Machine Learning Reveals the Parameters Affecting the Gaseous Sulfuric Acid Distribution in a Coastal City: Model Construction and Interpretation

• DOI: 10.1021/acs.estlett.3c00170
• Publication Date: 2023-04-12T19:17:30Z
• Authors: Chen Yang; Hesong Dong; Yuping Chen; Yonghong Wang; Xiaolong Fan; Yee Jun Tham; Gaojie Chen; Lingling Xu; Ziyi Lin; Mengren Li; Youwei Hong; Jinsheng Chen
• Keywords: Methanesulfonic acid; Cloud condensation nuclei; Sink (geography); Dimethyl sulfide

Although the fundamental mechanisms of atmospheric new particle formation events are largely associated with gaseous sulfuric acid monomer (SA), parameters affecting SA generation and elimination remain unclear, especially in coastal areas where certain sulfur-containing precursors abundant. In this study, we utilized machine learning (ML) combination field observations to map link between influencing parameters. The developed random forest (RF) model performed well creating simulations an R2 0.90, significant factors were ultraviolet, methanesulfonic (MSA), SO2, condensation sink, relative humidity descending order. Among five factors, MSA served as indicator for species from marine emissions. black box ML was broken determine marginal contribution these output using partial dependence plots centered-individual conditional expectation plots. These results indicated that had a positive impact on performance RF model, co-occurring relationship observed during nocturnal period. Our findings reveal emitted environment have should be considered areas.