Two decades of coastal oceanic nitrous oxide emissions inferred from atmospheric measurements and high-resolution inverse modelling

Nitrous oxide (N₂O) is a major greenhouse gas and ozone-depleting substance. Oceanic sources have been estimated to comprise one-third of natural emissions globally and its uncertainties are larger than for any other source. Previous estimates of ocean N₂O fluxes have been made through oceanographic measurements and using biogeochemical models. However, emissions from coastal areas, which are thought to comprise a significant portion of the total oceanic flux, are traditionally hard to capture through these methods due to their episodic nature and limited spatial scale.

Here, we use a novel method combining high-frequency atmospheric mole fraction measurements with a high-resolution atmospheric transport model and inverse modelling scheme to infer coastal and near-shore ocean N₂O emissions. Atmospheric measurements offer long-term and continuous monitoring and integrate fluxes over larger areas than traditional oceanographic measurements. If anthropogenic influences can be removed, this type of dataset has the potential for long-term quantification of coastal N₂O emissions. In this study, we estimate emissions from six oceanic regions, three of which are part of the four major eastern boundary upwelling systems (EBUS). We show that the three EBUS regions in California, Northwest Africa and South Africa exhibit very large ocean N₂O emissions based on long-term atmospheric measurements from Trinidad Head, California, Cape Verde and Namibia. These emissions are much larger than previously estimated, highlighting the value of this atmospheric measurement-based method. Where possible, we show that enhancements in N₂O mole fraction are associated with depletion events in the atmospheric oxygen/nitrogen (O₂/N₂)ratio, supporting that these enhancements are due to coastal upwelling. In contrast to the EBUS sites, we find that large N₂O emissions do not regularly occur in the ocean surrounding the station of Cape Grim, Tasmania. Using data from Mace Head, Ireland and Cape Ochi-Ishi, Japan, we also show that the North Atlantic and the western North Pacific are heavily influenced by anthropogenic sources, and only a small subset of the data can be used to constrain ocean emissions. Our derived N₂O emissions for these regions are generally smaller than those predicted by the current generation of ocean models.