Field measurement of ammonia emissions after nitrogen fertilization—A comparison between micrometeorological and chamber methods

Abstract

Abstract Field application of ammonium based nitrogen (N) fertilizers can cause high ammonia (NH3) losses into the atmosphere, posing a nutrient loss and threats to the environment. Reliable NH3 flux determination is necessary for improving management practices and emission inventories. However, all deployed measurement techniques are afflicted with different inaccuracies and shortcomings. Thus, in this study different methodical approaches for field measurements of ammonia emissions were compared under varying field conditions and with different N fertilizer types. NH3 fluxes were measured in three field trials applying urea, pig slurry and anaerobic digestate. The NH3 flux was measured by 4 approaches using (a) the combination of passive flux sampler with backward Lagrangian stochastic dispersion flux model (PFSbLS), (b) passive flux sampler with flux calculation by ZINST approach (PFSZINST), (c) the open path Fourier Transform Infrared spectroscopy with backward Lagrangian stochastic dispersion flux model (FTIRbLS) and (d) a calibrated dynamic chamber method (Drager tube measurement, DTM), respectively. The emission results by PFS (with both, bLS and ZINST) and FTIR were in good agreement with mean ratio of all three trials of 1.07 (range of 0.54–2.24) and 1.12 (range of 0.87–1.59) for FTIRbLS/PFSbLS and PFSZINST/PFSbLS, respectively. In urea trial, cumulative NH3 loss determined with DTM was close to that derived from PFSbLS, with a ratio of 1.18. However, in the trials of pig slurry and anaerobic digestate, in which the organic fertilizers were incorporated into the soil, the DTM yielded much lower results than PFSbLS, with the DTM/PFSbLS of 0.56 and 0.30, respectively. This underestimation was probably due to the high heterogeneity of the fluxes at different locations of the experimental site after slurry incorporation. FTIR as well as PFS integrate over large areas, making them robust against spatial heterogeneity. The DTM, in contrast, should only be deployed with surface applied and evenly distributed fertilizers. Results from bLS and ZINST calculations based on PFS raw data were in good agreement, but application of ZINST is clearly restricted by its strict prerequisites, such as deriving empirical ratio of the horizontal flux at the ZINST height to the emission rate from the plot with specific size and surface roughness as well as requiring large uniform fields. FTIRbLs provided the highest temporal resolution and highest sensitivity at low concentrations of all methods, and is thus the best choice when these are required.