Applicability of a closed-path quantum cascade laser spectrometer for eddy covariance (EC) flux measurements of nitric oxide (NO) over a cropland during a low emission period

Croplands are important sources of atmospheric nitric oxide (NO). However, high-frequency measurements of NO fluxes over croplands using the eddy covariance (EC) technique are still scarce, mainly due to instrumental limitation. In this study, a closed-path NO analyzer based on a quantum cascade laser (QCL) absorption spectrometer was employed for EC flux measurements over a subtropical vegetable field during a two-month summer period with the lowest NO emission intensity of the year. The purpose was to investigate the detection limit of the EC system based on this NO analyzer and evaluate its applicability for measuring the turbulent fluxes of NO under field conditions. The performance of the analyzer was stable, showing an average precision (0.1 s) of 0.338 nmol mol(-1) and a corresponding flux detection limit of 5.6 mu g N m(-2) h(-1) at the 95% confidence interval. The measured turbulent NO fluxes ranged from -7.1 to 61.4 mu g N m(-2) h(-1) (median: 3.5 mu g N m(-2) h(-1)), with a relative random error of 386% before field ploughing and 76% thereafter. The systematic errors due to the high-frequency loss and the use of lag times of carbon dioxide for NO flux calculation were estimated at 12% and 3%, respectively. During the measurement period, 37% of the observed half-hourly fluxes were larger than the detection limit; the magnitude of these fluxes is comparable with that measured by the static chambers. Nevertheless, this EC system could be still qualified for measuring turbulent NO fluxes over common croplands if the flux averaged at daily or longer timescales are of interest, because either flux detection limit or random error would decrease by an order of root n, wherein n is the number of half-hourly fluxes being taken for averages. This study shows that the closed-path dual-QCL analyzer could be an effective option for EC measurements of turbulent NO fluxes with the advantages of (i) stable performance, (ii) high precision and fast response, and (iii) feasible instrumental maintenance for long-term field measurements. However, the observed turbulent fluxes still underestimated the soil NO emissions likely due to chemical reaction loss of NO below the sensor height. Further studies are necessary to address this systematic error.