Earth System Models and AI-Enabled Development
We develop and improve Earth system models, atmospheric chemistry transport and air-quality models, particle dispersion models, and hybrid methods that combine machine learning with numerical simulation for Earth system research and air-quality prediction.
Why it matters
Numerical models connect observations, emissions, chemical mechanisms, and climate processes, making them essential for understanding air pollution and climate change, evaluating policy scenarios, and predicting air quality. Yet these systems are highly complex and some processes remain difficult to parameterize. Our research advances atmospheric chemistry modules in Chinese Earth system models and applies machine learning to constrain atmospheric composition, improve forecasts, correct model mechanisms, and develop new parameterizations.
Topic
Atmospheric Chemistry Modules in Earth System Models
We develop atmospheric chemistry modules for China Meteorological Administration Earth system models, improving atmospheric composition simulation and coupling with other Earth system components.
Our contributions
- The group has developed the first and second generations of the BCC-GEOS-Chem atmospheric chemistry modules, BCC-GEOS-Chem v1.0 and v2.0.
Paper
Featured
Latest
Our teamDevelopment of the global chemistry-climate coupled model BCC-GEOS-Chem v2.0: improved atmospheric chemistry performance and new capability of chemistry-climate interactions
Representative work for Atmospheric Chemistry Modules in Earth System Models.
Our teamDevelopment of the global atmospheric chemistry general circulation model BCC-GEOS-Chem v1.0: model description and evaluation
Representative work for Atmospheric Chemistry Modules in Earth System Models.
Related people
Hybrid Numerical Modeling and Machine Learning
We develop AI algorithms that work with numerical models to improve simulations of radicals, ozone, and other reactive constituents, and to advance AI-enabled air-quality prediction.
Our contributions
- We use machine learning to identify and correct biases in modeled ozone chemical production.
Paper
Our teamDivergent Ozone Predictions in China Under Carbon Neutrality: Why Chemical Mechanisms Disagree
Representative work for Hybrid Numerical Modeling and Machine Learning.