An inverse dispersion technique for the determination of ammonia emissions from urea-applied farmland

A backward Lagrangian stochastic (BLS) model for estimating ammonia emission in the field, utilizing data from ammonia concentration measurements by the open-path tunable diode laser method, was assessed by comparing it with the benchmark provided by simultaneous monitoring by the micrometeorological mass balance (MMB) technique. The results indicated that after rejecting data with u* ≤ 0.15 m s−1 or |L| < 10 m, the ammonia emissions derived by the BLS model were very similar to those by MMB (regression gradient = 0.966, R2 = 0.959). The BLS model was also compared with the static chamber (SC) method of monitoring changes in gas concentration (enrichment/depletion) over time; it was found that these methods yielded significantly different results for both diurnal ammonia fluctuations and total ammonia loss. The results also showed that the mean time had little effect on the accuracy of the BLS method, but a short averaging time was preferable when investigating diurnal change of ammonia emissions.