The rapid growth of urban areas and its population caused increase in interest and research efforts to better understand dispersion of pollutants in these areas. Besides efforts to predict concentrations of plume for a known source parameters and meteorological conditions (forward problem), in recent years a lot of work was done on creating the quick and efficient way to solve the inverse problem. The goal is to estimate source location and strength by using the available measurement data i.e. concentration measurements and knowledge of wind conditions. A lot of techniques were developed in last few years that reasonably well solve this problem, but most of them require prohibitively large amount of time in order to be applied in real life situations. Since our goal is to develop methodology for use in emergency situations, we use collector footprint approach to backtrack the source. Sensor positions are used as new source locations and after wind field was reversed the source strength was calculated. The source location was estimated by overlapping the intersecting backward plumes. Method was tested on real world data obtained during the filed campaign Joint Urban 2003 conducted in Oklahoma City. This enabled us to test the performance of the model when uncertainties are present due to measurement errors as well as due to the nature of flow in inhomogeneous roughness canopies under real atmospheric conditions.