Diurnal Variation and Sampling Frequency Effects on Nitrous Oxide Emissions Following Nitrogen Fertilization and Spring‐Thaw Events. Issue 3 (1st May 2019)

• Record Type: Journal Article
• Authors: Ferrari Machado, Pedro Vitor; Wagner-Riddle, Claudia; MacTavish, Robert; Voroney, Paul R.; Bruulsema, Thomas W.
• Is Part Of: Soil Science Society of America Journal. Volume 83:Issue 3(2019)
• Page Start: 743
• Page End: 750
• Publication Date: 2019-05-01
• Subjects: Soils -- United States -- Periodicals; Soil science -- Periodicals; Periodicals; 631.4973
• DOI: 10.2136/sssaj2018.10.0365
• Languages: English
• ISSNs: 0361-5995

Abstract

Abstract : Core Ideas: High‐frequency N2 O emission data was subsampled to assess sampling frequency errors. Mid‐morning sampling was adequate to derive daily average N2 O emission during events. Twice‐weekly sampling gave an uncertainty of –6 to +12% during spring thaw. Twice weekly + sample after >10 mm rainfall had –1 to +19% error after fertilization. An infrequent sampling protocol can introduce bias into N2 O studies. Different methods are used to measure nitrous oxide (N2 O), a potent greenhouse gas emitted from agricultural soils. While some methods (e.g. micrometeorological methods) conduct near‐continuous measurements, manual chambers measure discontinuously. Estimates of N2 O emissions based on discontinuous measurements can carry errors due to: (i) diurnal variation and (ii) integration of emissions over time. This study evaluated these two sources of uncertainties and identified the optimal sampling strategy for emission following spring‐thaw events (ST) and nitrogen fertilization (NF). Two times of day (mid‐morning [MM] and mid‐afternoon [MA]) and three sampling frequencies suggested in the literature on N2 O emissions (bi‐weekly [BW], weekly [W], twice weekly [TW] for the ST, and W, TW, and TW plus extra sample after rainfall events >10 mm for the NF) were created by subsampling a high‐frequency reference dataset. We show that the mid‐morning sampling strategy effectively represented the daily N2 O emission average, while the mid‐afternoon strategy overestimated fluxes by 14 and 32% for the ST and NF, respectively. For the ST, the weekly mid‐morning sampling protocol resulted in errors ranging from –3 to +33%, with lower uncertainties when sampling frequency increased to TW (–6 to +12%). The TW mid‐morning sampling also was adequate for the NF datasets (–5 to +19%), with narrower uncertainty levels when an additional sample was taken after >10 mm rainfall (–1 to +19%). These results provide increased confidence in selecting the sampling strategy for discontinuous measurements following N fertilization and STs in cropping systems.