Improvements of response surface modeling with self-adaptive machine learning method for PM 2.5 and O 3 predictions

• Published in: Journal of environmental management 2022-02, Vol.303, p.114210
• Main Authors: Li, Jinying, Dai, Youzhi, Zhu, Yun, Tang, Xiangbo, Wang, Shuxiao, Xing, Jia, Zhao, Bin, Fan, Shaojia, Long, Shicheng, Fang, Tingting
• Format: Article
• Language: English
• Subjects: Air Pollutants - analysis; Air Pollution - analysis; Environmental Monitoring; Machine Learning; Ozone - analysis; Particulate Matter - analysis; Pilot Projects
• ISSN: 1095-8630
• DOI: 10.1016/j.jenvman.2021.114210

Quickly quantifying the PM or O response to their precursor emission changes is a key point for developing effective control policies. The polynomial function-based response surface model (pf-RSM) can rapidly predict the nonlinear response of PM and O to precursors, but has drawbacks of overload computation and marginal effects (relatively larger prediction errors under strict control scenarios). To improve the performance of pf-RSM, a novel self-adaptive RSM (SA-RSM) was proposed by integrating the machine learning-based stepwise regression for establishing robust models to increase the computational efficiency and the collinearity diagnosis for reducing marginal effects caused by overfitting. The pilot study case demonstrated that compared with pf-RSM, SA-RSM can effectively reduce the training number by 70% and 40% and the fitting time by 40% and 52%, and decrease the prediction error by 49% and 74% for PM and O predictions respectively; moreover, the isopleths of PM or O as a function of their precursors generated by SA-RSM were more similar to those derived by chemical transport model (CTM), after successfully addressing the marginal effect issue. With the improved computation efficiency and prediction performance, SA-RSM is expected as a better scientific tool for decision-makers to make sound PM and O control policies.