WeMott, Sherry D., authorMagzamen, Sheryl, advisorKoslovsky, Matthew, committee memberRojas-Rueda, David, committee member2022-08-292022-08-292022https://hdl.handle.net/10217/235660Though individuals in the United States spend a majority of their time indoors, epidemiologic studies often use ambient air pollution data for exposure assessment. We used several modeling approaches to predict indoor black carbon (BC) from outdoor BC and housing characteristics to support future efforts to estimate personal air pollution exposure given time spent indoors. Households from the Healthy Start cohort in Denver, CO were recruited to host two paired indoor/outdoor low-cost air samplers for one-week sampling periods during spring 2018, summer 2018, and winter 2019. Participants also completed questionnaires about housing characteristics like building type, flooring, and use of heating and cooling systems. Sampled filters were analyzed for BC using transmissometry. Ridge, Lasso and multiple regression techniques were used to build the best predictive model of indoor BC given the available set of covariates. Leave-one-out cross-validation was used to assess the predictive accuracy of each model. A total of 27 households participated in the study, and BC data were available for 39 filters. We had limited comparable data on seasonality as winter data were excluded from the analysis due to high variability and low confidence in outdoor measurements. Shortened runtimes and other performance issues suggest insufficient weatherproofing of our monitors for low temperatures. Of the three modeling approaches, Ridge LSE showed the best predictive performance (MPSE 0.50). The final inference model included the following covariates: outdoor PM2.5, outdoor BC, hard floors, and pets in the home (adj. R2=0.27). These factors accounted for approximately 27% of the variability in indoor BC concentrations measured in Denver, CO homes. In the absence of personal monitoring, household characteristics and time-activity patterns may be used to calibrate ambient air pollution concentrations to the indoor environment for improved estimation of personal exposure.born digitalmasters thesesengCopyright and other restrictions may apply. User is responsible for compliance with all applicable laws. For information about copyright law, please see https://libguides.colostate.edu/copyright.birth cohortindoor air pollutionpredictive modelingblack carbonambient air pollutionPM2.5Text