Random Forest Estimation and Trend Analysis of PM[sub.2.5] Concentration over the Huaihai Economic Zone, China

• Published in: Sustainability 2022-07, Vol.14 (14)
• Main Authors: Li, Xingyu, Li, Long, Chen, Longgao, Zhang, Ting, Xiao, Jianying, Chen, Longqian
• Format: Article
• Language: English
• Subjects: Distribution; Environmental aspects; Forecasts and trends; Particles
• ISSN: 2071-1050; 2071-1050
• DOI: 10.3390/su14148520

Consisting of ten cities in four Chinese provinces, the Huaihai Economic Zone has suffered serious air pollution over the last two decades, particularly of fine particulate matter (PM[sub.2.5]). In this study, we used multi-source data, namely MAIAC AOD (at a 1 km spatial resolution), meteorological, topographic, date, and location (latitude and longitude) data, to construct a regression model using random forest to estimate the daily PM[sub.2.5] concentration over the Huaihai Economic Zone from 2000 to 2020. It was found that the variable expressing time (date) had the greatest characteristic importance when estimating PM[sub.2.5]. By averaging the modeled daily PM[sub.2.5] concentration, we produced a yearly PM[sub.2.5] concentration dataset, at a 1 km resolution, for the study area from 2000 to 2020. On comparing modeled daily PM[sub.2.5] with observational data, the coefficient of determination (R[sup.2]) of the modeling was 0.85, the root means square error (RMSE) was 14.63 μg/m[sup.3], and the mean absolute error (MAE) was 10.03 μg/m[sup.3]. The quality assessment of the synthesized yearly PM[sub.2.5] concentration dataset shows that R[sup.2] = 0.77, RMSE = 6.92 μg/m[sup.3], and MAE = 5.42 μg/m[sup.3]. Despite different trends from 2000–2010 and from 2010–2020, the trend of PM[sub.2.5] concentration over the Huaihai Economic Zone during the 21 years was, overall, decreasing. The area of the significantly decreasing trend was small and mainly concentrated in the lake areas of the Zone. It is concluded that PM[sub.2.5] can be well-estimated from the MAIAC AOD dataset, when incorporating spatiotemporal variability using random forest, and that the resultant PM[sub.2.5] concentration data provide a basis for environmental monitoring over large geographic areas.