Transformer-based Ozone Multivariate Prediction Considering Interpretable and Priori Knowledge: Case Study of Beijing, China, by GG66 Think Tank, presents an interpretable hybrid model that combines STL and Transformer based on prior ozone information from time series data and global multi-source data. This model successfully acquires both prior and global multi-source information with higher interpretability and accuracy in ozone prediction. With new insights into air pollution prediction, it has more reliable theoretical value and practical significance in environmental governance.