An automated system for selective and continuous measurements of vertical thoron profiles for the determination of transport times near the ground
Abstract. The quantification of in-canopy transport times is of major importance for the investigation of sources, sinks and net fluxes of reactive trace gases within plant canopies. The Damköhler number, which compares timescales of chemical reactions with transport times, is a widely applied measure to evaluate flux divergences. In this study we present and evaluate a novel automated measurement system for selective vertical thoron (Tn) profiles near the Earth's surface and demonstrate its suitability for the direct and reliable determination of transport times within a natural grassland canopy. For the first time, we perform a rigorous determination of systematic and random uncertainties of Tn (and Rn) concentrations under field conditions for this type of measurement system. The obtained median precisions for three concentration classes (> 100 Bq m−3, 100–15 Bq m−3, < 15 Bq m−3) were 8.8%, 23.2% and 132.1% for Tn (and 16.6%, 25.0%, 99.2% for Rn). We calculate in-canopy transport times (τ) and propagate their uncertainty from the individual errors of the Tn concentration measurements. A quality assessment of τ for the field experiment during a period of 51 days revealed good data quality with 44% of the relative uncertainties below 50%. The occurrence of transport time uncertainties higher than 100% was caused by absolute Tn gradients lower than 70 Bq m−3 m−1, which was found for 22% of all determined transport times. In addition, the method was found to be highly sensitive to the Tn concentrations at the upper of the two inlet heights (zu). Low values of CTnzu result in high absolute uncertainties of the transport time. A comparison with empirical parameterizations revealed a much lower scatter for the τ values determined from our measurements. We found an excellent agreement with τ values obtained by the in-canopy resistance approach used, e.g., in the SURFATM model during daytime, while the SURFATM model significantly overestimated transport times during nighttime.
