Sensing air quality: Spatiotemporal interpolation and visualization of real-time air pollution data for the contiguous United States

Abstract

Air pollution has been a major risk to public health. It is imperative to monitor the spatiotemporal patterns of regional air pollution. However, air pollution data are often collected at a limited set of locations and archived at different times. In order to monitor pollution in the continuous space-time domain, traditional interpolation methods tend to treat space and time separately when estimating the pollution data at un-sampled locations and times. But such methods may not be able to yield satisfactory interpolation results. In addition, a vast amount of air quality data have been available to request in real time with the advance of modern sensors, but existing approaches are limited in their ability to process large data sets and support real-time visualizations. In this research, we investigate and compare several spatiotemporal interpolation methods with the goal to conduct interpolation on real-time air pollution data at a large geographic area. Both accuracy and efficiency are evaluated in this study. Based on the findings, we developed a visualization approach using a proposed method that allows real-time summarization and presentation of hourly air pollution data across the contiguous United States. A web application is developed that provides a portal to the public to visualize air quality.