Main content area

Ammonia emissions from a beef feedlot: Comparison of inverse modeling techniques using long-path and point measurements of fenceline NH3

Shonkwiler, Kira B., Ham, Jay M.
Agricultural and forest meteorology 2018 v.258 pp. 29-42
ammonia, ammonium nitrogen, beef, beef cattle, crude protein, dust, ecosystems, emissions, emissions factor, feedlots, models, national parks, nitrogen, Colorado, Texas
Nitrogen (N) deposition in Rocky Mountain National Park, USA (RMNP) and adjacent alpine ecosystems has led to increased focus on ammonia (NH3) emissions from livestock operations in this region. A study was conducted at a commercial beef feedlot in Northern Colorado, USA to quantify NH3 fluxes, estimate emission factors (EFs), and evaluate measurement methodologies. Point and line-integrated fenceline NH3 measurements were taken over a 78-day study at a large feedlot east of RMNP focusing on transport towards the park. Ammonia fluxes were estimated using two inverse models, a backward Lagrangian stochastic approach (WindTrax) and a simpler inverse dispersion approach (FIDES) using input data from both NH3 datasets. Line-integrated NH3 concentrations were obtained from a long-path laser and single-point concentrations were collected with a cavity ring-down analyzer, both positioned on the downwind fenceline. Mean feedlot NH3 concentration (±1 standard deviation, σ) was 601±246μgm−3 (1013±421 ppbv), where peak values often exceeded several ppmv. Concentrations from the long-path sensor were slightly higher than the point sensor when winds were from the feedlot (p=0.0005), but 2.5 times less data were collected with the long-path instrument due to dust or laser-reflector alignment issues. Regardless of the model used, fluxes showed a diel pattern peaking in the afternoon with early morning minimums. Emissions varied more by inverse model than by sensor, with WindTrax emissions 25% higher than FIDES. Composite 24-h mean emissions (±1σ) from FIDES were 48±24μgm−2s−1 (EF=71±35ghd−1d−1), while WindTrax averaged 60±30μgm−2s−1 (EF=89±44ghd−1d−1) (p<<0.0001). Overall 24-h mean EF for the summer (across both models) was 80±39ghd−1d−1. When dietary N was considered (13.25% crude protein), based on 24-h composites FIDES estimated NH3-N released to the atmosphere to be 35.6% of fed-N, while WindTrax showed 44.6% fed-N. The overall mean across models was 40.0% fed-N.