|
Environmental sensors are employed for a variety of industrial safety and disaster response applications to detect emissions of concern that may indicate leaks, lack of regulatory compliance or illicit activities. Effusive emissions of interest include precursors, reaction intermediates and products, decomposition products, growth media, fermentation metabolites, solvents, and cleaners used in a variety of processes of interest. For all volatile organic compound/semivolatile organic compound (VOC/sVOC) sensors, a vital question is sensitivity to false alarms resulting from the presence of background clutter. To assess this sensitivity, “normal background” must be understood in terms of parameters such as: physical position (latitude, longitude, nearby geographical feature such as lakes, mountains, deserts); temporally varying factors (season, time of day, weather); and industrial features (surrounding population density, nearby presence of known or unknown activities that emit characteristic patterns of chemical effluent). It is only by understanding the scope and variability of the normal background that one can determine if it is possible for a given sensor modality to discriminate a specific industry or classes of industries from normal background either absolutely, through specific chemical indicators or indicator patterns, or as change detection, if the baseline of an area or area class is known. The Intelligence Advanced Research Projects Activity (IARPA) is conducting full VOC/sVOC mapping of a number of representative sites with varying climates, urban density, and local geographical features, as well as nearby contributing industries such as gas stations, a bakery, a brewery, a sewer treatment plant, and agricultural activities. This presentation will discuss the choice of sites, sampling methodology used, and preliminary comparative results.
|
