Spatiotemporal Modeling and Monitoring of Atmospheric Hazardous Emissions Using Sensor Networks

Abstract

A spatiotemporal methodology is presented for the analysis and visualization of atmospheric emissions in a metropolitan area. Numerical transport and dispersion models are used to build a library of time-dependent emissions of hazardous gases under various atmospheric conditions and from multiple potential sources in Washington DC. This library comprises representative emergency events that may involve natural or man-made hazardous emissions. To represent and analyze the events of this library we use the model of the spatiotemporal helix, which provides concise summaries of complex spatiotemporal events. We demonstrate the ability to compare emerging situations to library entries in order to predict their future evolution, thus recognizing potentially hazardous conditions early in their development.