Hydrogen soil flux measurements using the aerodynamic gradient technique at a grassland in central Scotland, August 2024-September 2025

Several stages of data filtering were applied to improve the quality of retained AG flux data. Initial filtering was based on wind conditions for which data was discarded when u_* < 0.1 m s-1, L < ± 2 m, the variance in vertical wind speed > 1 m s-1 or wind was from the direction obscured by the cabin (130 - 150 ˚N) or fence line between the fields (305 - 315 ˚N). As EC data was used to quality check AGM fluxes, all data was discarded when CO2 flux data was flagged using the 0-1-2 quality flagging system. Where quality flags for sensible heat (H), momentum ( 𝜏) or CO2 flux were >1, data was discarded from the analysis. Plausible limits on CO2 flux at the site were deemed to be within -50 to 25 µmol m-2 s-1. Any EC flux data beyond these limits was also used to discard data. A further filtering of the H2 data was applied based on the comparison of CO2 flux measured between the AG and EC methods. Periods when measured CO2 fluxes by both AG and EC methods differed by more than ± 10 µmol m-2 s-1 were taken as indicative that conditions for AGM calculations were not ideal, thus flux data was discarded from the analysis. After data filtering, approximately 38 % of measured H2 flux data remained.