• Deep Learning-based Prediction of PM10 Fluctuation from Gwanak-gu Urban Area, Seoul, Korea
  • Han-Soo Choi1 ·Myungjoo Kang2 ·Yong Cheol Kim3 ·Hanna Choi3, *

  • 1 Research Institute of Mathematics, Seoul National University, Seoul 08826, Korea
    2 Department of Mathematical Sciences, Seoul National University, Seoul 08826, Korea
    3 Korea Institute of Geoscience and Mineral Resources, Daejeon 34132, Korea 

  • 서울 관악구 도심지역 미세먼지(PM10) 관측 값을 활용한 딥러닝 기반의 농도변동 예측
  • 최한수1 ·강명주2 ·김용철3 ·최한나3, *

  • 1 서울대학교 수학연구소
    2 서울대학교 수리과학부
    3 한국지질자원연구원


Since fine dust (PM10) has a significant influence on soil and groundwater composition during dry and wet deposition processes, it is of a vital importance to understand the fate and transport of aerosol in geological environments. Fine dust is formed after the chemical reaction of several precursors, typically observed in short intervals within a few hours. In this study, deep learning approach was applied to predict the fate of fine dust in an urban area. Deep learning training was performed by combining convolutional neural network (CNN) and recurrent neural network (RNN) techniques. The PM10 concentration after 1 hour was predicted based on three-hour data by setting SO2, CO, O3, NO2, and PM10 as training data. The obtained coefficient of determination value, R2, was 0.8973 between predicted and measured values for the entire concentration range of PM10, suggesting deep learning method can be developed into a reliable and viable tool for prediction of fine dust concentration. 

Keywords: fine dust (PM10), precursor, deep learning, convolutional neural network, recurrent neural network, prediction

This Article

  • 2020; 25(3): 74-83

    Published on Sep 30, 2020

  • 10.7857/JSGE.2020.25.3.074
  • Received on Aug 26, 2020
  • Revised on Sep 3, 2020
  • Accepted on Sep 22, 2020

Correspondence to

  • Hanna Choi
  • Korea Institute of Geoscience and Mineral Resources, Daejeon 34132, Korea 

  • E-mail: pythagoras84@kigam.re.kr