Measuring ammonia losses from winter wheat with eddy covariance: a comparative analysis with the integrated horizontal flux method

Primary sources of ammonia (NH3) emissions originate from agriculture, impacting the environment, climate, and human health, thereby concomitantly reducing fertilizer nitrogen use efficiency. Accurate measurements under field conditions are required to provide a basis process understanding and for recommendations to policymakers and farmers. However, uncertainties remain regarding the accuracy and reliability of different low-cost NH3 measurement methods and new application of the eddy covariance method for emissions from low-intensity sources, such as synthetic fertilizers. In this study we focused on the quantification of NH3 concentrations and fluxes determined by a quantum cascade laser spectrometer (QCL) with passivated inlet line within an eddy covariance setup and the comparison to low-cost passive diffusion samplers (ALPHA sampler) used for emission estimations with the integrated horizontal flux (IHF) method. Measurements were carried out in Central Germany during the vegetation periods in 2021 - 2023 in a winter wheat crop field, which received three urea fertilizer applications (to a total of 170 kg N ha-1) per year. The atmospheric NH3 concentrations measured by the QCL and the ALPHA samplers were compared. Then, the cumulative losses of NH3 over the measurement periods (March - July) in the different years (2021 - 2023) were calculated and compared between the two approaches (QCL-eddy covariance and ALPHA-IHF). The NH3 concentration measurements showed that the ALPHA samplers generally yielded lower concentration values compared to the time-integrated QCL values. While the relative mismatch decreased with higher concentrations (>20 ppb), significant deviations were observed in the lower concentration regime. When ALPHA concentrations were corrected for measurement height to precisely align with QCL sampling height, a systematic underestimation was found. Reasons for the differences are currently under investigation and may be explained by the vertical and horizontal sampler separation from the main eddy flux tower and possibly due to varying environmental conditions.   First results of NH3 eddy covariance fluxes (kg N ha-1 period-1) showed clear diurnal courses and emission peaks around noon on the days after each urea application throughout all years. High-frequency losses using a co-spectral method in the process of eddy flux calculation were estimated to be in the range of 25 to 30 %. The performance of both methods (ALPHA-IHF and QCL-eddy covariance) in estimating NH3 losses from field-scale fertilizer applications is discussed, along with the sensitivity of input concentrations on NH3 emission estimates. Our study is a step towards better comparability and integration of different NH3 measurement techniques and is expected to provide useful tools for robust estimation of NH3 emission factors for synthetic fertilizer applications.