Microenvironment Tracker (MicroTrac) model to estimate time-location of individuals for air pollution exposure assessments: model evaluation using smartphone data

Background: A critical aspect of air pollution exposure assessments is determining the time spent in various microenvironments (ME), which can have substantially different pollutant concentrations. We previously developed and evaluated a ME classification model, called Microenvironment Tracker (MicroTrac), to estimate time of day and duration spent in eight MEs (indoors and outdoors at home, work, school; inside vehicles; other locations) based on input data from global positioning system (GPS) loggers. Objective: In this study, we extended MicroTrac and evaluated the ability of using geolocation data from smartphones to determine the time spent in the MEs. Method: We performed a panel study, and the MicroTrac estimates based on data from smartphones and GPS loggers were compared to 37 days of diary data across five participants. Results: The MEs were correctly classified for 98.1% and 98.3% of the time spent by the participants using smartphones and GPS loggers, respectively. Significance: Our study demonstrates the extended capability of using ubiquitous smartphone data with MicroTrac to help reduce time-location uncertainty in air pollution exposure models for epidemiologic and exposure field studies.