1. D. Famulari, E. Nemitz, A. Ibrom, A. Vermeulen, A. Hensen, P. Van Den Bulk, P. Laville, B. Loubet, O. Masher, A. Grossel, E. Rabot, M. Laborde, A. Lohila, T. Laurila, I. Mammarella, S. Haapanala, S. Neri, L. Ferrante, N. Cowan 1, M. Anderson 1, C. Helfter 1 Inter-comparison campaign for the measurement of N 2 O fluxes from a grassland in Scotland using micrometeorology Urbino, 19 September 2013, ACCENT -Plus Symposium

2. Objectives of the InGOS field campaign Improving the estimate of emission inventories for N 2 O using micrometeorological methods, alternative to more commonly used enclosure methods. Standardise and improve QA/QC on non-CO 2 GHG flux measurements Give EU groups the chance to inter-compare their equipment and methods such as REA, aerodynamic gradient and eddy covariance Introduce a protocol for the measurements of N 2 O Eddy Covariance (EC) fluxes. Assess what field setup is optimum for eddy covariance What procedures including corrections are advised in the post-processing of data

3. Objectives of the InGOS field campaign Improving the estimate of emission inventories for N 2 O using micrometeorological methods, alternative to more commonly used enclosure methods. Standardise and improve QA/QC on non-CO 2 GHG flux measurements Give EU groups the chance to inter-compare their equipment and methods such as REA, aerodynamic gradient and eddy covariance Introduce a protocol for the measurements of N 2 O Eddy Covariance (EC) fluxes. Assess what field setup is optimum for eddy covariance What procedures including corrections are advised in the post-processing of data

4. Instrumentation and project partners CountryInstituteInstrument Species measured Method Switzerland, The Netherlands, Germany, Australia Ecotech/ Bremen University Spectronus - FTIR trace gas analyser all Relaxed Eddy Accumulation FranceINRA Orleans lab-built CW-QCL spectrometer "SPIRIT" N 2 O, CO 2, H 2 O Aerodynamic gradient ItalyWest Systems Srl Trasportable accumulation chamber based fluxmeter. N 2 O, CO 2, CH 4 Dynamic enclosure FinlandFMICRD Los Gatos ResearchN 2 O, CO, H 2 OEddy Covariance Finland University of Helsinki CRD Los Gatos Research + dryer N 2 O, CO, H 2 O Eddy Covariance Denmark Technical University of Denmark CRD Los Gatos ResearchN 2 O, CO, H 2 OEddy Covariance NetherlandsECNAerodyne Pulsed QCLN 2 O, CH 4, H 2 OEddy Covariance FranceINRA GrignonAerodyne CW-QCLN 2 O, CH 4, H 2 OEddy Covariance United KingdomNERC - CEHAerodyne CW-QCLN 2 O, CO 2, H 2 OEddy Covariance

5. Field cabin setup 3 x Los Gatos Research Lasers on same absorption lines for N 2 O, CO and H 2 O Same model, but different firmware version/calibration, sample drying CW-QCL from Aerodyne Research x 2 systems: one optimised on absorption lines for CH 4, N 2 O, H 2 O, the other for H 2 O, CO 2, N 2 O Pulsed-QCL from Aerodyne Research optimised on absorption lines for CH 4, N 2 O, H 2 O. All systems logging on a communal PC running a custom-made LabView acquisition program, able to store data synchronously from the sonic anemometers and all EC-systems at 10Hz.

6. Easter Bush, Edinburgh Scotland 2013 N field S field N Intensively managed grassland with a permanent meteorological field station monitoring long term EC fluxes for CO 2, several years of N 2 O by enclosure techniques. Measurements started 3 rd June 2013 and finished on 30 th June 2013: first week of background measurements with sheep grazing on both fields Both fields fertilised on 11 th June, with NH 4 NO 3 (34.5% N) at a rate of 150 kg/ha Subsequent N 2 O emissions measured for the following 3 weeks.

7. Easter Bush meteorology June 2013 Easter Bush has a typical oceanic climate. In June 2013, the weather was unusually dry and sunny, with little rainfall, as a consequence of high pressure conditions.

8. Turbulent fluxes June 2013 Turbulent fluxes of momentum and heat were in the typical range for the Easter Bush field site, representing a fairly stable summer month.

9. Field setup: inlets Ultra sonic anemometers: 2 x Gill HS-50 disposed at a 90 o angle, one along the fence line and one pointing into one of the fields (prevailing wind dir) Inlet tubing was 3/8” OD, different materials were used: PE, Dekabon. All lines were heated and insulated up to the analysers located indoors. One LGR system moved between 2 sonics. Flow rates: Main common inlet = 43 l/min (4 combined) DTU inlet = 30 l/min (1 system) ECN inlet = 13.7 l/min (2 system, 1 REA)

10. Time Response: field study Objective: Estimate the time lag and high frequency attenuation for the N 2 O sensors with a simple field experiment. Then compare the time lags obtained with the post-processed data by maximization of covariances method. N 2 O sensor fcfc  tube Hz s LGR_DTU0.71.2 LGR_FMI0.461.2 LGR_UHEL0.670.9 cwQCL_CEH1.140.825 cwQCL_INRA1.061.2

11. Cross-Calibration of all systems We used ordinary compressed air tanks as well as BOC- rated standard gas mixtures containing near-ambient levels of CH 4, CO 2, N 2 O. FTIR used as absolute value measurement, and reference to all others. The compressed gas was overflowed at the inlet end near the sonic anemometer, to let all inlets sub-sample from it. Measured values [ppb]Calibration factor rel. to FTIR Cylinder 1Cylinder 2Cylinder 3Cylinder 1Cylinder 2Cylinder 3 18.2 ppm CO LGR-FMI387334.3381.50.9290.9440.927 LGR-UHEL365-63600.985-52.590.982 LGR-DTU408220.5401.50.8811.4310.881 cwQCL-CEH369.1324363.20.974 cwQCL-INRA378371.20.9510.953 pQCL-ECN3993920.9010.902 cwQCL- INRA(Spirit)3710.969 FTIR315.52

12. N 2 O precision and stability Instrument0.1 s best15 min RMSE ECN QCL767213636137 CEH cw-QCL200193820 INRA cw-QCL2071310223 DTU LGR5403225656 FMI LGR10462719491 UHEL LGR6214017281 As an outcome of the calibration exercise, an Allan variance study determines the precision at different rates of acquisition (here shown in pptV) for all systems

13. N 2 O fluxes time series

14. N 2 O cumulative fluxes Averaged N 2 O emissions from 6 to 27 June 2013: only overlapping measuring periods were considered.

15. Comparison of different sensors fluxes [nmol / m 2 s] Each sensor (Y axes) against the “average sensor” (X axes)

16. Summary 6 N 2 O eddy covariance systems have been compared during a field campaign over 25 days The N 2 O fast monitors were inter-calibrated via gaseous standards, and statistical analysis showed precisions ranging from 0.2 and 1 ppbV at 10Hz from 1 instance of calibration. The averaged N 2 O emission over the period was evaluated at 26.35 g N 2 O-N ha -1 day -1 (EF ~1.4) With the latest generation lasers, it is possible to measure very small fluxes, potential for non-agricultural fluxes and uptake studies. Potential benefit from physically drying the sample when using CRD LGR systems All LRG systems work slightly differently (i.e. output formats, CO correction?) despite being the same model.

17. On-going and future work Assessment of the impact of the corrections to apply to the datasets: water effects (spectroscopic and WPL), spectral corrections Comparison of magnitude of error correction components Compilation of a wish list for field deployment and manufacturers Dataset available to the community for possible gap-filling work Comparison of three micrometeorological techniques (REA, EC, AGM), and a dynamic enclosure method. Subsets of instruments will be assessed for performance on CO 2, CH 4, CO, both for concentration and fluxes.

18. Thank you!