Influencing factors of the gas–particle distribution of oxygenated organic molecules in the urban atmosphere and deviation from equilibrium partitioning: a random forest model study

<p>Gas–particle partitioning governs the fate of oxygenated organic molecules (OOMs) and the formation of organic aerosols. We employed a Chemical Ionization Mass Spectrometer equipped with a Filter Inlet for Gases and AEROsol (FIGAERO-CIMS) to measure the gas–particle distribution of OOMs in a winter campaign in the urban atmosphere. The observed gas–particle (<span class="inline-formula"><math xmlns="http://www.w3.org/1998/Math/MathML" id="M1" display="inline" overflow="scroll" dspmath="mathml"><mrow><mi mathvariant="normal">G</mi><mo>/</mo><mi mathvariant="normal">P</mi></mrow></math><span><svg:svg xmlns:svg="http://www.w3.org/2000/svg" width="24pt" height="14pt" class="svg-formula" dspmath="mathimg" md5hash="7f37efa4c6351e6ac36244d5c8977f16"><svg:image xmlns:xlink="http://www.w3.org/1999/xlink" xlink:href="acp-25-9601-2025-ie00001.svg" width="24pt" height="14pt" src="acp-25-9601-2025-ie00001.png"/></svg:svg></span></span>) ratios show a narrower range than the equilibrium <span class="inline-formula"><math xmlns="http://www.w3.org/1998/Math/MathML" id="M2" display="inline" overflow="scroll" dspmath="mathml"><mrow><mi mathvariant="normal">G</mi><mo>/</mo><mi mathvariant="normal">P</mi></mrow></math><span><svg:svg xmlns:svg="http://www.w3.org/2000/svg" width="24pt" height="14pt" class="svg-formula" dspmath="mathimg" md5hash="404ed04d9f5f98b5ad69044c8ac2befc"><svg:image xmlns:xlink="http://www.w3.org/1999/xlink" xlink:href="acp-25-9601-2025-ie00002.svg" width="24pt" height="14pt" src="acp-25-9601-2025-ie00002.png"/></svg:svg></span></span> ratios predicted from saturation mass concentration <span class="inline-formula"><i>C</i>^*</span> and organic aerosol content. The difference between observed and equilibrium <span class="inline-formula"><math xmlns="http://www.w3.org/1998/Math/MathML" id="M4" display="inline" overflow="scroll" dspmath="mathml"><mrow><mi mathvariant="normal">G</mi><mo>/</mo><mi mathvariant="normal">P</mi></mrow></math><span><svg:svg xmlns:svg="http://www.w3.org/2000/svg" width="24pt" height="14pt" class="svg-formula" dspmath="mathimg" md5hash="cdc7eb4a9b39794ffca46e4e87345259"><svg:image xmlns:xlink="http://www.w3.org/1999/xlink" xlink:href="acp-25-9601-2025-ie00003.svg" width="24pt" height="14pt" src="acp-25-9601-2025-ie00003.png"/></svg:svg></span></span> ratios could be up to 10 orders of magnitude, depending on the <span class="inline-formula"><i>C</i>^*</span> parameterization selection. Our random forest models identified relative humidity (RH), aerosol liquid water content (LWC), temperature, and ozone as four influential factors driving the deviations of partitioning from the equilibrium state. Random forest models with satisfactory performance were developed to predict the observed <span class="inline-formula"><math xmlns="http://www.w3.org/1998/Math/MathML" id="M6" display="inline" overflow="scroll" dspmath="mathml"><mrow><mi mathvariant="normal">G</mi><mo>/</mo><mi mathvariant="normal">P</mi></mrow></math><span><svg:svg xmlns:svg="http://www.w3.org/2000/svg" width="24pt" height="14pt" class="svg-formula" dspmath="mathimg" md5hash="98737caae05f7a105cc818a9e204b0a8"><svg:image xmlns:xlink="http://www.w3.org/1999/xlink" xlink:href="acp-25-9601-2025-ie00004.svg" width="24pt" height="14pt" src="acp-25-9601-2025-ie00004.png"/></svg:svg></span></span> ratios. Intrinsic molecule features far outweigh meteorological and chemical composition features in the model's predictions. For a given OOM species, particle chemical composition features, including pH, RH, LWC, organic carbon, potassium, and sulfate, dominate over meteorological and gaseous chemical composition features in predicting the <span class="inline-formula"><math xmlns="http://www.w3.org/1998/Math/MathML" id="M7" display="inline" overflow="scroll" dspmath="mathml"><mrow><mi mathvariant="normal">G</mi><mo>/</mo><mi mathvariant="normal">P</mi></mrow></math><span><svg:svg xmlns:svg="http://www.w3.org/2000/svg" width="24pt" height="14pt" class="svg-formula" dspmath="mathimg" md5hash="26ce69c41bf84e740a9ba5db56516dc1"><svg:image xmlns:xlink="http://www.w3.org/1999/xlink" xlink:href="acp-25-9601-2025-ie00005.svg" width="24pt" height="14pt" src="acp-25-9601-2025-ie00005.png"/></svg:svg></span></span> ratios. We identified the positive or negative effects as well as the sensitive ranges of these influential features using SHapley Additive exPlanations (SHAP) analysis and curve fitting with a generalized additive model (GAM). Our models found that temperature does not emerge as a significant factor influencing the observed <span class="inline-formula"><math xmlns="http://www.w3.org/1998/Math/MathML" id="M8" display="inline" overflow="scroll" dspmath="mathml"><mrow><mi mathvariant="normal">G</mi><mo>/</mo><mi mathvariant="normal">P</mi></mrow></math><span><svg:svg xmlns:svg="http://www.w3.org/2000/svg" width="24pt" height="14pt" class="svg-formula" dspmath="mathimg" md5hash="be91662a34371b29d0e2554cf6a31b03"><svg:image xmlns:xlink="http://www.w3.org/1999/xlink" xlink:href="acp-25-9601-2025-ie00006.svg" width="24pt" height="14pt" src="acp-25-9601-2025-ie00006.png"/></svg:svg></span></span> ratios, suggesting that other factors, most likely associated with particle composition, inhibit the gas–particle partitioning of OOMs in response to temperature change.</p>