Lessons from the Field: Insights for Modelling N2o Emissions from Riverine Ecosystems.

Nitrous oxide (N₂O) emissions are globally significant to climate and ozone depletion.  Aquatic emissions represent a win-win opportunity for mitigation, with efforts to improve water quality expected to reduce N₂O emissions – but a key step in mitigation is a better a priori understanding of what those emissions actually are. Approaches to modelling N₂O emissions in aquatic ecosystems are easy to criticize. Complex, process-based models are impacted by overparameterization and data limitations, yet simpler methods may overlook key controls. Along the way, we must consider modelling goals – from estimating national and global emissions to understanding how to mitigate emissions, or identify hotspots. IPCC-based methods for estimating indirect agricultural emissions have been highly criticized – yet no simple, tractable replacement has been identified that allows reliable emissions estimation with limited data. Here, we revisit our past work – assessing temporal and spatial variation in N₂O emissions from rivers and streams, and comparing these to estimates using the empirically-derived EF5-g – supplementing the analyses with new data, and data from varied ecosystems, including systems strongly impacted by anoxia. We assess several proposed approaches to estimating N₂O emissions in watersheds with a variety of hydrological regimes, land covers, and anthropogenic impacts. We use instantaneous field-based flux data to identify and assess hotspots and hot moments. Then, we return to the scale for which the EF5-g was really intended – assessing relationships between annual emissions and nitrate export across a suite of ecosystems for which we could obtain data. We end by identifying key data gaps and opportunities to help achieve robust, but simple, N₂O accounting in freshwater systems.