Surprisingly low frequency attenuation effects in long tubes when measuring turbulent fluxes at tall towers

The eddy covariance technique relies on the fast and accurate measurement of gas concentration fluctuations. While for some gasses robust and compact sensors are available, measurement of, e.g., non CO2 greenhouse gas fluxes is often performed with sensitive equipment that cannot be run on a tower without massively disturbing the wind field. To measure CO and N2O fluxes, we installed an eddy covariance system at a 125 m mast, where the gas analyser was kept in a laboratory close to the tower and the sampling was performed using a 150 m long tube with a gas intake at 96 m height. We investigated the frequency attenuation and the time lag of the N2O and CO concentration measurements with a concentration step experiment. The results showed surprisingly high cut-off frequencies (close to 2 Hz) and small low-pass filter induced time lags (< 0.3 s), which were similar for CO and N2O. The results indicate that the concentration signal was hardly biased during the ca 10 s travel through the tube. Due to the larger turbulence time scales at large measurement heights the low-pass correction was for the majority of the measurements < 5%. For water vapour the tube attenuation was massive, which had, however, a positive effect by reducing both the water vapour dilution correction and the cross sensitivity effects on the N2O and CO flux measurements. Here we present the set-up of the concentration step change experiment and its results and compare them with recently developed theories for the behaviour of gases in turbulent tube flows.