During 2018–2019, Energy Foundation China supported Planet-Data Inc. to investigate temporal and spatial distribution of PM2.5 in Jinan, the capital city of eastern Chinese province Shandong that has suffered smog episodes.

Using satellite data, this report in June 2019 combines Aerosol Optical Depth, meteorology, and land use statistics to evaluate Jinan’s PM2.5 concentration from 2003 to 2017. Meteorological factors significantly affect monthly PM2.5 predictions, the report finds, while industrial electricity consumption is positively correlated with annual PM2.5 concentration (correlation coefficient=0.68), with a one-year lag. These model calculated findings have been verified by 2013–2017 data from air quality monitoring stations on the ground.

This project was an attempt to use machine learning models to supplement PM2.5 temporal and spatial distribution data before 2013, when China’s air quality monitoring stations began to record PM2.5 concentration on a regular basis. The historical data produced by this project can be helpful for identifying sources of PM2.5 pollution and determining emissions control policies.