Estimating lateral nitrogen transfers over the last century through the global river network using a land surface model

International audience

英语. Lateral nitrogen (N) transport from land to oceans through rivers is an important component of the global N cycle. We developed a new model of this aquatic system, called LSM_Nlateral_Off (Land Surface Model Nitrogen lateral Offline), which simulates the routing of water in rivers, and the pertaining transport of dissolved inorganic N (DIN), dissolved organic N (DON), and particulate organic N (PON) as well as the accompanying biogeochemical processes of DON and PON decomposition and denitrification during transit from land to oceans through the global river network. Evaluation against global observation-based datasets shows that the model effectively captures both the magnitude and seasonal variations in riverine water discharges and total nitrogen (TN) flows. Our model was then applied to reconstruct the historical evolution of global N flows and transformations from land to rivers and, ultimately, the oceans. Model simulation results indicate that, driven by anthropogenic activities (e.g. application of mineral fertilisers and manure, sewage water injection in rivers, and land use change) and indirect effects of climate change and rising atmosphere CO 2 , TN exports increased from 27.5 Tg N yr -1 during the 1901-1920 period to 40.0 Tg N yr -1 during the 1995-2014 period, with DIN contributing most (80 %) of this increase. Simulation results reveal substantial spatial heterogeneities in annual mean TN flows and denitrification rates, while their seasonal amplitude is of similar magnitude to the large-scale spatial variability. Compared to previously published regional or global aquatic N transfer models (IMAGE-GNM, FrAMES-N, MBM, DLEM, and GlobalNEWS2), our model produces similar global-and continentalscale TN exports to the ocean but shows distinct patterns at the finer scale of river basins. LSM_Nlateral_Off is here coupled to the land surface model (LSM) ORCHIDEE, but the offline approach implemented in this work facilitates its coupling with other land surface models in the future such as those synthesised by the Global N 2 O Model Intercomparison Project (NMIP). Our modelling approach provides a comprehensive simulation of N transport and transformations from terrestrial ecosystems to oceans at 0.5°spatial resolution and daily temporal resolution, globally