Evaluation of back trajectories associated with ozone transport during the 1993 North Atlantic regional experiment

Back trajectories traditionally have been used to identify possible sources of pollutants measured at a receptor location; however, they neglect turbulent diffusion processes that affect pollutant transport. After evaluating the performance of a coupled mesoscale and Lagrangian particle dispersion model with observations taken during the 1993 Northern Atlantic Regional Experiment (NARE), a series of forward and back trajectories are analyzed to determine the errors associated with the back trajectory calculations. The forward trajectories are based on the mean and turbulent wind components and their initial positions correspond to several urban areas in eastern North America. The back trajectories employ only the mean wind components and their initial positions correspond to the final positions of the forward trajectories within a sampling volume near Yarmouth, Nova Scotia. Large differences between the forward and back trajectory positions occurred since the back trajectories did not include the irreversible turbulent diffusion processes. The back trajectories had no mechanism to identify surface source regions, with the consequence that the back trajectories would either pass over the “true” source region, or miss it altogether as a result of locating the particle at an incorrect horizontal position. A horizontal displacement error of ± 100–400 km and a vertical displacement error of ± 0.2–1 km occurred after 2 days of transport. These errors leveled off between 2 and 3 days, but grew larger again after 3 days of transport. The model results suggest that the use of back trajectories by themselves may be an inappropriate approach for estimating the source location of a plume emitted near the surface.