Numerical modeling is a very useful approach to understand atmospheric processes, and also for weather and climate prediction. Since 1950s, atmospheric model becomes more and more important for research of atmospheric science. It also becomes a critical component of Earth system model. Nowadays, it has been widely used for understanding the mechanisms of extreme events, climate change, air pollution, and etc. However, atmospheric model has never been perfect, and includes many simplified and uncertain parameterizations of complex and multi-scale atmospheric processes. Therefore, it is necessary to understand the uncertain sources of atmospheric model so that simulation result can be better interpreted. In this seminar, some uncertainties associated with modeling atmospheric aerosol and its climatic impact will be discussed. One uncertainty quantification (UQ) framework will also be introduced to investigate the sensitivity of modeling results to the selected parameters. These studies can provide useful and practical guidance on the improvement of physical and chemical parameterizations to reduce model uncertainties in simulating extreme events, climate change, air quality, and etc. Numerical modeling is a very useful approach to understand atmospheric processes, and also for weather and climate prediction. Since 1950s, atmospheric model becomes more and more important for research of atmospheric science. It also becomes a critical component of Earth system model. Nowadays, it has been widely used for understanding the mechanisms of extreme events, climate change, air pollution, and etc. However, atmospheric model has never been perfect, and includes many simplified and uncertain parameterizations of complex and multi-scale atmospheric processes. Therefore, it is necessary to understand the uncertain sources of atmospheric model so that simulation result can be better interpreted. In this seminar, some uncertainties associated with modeling atmospheric aerosol and its climatic impact will be discussed. One uncertainty quantification (UQ) framework will also be introduced to investigate the sensitivity of modeling results to the selected parameters. These studies can provide useful and practical guidance on the improvement of physical and chemical parameterizations to reduce model uncertainties in simulating extreme events, climate change, air quality, and etc.