Seasonally Resolved Excess Urban Methane Emissions from the Baltimore/Washington, DC Metropolitan Region.

Urban areas are increasingly recognized as an important source of methane (CH), but we have limited seasonally resolved observations of these regions. In this study, we quantify seasonal and annual urban CH emissions over the Baltimore, Maryland, and Washington, DC metropolitan regions. We use CH atmospheric observations from four tall tower stations and a Lagrangian particle dispersion model to simulate CH concentrations at these stations. We directly compare these simulations with observations and use a geostatistical inversion method to determine optimal emissions to match our observations. We use observations spanning four seasons and employ an ensemble approach considering multiple meteorological representations, emission inventories, and upwind CH values. Forward simulations in winter, spring, and fall underestimate observed atmospheric CH while in summer, simulations overestimate observations because of excess modeled wetland emissions. With ensemble geostatistical inversions, the optimized annual emissions in DC/Baltimore are 39 ± 9 Gg/month (1 ), 2.0 ± 0.4 times higher than the ensemble mean of bottom-up emission inventories. We find a modest seasonal variability of urban CH emissions not captured in current inventories, with optimized summer emissions ∼41% lower than winter, broadly consistent with expectations if emissions are dominated by fugitive natural gas sources that correlate with natural gas usage.