Validation and uncertainty quantification of three state-of-the-art ammonia surface exchange schemes using NH₃ flux measurements in a dune ecosystem

Abstract. Deposition of reactive nitrogen causes detrimental environmental effects, including biodiversity loss, eutrophication, and soil acidification. Measuring and modeling the biosphere-atmosphere exchange of ammonia, the most abundant reduced nitrogen species, is complex due to its high reactivity and solubility, often leading to systematic discrepancies between model predictions and observations. This study aims to determine whether three state-of-the-art exchange schemes for NH₃ can accurately model NH₃ exchange in a dune ecosystem (Solleveld) and detect factors causing the uncertainties in these schemes. The selected schemes are DEPAC by Van Zanten et al. (2010), and the schemes by Massad et al. (2010) and Zhang et al. (2010). Validation against one year of gradient flux measurements revealed that the Zhang scheme represented the NH₃ deposition at Solleveld best, whereas the DEPAC scheme overestimated the total deposition while the Massad scheme underestimated the total deposition. Yet, none of these schemes captured the emission events at Solleveld, pointing to considerable uncertainty in the compensation point parameterization and possibly in the modeling of NH₃ desorption processes from wet surface layers. The sensitivity analysis further reinforced these results, showing how uncertainty in essential model parameters in the external resistance (R_w) and compensation point parameterization propagated into diverging model outcomes. These outcomes underscore the need to improve our mechanistic understanding of surface equilibria represented by compensation points, including the adsorption-desorption mechanism at the external water layer, and specific recommendations are provided for future modeling approaches and measurement setups to support this goal.