Inverse Modelling for Identification of Multiple-Point Releases from Atmospheric Concentration Measurements

Abstract

An inverse modelling methodology is proposed for the identification of multiple-point sources releasing similar tracer, in which influences from the various emissions are merged in each detector’s measurement. The identification is addressed from a limited merged set of atmospheric concentration measurements. The methodology is blended with the natural information provided by the geometry of the monitoring network in terms of the weight functions that interpret the associated visibility/illumination of the monitoring network. The release parameters, locations and intensities of the multiple-point sources are estimated by minimizing the objective function within the least squares framework. The methodology has been successfully applied to identify the two- and three- point simultaneous emissions from synthetic measurements generated by the model without noise or with controlled noise artificially added, and from pseudo-real measurements generated from the Indian Institute of Technology low wind diffusion experiment by combining several of single-point release runs. With the synthetic measurements, all the release parameters are retrieved exactly as those prescribed in all the runs. With the pseudo-real measurements, the release locations are retrieved with an average error of 13 m and intensities are estimated on an average within a factor of 1.5. In a sensitivity analysis, it is shown that the incurred errors in the retrieval of the two- and three-point sources with the pseudo-real data correspond to the 10–30 % Gaussian distributed random noise in the observations. Theoretical and computational comparisons are given between the weighted and non-weighted classical formulations. In addition, an alternative strategy is proposed in order to reduce the computational time required in the source estimation.