On the contribution of river network to nitrous oxide emissions

Nitrous oxide (N2O) is a potent greenhouse gas (GHG) that contributes to stratospheric ozone destruction. Anthropogenic nitrogen (N) loading to inland river networks is a potentially important source of this gas to the atmosphere via microbially-mediated denitrification of reactive nitrate (NO3-) to N2O and dinitrogen (N2) gas. While recent studies have shown that inland waters, oceans, and the terrestrial landscape contribute similarly to global carbon dioxide (CO2) emissions, N2O emissions from inland river networks are poorly quantified. Moreover, field evidence underlined the key role played by the hydro-biogeochemical interaction between surface (i.e., the water column) and subsurface (i.e., benthic and hyporheic zone) riverine environments in controlling these emissions. To consider all these factors, we propose the generalization of a predictive model, based on Damköhler numbers (i.e., on the ratio between time of transport and time of reaction), to capture the primary effects of reach-scale biogeochemical and hydro-morphological characteristics in regulating the release of N2O to the atmosphere. The developed model was tested against available field data under mean annual streamflow and N load conditions as well as under extreme events such as drought conditions. The scalable characteristic of the proposed modeling approach allows to estimate, continuously, N2O emissions from streams and rivers from easily retrievable stream flow and channel morphology datasets. The model input parameters can be measured in the field or derived, mainly for large scale applications, from available data also through artificial intelligence techniques such as Machine Learning. Model outputs underline the importance to holistically look at the riverine environments to predict their role in controlling DIN (Dissolved Inorganic Nitrogen) fate and the production/emission of N2O. Results are socially relevant to policy makers combating high nitrogen concentrations in surface waters. Furthermore, the proposed approach is useful in supporting river restoration projects, evaluating land use effects and beneficial positive effects of non-point source management.