|
Particle and Aerosol Research
Vol. 21, No. 3, September 2025, Pages 81-92
|
ISSN : 1738-8716 (Print)
ISSN : 2287-8130 (Online)
|
Air Quality Prediction of Chuncheon as Gangwon Province is promoted to Special Self-Governing Province
Seok Ho Leea), Na Rae Choia),b) *
a)Department of Environmental Engineering, Kangwon National University
a),b)Gangwon particle pollution Research and Management Center, Region of Korea, Kangwon National University
*Corresponding author.
Tel.: 033-250-6353, E-mail: narae@kangwon.ac.kr
Received 03 April 2025, Revised 05 June 2025, Accepted 03 September 2025, Available online 30 September 2025
http://dx.doi.org/10.11629/jpaar.2025.21.3.081
Abstracts
This study aims to predict air quality changes in Chuncheon City of Gangwon Special Self-Governing Province and analyze the impact of urban development indicators on air pollutant concentrations. Using a Random Forest model based on data from 2000 to 2024, we predicted the concentrations of PM10, NO2, and O3 until 2034. The model demonstrated high reliability with R2 values of 0.92, 0.85, and 0.83 for PM10, NO2, and O3, respectively, and Mean Absolute Percentage Errors (MAPE) of 7.07%, 2.60%, and 9.59%. Prediction results indicate that PM10 concentrations are expected to exceed the air quality standard (46 §¶/m3) by 2032, while NO2 and O3 concentrations will remain within acceptable limits. Scenario analyses based on 10%, 20%, and 30% increases in population, GRDP, and vehicle registration showed only minor variations from baseline predictions, suggesting the model predominantly extrapolates from long-term temporal trends in the training data. Feature importance analysis revealed that vehicle registration numbers had the greatest impact on PM10 (0.65) and NO2 (0.40) concentrations, while GRDP had the strongest influence on O3 levels (0.45). These findings suggest that future air quality management in Chuncheon should prioritize traffic volume control due to its tourism industry characteristics while systematically managing ozone precursors during industrial development. The limitations of this study include the potential inability of the model to fully reflect the impacts of rapid institutional changes, highlighting the need for more adaptive modeling methodologies in future research.
Keywords
Air Quality Prediction, Random Forest, Machine Learning, Urban Development, Gangwon Special Self-Governing Province
File
02-³í¹® 21±Ç3È£(ÃÖ³ª·¡)-5-°Ô½Ã¿ë.pdf
Download