Acta Geographica Sinica ›› 2021, Vol. 76 ›› Issue (1): 191-205.doi: 10.11821/dlxb202101015

• Ecosystem Services • Previous Articles     Next Articles

High resolution PM2.5 estimation based on the distributed perception deep neural network model

LIU Jiwei(), MIN Suqin, JIN Mengdi   

  1. College of Data Science and Intelligent Media, Communication University of China, Beijing 100024, China
  • Received:2019-12-02 Revised:2020-11-03 Online:2021-01-25 Published:2021-03-25
  • Supported by:
    The Fundamental Research Funds for The Central Universities of China(3132018XNG1830)


The individual exposure level of fine particulate matter (PM2.5) is a key issue in the study of health effects. However, the lack of historical data and the small coverage of ground monitoring stations have hindered the development of related research. Based on the aerosol optical depth (AOD) provided by NASA remote sensing data, multi-source data such as ground monitoring and meteorological data were integrated for modeling to estimate near-ground PM2.5 concentration. The results have wide spatial coverage, strong time continuity and low method cost. Based on the Beijing-Tianjin-Hebei-Shandong region in 2018, this paper introduces 50 features such as meteorological elements, NDVI, time nodes and spatial markers to analyze the relationship between AOD and PM2.5. In view of the information loss caused by the single traditional interpolation method, the spatiotemporal multi-view interpolation method is used to improve the accuracy and coverage of interpolation. Considering the complex relationship caused by hysteresis of features, and the correlation and partial correlation between features, this paper uses a distributed perception deep neural network model (DP-DNN) to separately capture higher-order features between multiple-source features. The results show that: (1) The relative error of the spatiotemporal multi-view interpolation method is 27.5%, and the average proportion of missing data decreases from 52.1% to 4.84%. (2) In terms of time prediction, mean absolute error, relative error, mean square error and root mean square error of DP-DNN are 7.7 μg/m 3, 46.8%, 766.2 μg 2/m 6 and 26.9 μg/m 3, respectively. In terms of space prediction, they are 16.6 μg/m 3, 41.8%, 691.5 μg 2/m 6 and 26.6 μg/m 3, respectively. In aspects of accuracy, robustness, generalization ability and time consuming, the results are superior to linear statistical models and common deep learning architecture.

Key words: aerosol optical depth (AOD), PM2.5 prediction, multi-view interpolation, distributed perception, deep learning, prediction of spatiotemporal migration