Journal of Broadcast Engineering (방송공학회논문지)

The Korean Institute of Broadcast and Media Engineers (한국방송∙미디어공학회)
권호 표지
DOI QR코드

DOI QR Code

Early Prediction of Fine Dust Concentration in Seoul using Weather and Fine Dust Information

기상 및 미세먼지 정보를 활용한 서울시의 미세먼지 농도 조기 예측

  • 이한주 (세종대학교 일반대학원 소프트웨어학과) ;
  • 지민규 ((주)셀버스) ;
  • 김학동 (세종대학교 일반대학원 디지털콘텐츠학과) ;
  • 전태흘 (세종대학교 일반대학원 소프트웨어융합학과) ;
  • 김청원 (세종대학교 소프트웨어학과)
  • Received : 2023.03.28
  • Accepted : 2023.05.18
  • Published : 2023.05.30
Download PDF
⟨ Previous Next ⟩

Abstract

Recently, the impact of fine dust on health has become a major topic. Fine dust is dangerous because it can penetrate the body and affect the respiratory system, without being filtered out by the mucous membrane in the nose. Since fine dust is directly related to the industry, it is practically impossible to completely remove it. Therefore, if the concentration of fine dust can be predicted in advance, pre-emptive measures can be taken to minimize its impact on the human body. Fine dust can travel over 600km in a day, so it not only affects neighboring areas, but also distant regions. In this paper, wind direction and speed data and a time series prediction model were used to predict the concentration of fine dust in Seoul, and the correlation between the concentration of fine dust in Seoul and the concentration in each region was confirmed. In addition, predictions were made using the concentration of fine dust in each region and in Seoul. The lowest MAE (mean absolute error) in the prediction results was 12.13, which was about 15.17% better than the MAE of 14.3 presented in previous studies.

최근 미세먼지가 건강에 미치는 영향은 큰 화두가 되고 있다. 미세먼지는 코의 점막에 걸러지지 않고 인체 내부까지 침투하여 호흡기에 영향을 미치기 때문에 위험하다. 미세먼지는 산업과 직접적으로 연관되어있기 때문에 미세먼지를 제거하는 것은 사실상 불가능하다. 따라서 미세먼지 농도를 사전에 예측할 수 있다면 사전 조치를 취해 인체에 미치는 영향을 줄일 수 있다. 미세먼지는 하루 600km 이상 이동할 수 있는 특성을 가진다. 이러한 특성으로 인해 미세먼지는 인접 구뿐만 아니라 멀리 떨어져있는 구에도 직접적인 영향을 미친다. 본 논문에서는 풍향, 풍속 데이터와 시계열 예측 모델을 이용하여 서울특별시의 미세먼지 농도를 예측하고, 서울특별시의 미세먼지 농도와 지역별 미세먼지 농도의 상관관계를 확인했다. 또한, 각 지역별 미세먼지 농도와 서울특별시의 미세먼지 농도를 이용하여 예측을 진행했다. 예측 결과에서 가장 낮았던 MAE(평균 절대 오차)는 12.13으로 선행연구에서 제시된 MAE인 14.3 보다 약 15.17% 더 예측성능이 향상된 것을 확인했다.

Keywords

Acknowledgement

This work is financially supported by Korea Ministry of Environment (MOE) as 「Graduate School specialized in Climate Change.

References

  1. D. Shin, "Health Effects of Ambient Particulate Matter", Joarnal of Korean Medical Association, Vol.50, No.2, , pp.175-182, 2007. doi: https://doi.org/10.5124/jkma.2007.50.2.175
  2. G. Shin, J. Kim, and Y. Choi, "A Study on the Data Model Design of Fine Dust Related Disease", Jounal of The Korea Society of Information Technology Policy & Management (ITPM), Vol.10, No.1. pp.655-659, 2018. http://scholarworks.bwise.kr/ssu/handle/2018.sw.ssu/31956
  3. D. Loomis et al., "The carcinogenicity of outdoor air pollution.", Lancet Oncology, Vol.14, No.13, pp.1262 ,2013. doi: https://doi.org/10.1016/S1470-2045(13)70487-X
  4. J. Park, Y. Park, E. Lee and K. Lee, "Analysis on the Effects of Particular Matter Distribution on the Number of Outpatient Visits for Allergic Rhinitis", Health Policy and Management Vol.30, No.1, pp.50-61, 2020. doi: https://doi.org/10.4332/KJHPA.2020.30.1.50
  5. W. Jung, "South Korea's Air Pollution: Gasping for Solutions", Policy Brief, Vol.199, pp.1-4, 2017.
  6. S. Jeon, and Y. Son, "Prediction of fine dust PM 10 using a deep neural network model", The Korean journal of applied statistics, Vol.31, No.2, pp.265-285, 2018 doi: https://doi.org/10.5351/KJAS.2018.31.2.265
  7. P. Soh, J. Chang, and J. Huang, "Adaptive deep learning-based air quality prediction model using the most relevant spatial-temporal relations", Ieee Access, Vol.6, pp.38186-38199, 2018. doi: https://doi.org/10.1109/ACCESS.2018.2849820
  8. S. Kim, J. Lee, and J. Seo, "Deep-dust: Predicting concentrations of fine dust in Seoul using LSTM", arXiv preprint arXiv:1901.10106, 2019. doi: https://doi.org/10.48550/arXiv.1901.10106
  9. Z. Joharestani et al.,"PM prediction based on random forest, XGBoost, and deep learning using multisource remote sensing data", Atmosphere, Vol.10, No.7, pp.373, 2019. doi: https://doi.org/10.3390/atmos10070373
  10. T. Xayasouk, H. Lee, and G. Lee, "Air pollution prediction using long short-term memory (LSTM) and deep autoencoder (DAE) models", Sustainability, Vol.12, No.6, pp. 2570, 2020. doi: https://doi.org/10.3390/su12062570
  11. K. Lee, W. Hwang, and M. Choi, "Design of a 1-D CRNN Model for Prediction of Fine Dust Risk Level", The Society of Digital Policy and Management, Vol.19, No.2, pp.215-220, 2021. doi: https://doi.org/10.14400/JDC.2021.19.2.215
  12. H. Kim, and T. Moon, "Machine learning-based Fine Dust Prediction Model using Meteorological data and Fine Dust data", Vol.24, No.1, pp.92-111, 2021. doi: https://doi.org/10.11108/kagis.2021.24.1.092
  13. G. Park, "Discovering a fine dust pathway via directed acyclic graphical models", Journal of the Korean Data & Information Science Society, Vol.30, No.1, pp.67-76, 2019. doi: https://doi.org/10.7465/jkdi.2019.30.1.67
  14. A. Dairi et al., "Integrated multiple directed attention-based deep learning for improved air pollution forecasting", IEEE Transactions on Instrumentation and Measurement, Vol.70, pp.1-15, 2021. doi: https://doi.org/10.1109/TIM.2021.3091511
  15. S. Jo, M. Jeong, J. Lee, I. Oh, and Y. Han, "Analysis of Correlation of Wind Direction/Speed and Particulate Matter(PM10) and Prediction of Particulate Matter Using LSTM", Proceeding of Korean Institute of Information Scientists and Engineers, pp.1649-1651, 2020.
  16. H. Zhou et al., "Informer: Beyond efficient transformer for long sequence time-series forecasting", Proceedings of the AAAI conference on artificial intelligence, Vol.35, No.12, pp.11106-111115, 2021. doi: https://doi.org/10.1609/aaai.v35i12.17325