Obtaining accurate source information about the release of a contaminant is a crucial step to the mitigation of the release. In addition, accurate meteorological data is essential for driving the atmospheric transport and dispersion (AT&D) models used to predict subsequent puff position and spread. Our previous work has focused on using basic models such as the Gaussian puff equation to predict the future state of the contaminant puff. Here, we substitute a sophisticated model, SCIPUFF, to predict the transport and dispersion of the contaminant. The observation data are generated using the same model via an identical twin experimental setup. To these concentration data we then apply a genetic algorithm to back-calculate the source and meteorological information. In addition, we analyze the computing requirements for the analysis and compare these results with those for the basic Gaussian puff.