Using GIS and NOAA HYPLIT trajectories to create meaningful airsheds in Pennsylvania

Atmospheric airsheds, surface areas that contribute contaminants and/or pollutants to a defined location or the airmass above this location, are often defined statically and are byproducts of forward trajectories from emitters rather than back trajectories from monitoring locations. In contrast, we have created a model that specifically uses frequency occurrence of back trajectories in space to delineate airsheds and conduct regional studies of air pollution variability within the resulting airsheds. The model contains an easily understood interface between the NOAA HYPSLIT model and an industry standard GIS platform. The model was used to generate Pennsylvania’s air sheds based on regularly spaced sampling points, and an extraction script to generate meaningfully large climatological sample sets from HYSPLIT online archives. Trajectory generating coordinates were produced at varying resolutions in a regularized grid across Pennsylvania. From each coordinate, back trajectories were calculated based on noon transport for a twelve-day duration over a ten-year period. From these data, air sheds were delineated based on frequency surfaces of trajectories at an annual, seasonal, and monthly level. Meaningful boundaries were discernible at the seasonal and annual level and were consistent with the overall ten-year boundaries. The preliminary study of Pennsylvania suggests that a 10 and 20 km grid of source points produces consistent air sheds.