Research
My research lies at the intersection of atmospheric science, climate dynamics, and machine learning. I am particularly interested in how AI models learn atmospheric dynamics, how they behave for extreme and rare events, and what sets the intrinsic limits of weather predictability.
Current themes include AI weather prediction, out-of-distribution extremes, butterfly effects and perturbation growth, gray-swan events, and the role of physical reasoning in improving machine-learning-based forecasts.
Research Interests
- AI weather prediction and climate modeling
- Atmospheric predictability and forecast limits
- Extreme weather and out-of-distribution events
- Butterfly effects and perturbation growth
- Physics–AI integration for Earth system science
AI Weather Models
Understanding how AI models forecast the atmosphere, where they succeed, and where they struggle compared with traditional numerical models.
Rare Extremes
Studying gray-swan and out-of-distribution events such as unprecedented rainfall and extreme tropical cyclones.
Predictability
Investigating intrinsic and practical predictability limits, including the role of perturbation growth and nonlinear dynamics.
Publications
For a full and up-to-date publication list, please visit my Google Scholar profile.
Selected Work
-
Predicting Beyond Training Data via Extrapolation versus Translocation
AI weather models and unprecedented extreme rainfall events. -
Butterfly Effects and Finite Predictability in AI-Based Weather Prediction
Predictability diagnostics and perturbation growth in AI forecast systems. -
Can AI Weather Models Predict Out-of-Distribution Gray Swan Tropical Cyclones?
A controlled study of rare tropical cyclone intensity events.
You may later replace this section with an automatically updated publication list or add full citations manually.
Group
I am building a research group at Nanjing University focused on atmospheric predictability, AI for weather and climate, and extreme-event dynamics.
Prospective Students and Collaborators
I welcome inquiries from students and collaborators interested in atmospheric dynamics, machine learning, data-driven Earth system modeling, and weather extremes.
Open Directions
- AI and physics for predictability
- Rare-event learning in Earth system models
- Dynamics of extreme weather events
Opportunities
If you are interested in joining the group, please email me with your CV, research interests, and a brief introduction.
Teaching
Information about courses, lecture notes, and teaching materials will be posted here.
Planned / Possible Topics
- Atmospheric Dynamics
- Weather and Climate Predictability
- Machine Learning for Earth System Science
- Extreme Weather and Climate Risks
CV
You can download my CV here:
Replace Qiang_Sun_CV.pdf with your actual CV file in the same GitHub repository.
Contact
Qiang Sun
Assistant Professor
School of Atmospheric Sciences
Nanjing University
Email: qiangsun@nju.edu.cn
Nanjing University profile: Faculty Page
Google Scholar: Scholar Profile