N2O flux measurements over an irrigated maize crop: A comparison of three methods

International audienceThis paper presents the NitroCOSMES campaign, aimed at testing and evaluating the performance of three methods for monitoring N2O fluxes over an agricultural field. The experiment was conducted from May to August 2012 at a site located in the south-west of France. N2O fluxes from a 24 ha irrigated maize field were measured using eddy covariance (EC), automated chamber (AC) and static chamber (SC) methodologies. Uncertainties were calculated according to the specificities of each set-up. Measurements were performed over a large range of water-filled pore spaces (WFPS), soil temperatures, and mineral nitrogen availability, and offered the opportunity to compare methodologies over a wide range of N2O emission intensities. The average N2O fluxes were compared among the three methodologies during the same periods of measurement and for different intensities of emissions (low, moderate and high). Periods of comparison were determined according to the AC results. On average, the three methods gave comparable results for the low (SC: 14.7 ± 2.2, EC: 15.7 ± 10.1, AC: 17.5 ± 1.6 ng N2O-N m−² s−1) and the high (SC: 131.7 ± 22.1, EC: 125.3 ± 8, AC: 125.1 ± 8.9 ng N2O-N m−² s−1) N2O emission ranges. For the moderate N2O emission range, AC measurements gave higher emissions (57.2 ± 3.9 ng N2O-N m−² s−1) on average than both the SC (41.6 ± 6.6 ng N2O-N m−² s−1) and EC (33.8 ± 3.9 ng N2O-N m−² s−1) methods, which agreed better with each other. The relative standard deviation coefficient (RSD) indicated that EC methodology gave highly variable values during periods of low N2O emissions, from -52.2 ± 88.1 to 62.2 ± 50.7 ng N2O-N m−² s−1, with a mean RSD of 151%. Water vapour effects (dilution and spectroscopic cross-sensitivity) were discussed in an attempt to explain the high variability in low N2O emission measurements. Even after applying the Webb term correction, there could still be a spectroscopic cross-sensitivity effect of water vapour on the N2O trace gas signal because of the layout of the analysers, which was not determined during the experiment. This study underlined that EC methodology is a promising way to estimate and refine N2O budgets at the field scale and to analyse the effects of different agricultural practices more finely with continuous flux monitoring. It also highlighted the need to continue the effort to assess and develop chambers and EC methodologies, especially for the low N2O emission measurement range, for which values and systematic uncertainties remain high and highly variable