室内生物气溶胶中微生物浓度的预测模型——以青岛为例

The deposition of bioaerosols in the respiratory system may induce respiratory diseases. Given that modern populations spend approximately 90% of their time indoors, predicting indoor bioaerosol concentrations is crucial for health risk prevention and control. This study collected samples from offices and laboratories in three universities in Qingdao. Using DAPI staining and LIVE/DEAD~® BacLightTM staining combined with fluorescence microscopy, we measured the concentrations of total airborne microorganisms(TAMs), viable bacteria(VBs), and non-viable bacteria(NVBs). The experimental results showed that the concentration of VBs ranged from 0.10×104～1.97×104 cells/m3, while NVBs and TAMs ranging from 3.11×104～12.63×104 cells/m3 and 1.09×104～9.42×105 cells/m3, respectively. To investigate the relationship between indoor TAMs concentrations and environmental factors, we compared the performance of multiple linear regression(MLR), Ridge regression, and Poisson regression models. The results revealed that MLR had limited predictive capability, Ridge regression could not adequately explain all variables, while the Poisson regression model provided a more accurate explanation and prediction of bioaerosol concentrations. The analysis indicated that temperature, relative humidity(RH), particulate matter concentration, and human activities collectively influence microbial concentrations, with PM10 and RH showing the most significant effects. The low-cost predictive model developed in this study provides a theoretical foundation for indoor air quality management.