Investigating the effect of China’s clean coal technology policy on air quality is of great significance for promoting energy transformation and formulating follow-up policies. Utilizing 31 provincial cities data in Chinese mainland from 2013 to 2020, the spatial variation characteristic and change rate of air quality index (AQI) are discussed in this study. Amongst, the AQI in 2020 is predicted by deep learning approaches, to eliminate the uncertainty that COVID-19 bring about. The association analysis between AQI and socio-economic factors is also conducted, to clarify the internal mechanism of clean coal technology policy. The results show that 1) The AQI can be better predicted by the tailored Convolutional Neural Network-Long Short Term Memory (CNN-LSTM) network; 2) the air pollution in China shows an integration trend, embodying heavy and slight pollution in Northern and Southern China, respectively; 3) the clean coal technology policy has an average reduction effect of 18.82% on AQI. And there is a 2-year time lag before the policy takes any strong positive effects; 4) the clean coal technology policy mainly improved air quality through the way of emission reduction and de-industrialization. Practicable policy suggestions are put forward to supporting emission reduction, promoting energy transformation in China and applicable to other developing countries with scarce energy resources and severe air pollution.
Keywords Coal-based clean energy, air quality, clean coal technology policy, deep learning approaches, energy transformation