Ground-Based Measurements of Ammonia Emissions from 200 Dairy Farms in California

Emissions of ammonia (NH₃)—as well as methane (CH₄) and nitrous oxide (N₂O)—were measured from 200 unique dairy farms in the San Joaquin Valley, a major agricultural hotspot in California. These 200 farms house 2.4% of the total dairy cow population of the United States. Emissions were quantified using an electric vehicle equipped with a combination of custom and commercial open-path optical gas analyzers. We drove the vehicle on public roads downwind of dairy farms. Emissions were inferred from gas concentration measurements using a Gaussian plume atmospheric dispersion model and wind data from a 2D sonic anemometer mounted on top of the vehicle.

Previous studies have cited high seasonal, diurnal, and farm-to-farm variability of NH₃ emissions from livestock. To address this, our recent field campaign not only included 200 farms to better capture farm-to-farm variability, but also spanned the course of one year. 53 of the 200 farms were sampled at least once a season for all 4 seasons. In addition, our measurements spanned multiple times of the day, and intensive, continuous 24-hour measurements were conducted at 6 farms to examine diurnal variability. In total, there were 477 independent sampling periods of farms, consisting of 273 hours of downwind plume data and 4,364 transects of plumes after data quality filtering.

Measured NH₃ emissions broadly agreed with the statewide NH₃ inventory of 300,920 tons per year. Measured emissions from individual sites varied widely and spanned 3 orders of magnitude but correlated well with facility-scale emission inventories derived from IPCC guidelines and activity data from the Vista-CA database (R² = 0.47). The median ratio of measured emissions to inventory emissions was 1.2. Therefore, for annual statewide emissions, our measurements show that the EPA NEI 2017 reasonably estimates actual emissions.

Non-linear trends in the ratio of NH₃ to CH₄ with air temperature were observed, similar to other studies. However, we observed more spread in the data compared to some studies. Trends were only distinct in the summer and not very clear in the other seasons. This implies that our measurements of NH₃:CH₄ captured many emissions events that were influenced by factors other than air temperature. We are still examining these factors for our data. Other known factors that influence NH₃ emissions are wind speed and farm management practices.