A case study of eddy covariance flux of N20 measured within forestecosystems: quality control and flux error analysis

摘要

Eddy covariance (EC) flux measurements of ni-trous oxide (N20) obtained by using a 3-D sonic anemometerand a tunable diode laser gas analyzer for N20 were inves-tigated. Two datasets (Soro, Denmark and Kalevansuo, Fin-land) from different measurement campaigns including sub-canopy flux measurements of energy and carbon dioxide arediscussed with a focus on selected quality control aspects andflux error analysis. Although fast response trace gas analyz-ers based on spectroscopic techniques are increasingly usedin ecosystem research, their suitability for reliable estimatesof EC fluxes is still limited, and some assumptions have tobe made for filtering and processing data. The N20 con-centration signal was frequently dominated by offset drifts(fringe effect), which can give an artificial extra contribu-tion to the fluxes when the resulting concentration fluctua-tions are correlated with the fluctuations of the vertical windvelocity. Based on Allan variance analysis of the N20 sig-nal, we found that a recursive running mean filter with atime constant equal to 50 s was suitable to damp the influ-ence of the periodic drift. Although the net N20 fluxes overthe whole campaign periods were quite small at both sites(— 51.tg N M-2 h-1 for Kalevansuo and — 10 lug N I11-2 h-1for Sort ), the calculated sub-canopy EC fluxes were in goodagreement with those estimated by automatic soil chambers.However, EC N20 flux measurements show larger randomuncertainty than the sensible heat fluxes, and classificationaccording to statistical significance of single flux values in-dicates that downward N20 fluxes have larger random error.