Optimized Sampling for Temporally Varying Source Emissions

Coupled mesoscale atmospheric and Lagrangian particle dispersion models are used routinely to estimate source emissions from downwind effluent measurements. Successful characterization of any source is dependent upon the ability to detect an effluent plume above some minimum detectable level, with sufficient signal-to-noise resolution in order differentiate the effluent above an environmental background concentration. As a result, the locations where measurements are made should be selected to provide the best opportunity for obtaining collections at sufficient temporal and spatial resolution in order to describe the emission process. At downwind distances where a plume has become well mixed within the environment, longer duration sample collections can be utilized to quantify the source emission. Quantification using long duration samples, or alternatively, averaging multiple short duration samples reduces sensitivity to random model errors and unknown environmental conditions, although these have disadvantages when source emissions vary with time. Lower concentration emissions require sampling closer to the source and over shorter durations in order to provide a sufficient signal to noise ratio for analysis. Decreasing distance from the source, and temporal variations in emissions complicate source term estimates obtained from downwind measurements. Sensitivity in source characterization will be analyzed to determine the number of measurements and optimum collection locations in order to integrate independent discrete samples.