This study evaluates the accuracy of measuring trace gas emission from treatment lagoons using backward Lagrangian stochastic (bLs) technique.  The bLs technique was originally developed for relatively homogeneous terrains without any obstacles causing significant windflow disturbance.  The errors associated with using the bLs technique for emission sources with obstacles such as fences or building can be minimized by collecting downwind concentration data far away from the obstacles.  Because treatment lagoons in the Southeast region area are usually surrounded by tree lines, it may not be possible to obtain downwind concentration data with enough fetch.  Instead, we assume that the microclimate condition within the treatment lagoon boundary can be isolated and its wind turbulence adequately characterized by a 3-D sonic anemometer installed within the treatment lagoons. This assumption allows us to use the bLs technique; however, its accuracy in this type of application is not known at this time.  In this study, we compared the emission rate of CH4 estimated using the WindTrax, a computer based bLs technique, with known mass rates released through a 47m x 47 m floating pipe network on a 65m x 65m reservoir. Tunable diode laser absorption spectrometers (TDL) were used to collect path integrate concentration (PICs) data.  Experimental variables include laser height and location, berm height, and release rate.  The experiments are still on-going and the results will be presented at the meeting.