Sensing Air Quality: Spatiotemporal Interpolation of Air Pollution Data and Visualization
Proceedings of the Association of American Geographers Annual Meeting
In order to improve current spatiotemporal interpolation methods for public health applications, we combined the spatiotemporal interpolation methods and several machine learning methods, employed the efficient k-d tree structure to store data, and implement our methods on Apache Spark Platform. The preliminary results demonstrate a great computation ability and scalability of our method, which outperforms the previous work. In addition, we develop a visualization approach using spatiotemporal interpolation methods that allows real-time summarization and presentation of real-time air pollution data across the contiguous United States. A web application is developed to interpolate and visualize in the real-time variation of ambient air pollution using data from the U.S. Environmental Protection Agency (EPA)'s AirNow program. Future research will investigate spatial uncertainty, improve the interpolation accuracy and computation efficiency, as well as establish associations between air pollution exposure and adverse health effects.
Zhou, Xiaolu, Lixin Li, Weitian Tong, Jason Franklin.
"Sensing Air Quality: Spatiotemporal Interpolation of Air Pollution Data and Visualization."
Proceedings of the Association of American Geographers Annual Meeting Boston, MA: Association of American Geographers.