaborate across disciplines to understand the impacts of climate change.
Based on her summer research as a LEAP Momentum Fellow, Akila presented her poster, “Deep contrastive learning for microphysics scheme comparison in CESM ensembles,” at the 2025 AGU Annual Meeting (New Orleans, LA).
AKILA SAMPATH
Project 4: Understanding Cloud Microphysical Process in Climate Models (Fan)
Akila is a Ph.D. candidate in Information Systems at the University of Maryland, Baltimore County. Her research interests focus on AI/ML and explainable AI approaches to study Arctic sea ice and other affecting factors of the atmosphere and ocean. This summer, she will be conducting research to understand cloud microphysical processes in climate models using machine learning. In her free time, she enjoys spending time in nature and hiking. She hopes to pursue a future career in research labs or academia. She is proud to be persistent in pursuing her research interests that combine climate modeling and AI.
Projects
Click the image below to learn more about the Summer 2025 Momentum Fellowship Research Projects.
