Improving the representation of NOx emissions from soils in the LMDZORINCA model for global chemistry and climate purposes

Nitrogen oxides (NOₓ ≡ NO + NO₂) play a central role in atmospheric chemistry, with direct impacts on human health through respiratory stress and indirect effects on climate via tropospheric ozone production, methane lifetime, and secondary aerosol formation. To study the impacts of NOₓ emissions on the atmospheric chemistry and the climate, we developed a version of the LMDZORINCA Chemistry-Transport model in which soil NOₓ emissions are computed by the ORCHIDEE land surface model, in complement of anthropogenic emissions provided by inventory data.

Current estimates place global total NOx emissions at about 45–50 Tg N yr⁻¹, dominated by fossil fuel combustion. Soils are a substantial source of NOₓ, with global emissions estimated at ~10–15 Tg N yr⁻¹, of which ~3 Tg N yr⁻¹ are attributable to fertilizer and manure inputs, the remainder arising from natural processes. Due to air-quality and climate policies reducing NOₓ emissions from transport and industrial sectors, the relative importance of soil NOₓ emissions is expected to further increase in the future. Currently, most atmospheric chemistry models take into account agricultural soil NOₓ emissions, using inventory-based approaches. These inventories rely on fixed emission factors or highly simplified parameterizations to calculate NOₓ emissions from nitrogen inputs, thereby neglecting, or overly simplifying, the strong non-linear dependence of emissions on climate, soil biogeochemistry, and vegetation type. Furthermore, natural soil NOₓ emissions are often neglected or calculated using simplified parametrizations.

In this work, we use soil NOₓ emissions estimated by the ORCHIDEE terrestrial biosphere model as an input to the LMDZINCA atmospheric chemistry model. ORCHIDEE simulates the carbon-nitrogen cycle as well as soil microbial nitrification and denitrification processes, thus offering a mechanistic and ecologically grounded description of soil NOₓ emissions. We first evaluate ORCHIDEE soil NOX emissions against three different datasets : CAMS-GLOB-SOIL product, which is based on empirical parametrizations, the inventory-derived dataset for anthropogenic NOₓ emissions CEDS, and the DESCO dataset using top-down constraints from satellite-based NO2. In the current configuration of LMDZINCA, soil NOₓ emissions are based on the CEDS anthropogenic emission inventory, therefore not taking into account NOₓ emissions from natural soils. Replacing these current CEDS soil NOₓ emissions with ORCHIDEE soil NOₓ emission includes the very substantial natural soil NOₓ component which has previously not been accounted for in LMDZINCA. The resulting changes on atmospheric NO2 in recent years are then analysed and compared with satellite derived NO2 columns from OMI and TROPOMI. These differences of soil NOX emissions and resulting tropospheric NO2 changes are studied over the last two decades.

The offline implementation of ORCHIDEE soil NOₓ in LMDZINCA represents a first step toward a fully coupled nitrogen cycle within the IPSL framework, enabling future assessments of feedbacks between terrestrial and atmospheric nitrogen reservoirs.