다중선형회귀와 인공신경망을 이용한 2015년 이전 PM2.5 일일 평균 수치 추정 방법론 제안
Daily PM2.5 Estimation using Multiple Linear Regression and Artificial Neural Networks Before 2015
1
조회
0
다운로드
0
인용
0
참고문헌
키워드
초록
Since 2015, the PM2.5 measurement data has been publicly available nationwide in South Korea, but its use is restricted to after 2015, unlike other air pollutants. To overcome this limitation, multiple linear regression and artificial neural network models were developed to predict the daily average PM2.5 values in South Korea before 2015. The daily data of air pollution measurement(SO2, CO, O3, NO2, PM10) and meteorological observation data (temperature, humidity, wind speed, atmospheric pressure, precipitation, snowfall) were used as input variables to develop regional prediction models for five regions(Seoul, Incheon, Gwangju, Daejeon, Ulsan) and a national prediction model. The models were developed and validated using the air pollution measurement data after 2015, and applied to predict PM2.5 values before 2015. The multiple linear regression model showed R2 values of 0.80 nationwide, 0.73 in Seoul, and 0.67 in Incheon, which enabled estimation of daily average PM2.5 values before 2015. The artificial neural network model showed good prediction power with R2 values of 0.79 in Gwangju, 0.81 in Daejeon, and 0.72 in Ulsan. The regional prediction models showed good prediction power in most regions, and both the multiple linear regression and artificial neural network models showed good prediction power.
Abstract
Since 2015, the PM2.5 measurement data has been publicly available nationwide in South Korea, but its use is restricted to after 2015, unlike other air pollutants. To overcome this limitation, multiple linear regression and artificial neural network models were developed to predict the daily average PM2.5 values in South Korea before 2015. The daily data of air pollution measurement(SO2, CO, O3, NO2, PM10) and meteorological observation data (temperature, humidity, wind speed, atmospheric pressure, precipitation, snowfall) were used as input variables to develop regional prediction models for five regions(Seoul, Incheon, Gwangju, Daejeon, Ulsan) and a national prediction model. The models were developed and validated using the air pollution measurement data after 2015, and applied to predict PM2.5 values before 2015. The multiple linear regression model showed R2 values of 0.80 nationwide, 0.73 in Seoul, and 0.67 in Incheon, which enabled estimation of daily average PM2.5 values before 2015. The artificial neural network model showed good prediction power with R2 values of 0.79 in Gwangju, 0.81 in Daejeon, and 0.72 in Ulsan. The regional prediction models showed good prediction power in most regions, and both the multiple linear regression and artificial neural network models showed good prediction power.
참고문헌 (11)
- Jang, AS, "Impact of particulate matter on health", Journal of the Korean Medical Association, 57(9), pp.763~768 (2014)
- Kim, Y., "Different adverse effects of air pollutants on dry eye disease: Ozone, PM2.5, and PM10", Environmental Pollution, 265, pp.115039~ (2020)
- Lee, Y. H., "Using visibility to estimate PM2. 5concentration trends in Seoul and Chuncheon from 1982 to 2014", Journal of Korea Society for Atmospheric Enviroment, 34, pp.1156~1165 (2018)
- Lee, S. H., "An approach to estimating national-scale annual-average concentrations of PM2. 5 before 2015when national air quality monitoring data are available in South Korea", Journal of Korea Society for Atmospheric Enviroment, 34, pp.6806~6821 (2018)
- ANNUAL REPORT OF AIR QUALITY IN KOREA, "Ministry of Environment", Republic of Korea Ministry of Environment, 8~ (2020)
- Jee, J. B., "An Analysis of Meteorological Characteristics for Fine Particles on the Korean Peninsula during Wintertime, 2015-2021", Journal of Korea Society for Atmospheric Enviroment, 38(3), pp.394~413 (2022)
- Zheng, M., "Seasonal trends in PM2.5 source contributions in Beijing, China", Atmospheric Environment, 39, pp.223967~223976 (2005)
- Huang, G., "PM2.5 concentration forecasting at surface monitoring sites using GRU neural network based on empirical mode decomposition", Sicience of The Total Environment, 768, pp.144516~ (2021)
- Gunther, F., "Neuralnet : training of neural networks", The R Journal, 2(1), pp.30~ (2010)
- Shmueli, G., "Data mining for business analytics:concepts techniques and applications in R" (2017), John Wiley & Sons
- Huh, J. W., "Development of a Prediction Model for Daily PM2.5 in Republic of Korea by Using an Artificial Neural Network", Applied Sciences, 13(6), pp.3575~ (2023)