Analysis of uncertainty for N₂O fluxes measured with the closed‐chamber method under field conditions: Calculation method, detection limit, and spatial variability.

Abstract

The search for agricultural practices that mitigate N2O emissions, and the validation and development of models, would benefit if the uncertainty of emission estimates would be reduced relative to current levels. This uncertainty has different sources, such as the error of the flux measurements, the error caused by spatially scaling up measurements to the field or regional level, and the error caused by estimating emissions for the time intervals between measurements by interpolation. This paper focuses on the uncertainty of flux measurements and on flux spatial variability, which is a major cause of error when measured fluxes are scaled up spatially. The analysis focuses on 2415 flux measurements made with the closed chamber method over a monitoring campaign of approximately 3 years. Statistically significant nonlinearities in the changes of N2O concentrations during chamber closure were infrequent (8.3%). Further analysis of significant non‐linear concentration changes indicates that, for positive fluxes, nonlinearity might not always be an artifact caused by chamber placement, but that it can reflect natural temporal variability of the flux during chamber placement in a significant number of cases. The analysis of the coefficients of determination (R2) and of the normalized root mean squared errors (NRMSE) of the linear regressions shows that, below emission rates of 5 g N2O‐N ha−1 d−1, the uncertainty of flux estimates strongly increases. The flux detection limit was 3.5 g N2O‐N ha−1 d−1, which is consistent with the outcome of the analysis of the R2s and NRMSEs. Flux measurements based on less than five N2O concentrations per flux led to estimates with considerably larger confidence intervals. When the number of N2O concentrations per flux was reduced from five to four, the detection limit increased to 7.5 g N2O‐N ha−1 d−1. The individual fluxes of spatially replicated plots show strong dispersion around the mean: the average coefficient of variation for fluxes above 5 g N2O‐N ha−1 d−1 was 54.3% and the data suggest that the spatial variability of fluxes correlates positively with flux magnitude. Our results suggest that, for measurements performed with the closed static chamber method, (1) linear regression might generally lead to the best estimates of the average fluxes during closure time, and that the chamber sampling strategy might be designed accordingly, (2) there is considerable potential to reduce the uncertainty for fluxes lower than 5 g N2O‐N ha−1 d−1 by employing analytical instrumentation with higher precision, (3) flux uncertainty and detection limit are strongly affected by the number of concentrations measured for each flux, and (4) large numbers of replicated chambers could be particularly beneficial if high fluxes are expected.

N2O flux short-term response to temperature and topsoil disturbance in a fertilized crop: An eddy covariance campaign

“…Up to now, most in situ studies on crops were performed using manual closed chambers at low temporal resolution. However, this technique prevents researchers from adequately capturing the sudden variations in N 2 O emissions (Kroon et al, 2008) and can be source of error when measurements are scaled up spatially (Lammirato et al, 2018). Automated chambers that allow an improved temporal resolution have also been used (Laville et al, 2011, but the uncertainty related to spatial variability remains.…”

Global Research Alliance N2O chamber methodology guidelines: Recommendations for deployment and accounting for sources of variability

“…Similar to Venterea et al. (2020), along with raising the minimum detectable flux, Lammirato, Lebender, Tierlingand, and Lammel (2018) found the uncertainty of individual N 2 O flux estimates (calculated by linear regression over five headspace sampling points) increased with increasing chamber volume (perhaps indicating that headspace mixing is required; Clough et al., 2020).…”

Detection of nitrous oxide using infrared optical plasmonics coupled with carbon nanotubes

“…Field measurements typically are conducted using the "closed chamber method", 19 which has come under increasing scrutiny with regards to accuracy. 20 Therefore, there is a need to create a sensing platform to conduct eld/in situ measurements with high sensitivity and a low detection limit, which is also simple to operate and regenerate, portable, robust, and power efficient for remote eld measurements and data collection.…”