Examining the feasibility of prediction models by monitoring data and management data for bioaerosols inside office buildings

Exposure to bioagents can cause several health problems, including acute allergies, infectious diseases, and myctoxicosis. Nevertheless, all conventional methods for measuring airborne bioaerosols have significant limitations such as high cost, prolonged measurement time, and discontinuous measurements.This work develops a simple and cost-effective method for indoor airborne bioaerosols that uses monitoring data such as coarse particle (PM10), fine particle (PM2.5), and carbon dioxide (CO2) concentrations, and temperature (Temp), and relative humidity (RH) both indoors and outdoors. Some IAQ management data, such as the number of stories, air ventilation types, air exchange rate, potential indoor particulate sources, and population density were quantified in this study. Both monitoring data and management data are considered simultaneously, and multiple linear regression and nonlinear regression analyses are applied to develop prediction models for bacteria and fungi concentrations in office buildings. The indoor and outdoor air qualities of 37 office buildings in Taipei, Taiwan were sampled to develop the prediction models for buildings in Taipei Metropolitan.Results showed that the predictions of a single office building were better than those of all office buildings in the city. The prediction using multiple linear regression models performed best for both indoors bacteria and fungi concentrations. Furthermore, analytical results show that the prediction with both monitoring and management data inputs were better than with monitoring data only. This real-time prediction model can serve as a simple and cost-effective tool for predicting bioaerosol concentrations to identify and prevent IAQ problems.

► We develop a simple and cost-effective method for indoor airborne bioaerosols based on the monitoring data and management data.
►  The predicted concentrations were better for a single office building.
► The predicted concentrations were better when multiple linear regression models were used.
► The office buildings were best when monitoring data and management data were utilized.