Big Earth Data Method Uncovers Two Decades of PM2.5 Transport Pathways Across China

A new study has, for the first time, mapped long-term, large-scale transport pathways of PM2.5  pollution across China from 2000 to 2021. Published in Geoscience Frontiers, the research introduces an innovative data-driven framework based on the Dynamic Times Warping (DTW) method to trace how air pollutants have moved across the country during the two decades. The findings offer critical insights into the evolution of PM2.5 transport and provide scientific support for more effective air quality management.

The analysis reveals that PM2.5 transport pathways are most prevalent in Northwest China (34% of all identified pathways), followed by the Southwest (22%), and North (21%), with smaller shares in the Northeast (10%) and isolated cases elsewhere. Importantly, the number of transport pathways declined significantly after 2013, mirroring the sharp decline in annual mean PM2.5 concentrations during the same period. This trend reflects the positive impact of China's increasingly stringent environmental policies.

Unlike traditional chemical transport models, which requires complex meteorological or emission datasets, the DTW-based method is purely data-driven. It nevertheless effectively captures PM2.5 transport pathways, showing strong spatio-temporal agreement with established methods. This makes the method particularly useful in regions with sparse monitoring infrastructure such as Northwest China, where it can uncover potential sources and transport routes that were previously difficult to detect.

"This work demonstrates how Big Earth Data can reveal long-term patterns of pollution transport that are often missed by traditional methods," said the first author LIU Yiming. "By better understanding how pollutants move, we can design more effective strategies to protect air quality and public health."

The study also highlights the broader potential of the DTW approach, which can be extended to analyze the transport of other gaseous pollutants. Its adaptability and efficiency make it a promising tool for advancing regional environmental governance and supporting global sustainability goals.

Looking ahead, the research team plans to further enhance the DTW framework by improving cross-regional pathway generation, boosting computational efficiency, and incorporating scale effects. These improvements aim to provide a faster more precise picture of cross-regional pollutant transport, enhancing both the scientific rigor and practical applications of the model.

The study was led by PhD candidate LIU Yiming under the supervision of Prof. GUO Huadong from the Aerospace Information Research Institute (AIR), Chinese Academy of Sciences, in collaboration with the International Research Center of Big Data for Sustainable Development Goals (CBAS) and Peking University.