Airpred: A Flexible R Package Implementing Methods for Predicting Air Pollution

Large epidemiological studies have shown that exposure to air pollution, in particular fine particulate matter (PM _2.5 ), is harmful to human health. However, air pollution monitors which measure air pollutant concentrations are sparsely located, excluding large portions of the population, in particular non-urban populations, from studies. One approach to resolving this issue has been developing models to predict local PM _2.5 , NO ₂ , and ozone in unmonitored areas based on satellite, meteorological, and land use data. These prediction models are typically developed using large amounts of input data and are highly computationally intensive. We have developed a flexible R package that allows environmental health researchers to design and train spatio-temporal models capable of predicting multiple pollutants, including PM _2.5 . We utilize H2O, an open source big data R platform, to achieve both performance and scalability when used in conjunction with cloud or cluster computing systems.