Abstract. In-situ, quasi-continuous measurements of atmospheric hydrogen (H2) have been performed since 2007 at the Cabauw tall tower station in the Netherlands. Mole fractions of H2, CO and several greenhouse gases are determined simultaneously in air sampled successively at four heights, between 20 and 200 m above ground level. 222Rn measurements are performed in air sampled at 20 and 200 m. This H2 dataset represents the first in-situ, quasi-continuous measurement series of vertical profiles of H2 in the lower continental boundary layer. From the three-year long time series, we characterize the main features and variability patterns of H2 and CO on various time scales; the time series is too short to justify an attempt to determine multi-annual trends. Seasonal cycles are present in both H2 and CO, and their amplitude varies with the sampling height. The seasonality is evident in both the "baseline" values and in the short term (diurnal to synoptic time scales) variability, the latter being significantly larger during winter. The observed H2 short term signals and vertical gradients are in many cases well correlated to other species, especially to CO. On the other hand, H2 has at times a behaviour which differentiates it from all the other species measured, due to its particular distribution of sources and sinks, that is, with the main source in our area (anthropogenic emissions) and the main sink (soil uptake) both near ground level. The local to regional soil sink of H2 is observable as H2 depletion at the lower sampling levels in some of the stable nights, although the signals at Cabauw are smaller than observed at other stations. Positive vertical gradients are another consequence of the soil uptake. Our estimation for the regional H2 soil uptake flux, using the radon tracer method, is (−1.89 ± 0.26) × 10−5 g/(m2h), significantly smaller than other recent results from Europe. Local soil and weather characteristics might be responsible for the very low soil uptake of H2. Our result could also be biased by the absence of radon flux estimates that could reliably approximate the fluxes during the relevant time intervals in our study domain. H2/CO ratios of the traffic emissions computed from our data, with an average of 0.54 ± 0.07 mol:mol, are larger and more scattered than estimated in some of the previous studies in Europe. This difference can be explained by a different driving regime, due to the frequent traffic jams in the influence area of Cabauw. In contrast, the H2/CO ratios of the large scale pollution events, with an average of 0.36 ± 0.05 mol:mol, are very similar to results of previous studies; these ratios were observed to slightly increase with sampling height, possibly due to a stronger influence of soil uptake at the lower sampling heights.
Numerical comparison by different methods (second order Runge Kutta methods, Heun method, fixed point method and Ralston method) to differential equations with initial condition