Weekly agricultural emissions and ambient concentrations of ammonia: Validation of an emission inventory

Weekly inventories for emissions of agricultural ammonia were calculated for 139 4 × 4km grid cells over 52 weeks in the intensely farmed Lower Fraser Valley, BC. The grid cells were located both inside and outside an area that had been depopulated of poultry due to an outbreak of Avian Influenza prior to the start of the study. During the study period, ambient ammonia concentrations were measured hourly at two locations outside the cull area and one location inside the cull area. Large emission differences between grid cells and differences in temporal variation between cells were related to farming practices and meteorological factors such as temperature and rainfall. Weekly average ambient concentrations at the three sampling locations were significantly correlated with estimates of weekly emissions for many of the grid cells in the study area. Inside the cull area, ambient concentrations during the cull (week 1) were 37% of the concentrations after the cull (week 52), while outside the cull there was almost no difference between week 1 and week 52, suggesting that in normal (non-cull) conditions, about 60% of the ambient ammonia was due to poultry farms. Estimated emissions in weeks 1 and 52 for grid cells affected by the cull indicated that over 90% of the emissions came from poultry. The discrepancy in difference between week 1 and 52 for emissions and ambient concentrations could be due to atmospheric factors like transport, atmospheric reactions, dispersion or deposition; to errors in the inventory including farming data, emission factors; and omission of some non-poultry emission sources. Overall the study supports the ammonia emission inventory estimates. Detailed emission data helps in modeling ammonia in the atmosphere and is useful for developing abatement policy. •Weekly NH3 emissions in 4 × 4km grid cells during and after a major poultry cull.•Concentrations NH3 were measured at 2 locations outside and 1 inside the cull area.•Wide variation in emissions due to farm activities and weather factors.•Good correlation between emission estimates and ambient measurements.•Detailed inventories useful for policy decisions and improving model inputs.