A variety of measurements of methane in the Martian atmosphere have been made over the past 15 years, showing wildly varying indications of methane abundance, location and lifetime in the Martian atmosphere. Attempts have been made to use numerical tools such as general circulation models (GCMs) to identify source locations and timing of methane releases, but these remain inconclusive under the current approach of forward-trajectory plume modeling. Here we present results using a novel, complementary method of localizing methane surface sources by modeling passive tracer trajectories backwards in time from the locations where observations of atmospheric methane have been made. Such back-trajectory modeling employs both GCM modeled winds and a Lagrangian particle dispersion model to isolate potential upwind sources of the observed signals. This approach avoids many of the pitfalls inherent in forward-trajectory modeling approaches such as numerical diffusion and subgrid-scale motion which cannot be captured in the Eulerian framework of a GCM. We have chosen to focus on localization of the detection of methane by the Planetary Fourier Spectrometer near Gale crater around Ls=336° in MY 31. This observation is consistent with a near-coincident enhanced methane ‘spike’ observed by the Mars Science Laboratory TLS instrument. We have chosen to use the Stochastic Time-Inverted Lagrangian Transport (STILT) particle dispersion model in conjunction with the Mars Weather Research and Forecasting (MarsWRF) GCM for our back-trajectory modeling. To date, we have combined MarsWRF output with a more basic trajectory model, which advects particles based on bulk winds, and have found areas of enhanced tracer density to the north of Gale crater at prior times. Incorporation of turbulent processes in the planetary boundary layer will subject these preliminary results into test. And geological context will also be used to constrain the likelihood of these methane source locations.