Measuring frequently during peak soil N2O emissions is more important than choosing the time of day to sample
Abstract. Manual gas sampling from static soil chambers is commonly used to measure the flux of nitrous oxide (N2O) from soil. Because manual sampling is labour intensive, sampling frequencies are often insufficient to fully capture daily variability of N2O soil flux, which compromises the accuracy of estimates of daily and cumulative emissions. Knowledge of the diurnal fluctuation of N2O flux has been used to choose a flux sampling time that maximizes the accuracy of N2O flux estimates and thereby reduces the required frequency of flux measurements, but results of previous studies have been inconsistent. We analysed N2O soils emissions measured quasi-continuously over three years from a highly fertilized (> 200 kg N ha−1) maize system grown in southern Wisconsin, USA. This is the first study of N2O flux temporal variability that includes multiple difficult-to-measure peak emission events (
hot moments) and estimates the relative contribution of hot moments to cumulative emissions. Analysis of diurnal fluctuation in N2O flux was performed using all measured data (≈ 22 000 fluxes) as well as using subsets of the data grouped by flux magnitude. The relationship between the observed hourly average flux and the mean daily flux was assessed using linear regression. Results show that diurnal variation in N2O soil flux was closely associated with normalized flux size. During low emission periods, N2O soil fluxes exhibited a diurnal pattern such that N2O flux measured at particular times of day,
Preferred Measuring Times (PMTs), were not significantly different from the mean daily flux. During high emissions periods N2O flux did not exhibit a diurnal pattern and there was no PMT. High emissions periods included difficult-to-measure hot moments that did not exhibit a PMT and contributed up to 50 % of the cumulative emissions. We conclude that in order to accurately measure soil N2O flux in this type of system, it is necessary to sample frequently, particularly during peak flux events, and that constraining sampling to particular times of day provides little benefit.
This preprint has been withdrawn.
N2O Flux data https://doi.org/10.21231/6a7b-xy41
Model code and software
R Code, statistical analysis https://doi.org/10.21231/6a7b-xy41
Viewed (geographical distribution)