Site-Calibrated LandscapeDNDC Modeling of Yield-Scaled N₂O Emissions from Smallholder Cropping Systems in Sub-Saharan Africa

Sub-Saharan Africa (SSA) faces a critical dilemma: how to close yield gaps and ensure food security while minimizing agriculture's climate footprint. While nitrogen (N) fertilisation is essential for boosting crop productivity, it can also lead to increased nitrous oxide (N₂O) emissions, thereby further fueling climate change. SSA is highly vulnerable to changes in climate. Yield-scaled N₂O emissions offer a framework to evaluate agricultural climate efficiency, but regional estimates require robust modeling approaches that are calibrated to local conditions. Here, we calibrated LandscapeDNDC, a process-based biogeochemical model, using the most comprehensive dataset on N2O emissions and yields as obtained from 25 field experiments conducted across 12 locations in SSA. These experiments focused on the dominant food crops of maize, sorghum, millet, and rice, and legumes (soybeans and beans). Site-specific parameterization was achieved through Latin hypercube sampling via SPOTPY-LDNDC, followed by validation against an independent dataset of 256 treatment-years across 44 sites representing SSA's major agroecological zones. We assessed the model’s performance in terms of absolute N₂O emissions, yields, and yield-scaled emissions (YSE). We then applied sensitivity analysis to identify the primary drivers of emission variability. Our results show that LandscapeDNDC effectively captures the variability in N₂O and YSE across various cropping systems, highlighting its potential as a tool for national and regional GHG inventories. This could be an efficient way to improve greenhouse gas inventories, enabling better-targeted mitigation and sustainable intensification strategies.