Global agricultural N₂O emission reduction strategies deliver climate benefits with minimal impact on stratospheric O₃ recovery

The study addresses how mitigating agricultural N₂O emissions—an increasingly important greenhouse gas and stratospheric ozone-depleting substance—will affect stratospheric ozone recovery and climate. N₂O has risen from ~270 ppbv pre-industrially to 332 ppbv in 2019 and now contributes ~0.21 W m⁻² radiative forcing, with growth dominated by agriculture. While reducing N₂O is a key element of net-zero strategies, N₂O’s role in stratospheric chemistry—via NOx-driven O₃ loss and interactions with HOx and ClOx families—complicates expectations for ozone recovery, particularly as stratospheric chlorine continues to decline under Montreal Protocol compliance. Few studies have examined sustained N₂O emission reductions within chemistry–climate model frameworks across multiple future scenarios. This work investigates whether a realistic, spatially targeted ~5% global N₂O emission cut (~25% of direct agricultural emissions) can deliver climate benefits without impeding total column ozone (TCO) recovery.

The paper synthesizes prior understanding that N₂O is the dominant ozone-depleting substance emitted in the 21st century and a potent long-lived GHG. Earlier modeling showed increases in N₂O decrease stratospheric O₃, largely via NOx catalytic loss, and that coupling between NOx, ClOx, and HOx can modulate O₃ destruction (e.g., sequestration of chlorine into ClONO₂). Multi-model projects (CCMI, AerChemMIP) provide context for model sensitivity to N₂O perturbations. The literature indicates that while cross-family coupling exists, its net effect on O₃ tends to be smaller than direct NOx-driven loss changes. Few studies assessed sustained N₂O emission reductions concurrently with evolving GHGs and ODSs under future SSP scenarios, motivating the present analysis.

Agricultural N₂O abatement scenario: Employed the NMIP dataset (seven process-based biosphere models) for 2010–2016 as baseline direct agricultural N₂O emissions at 50×50 km, monthly resolution. Two mitigation strategies were combined with spatial separation: (1) Enhanced rock weathering (ERW) via crushed basalt applied to croplands in five regions (North America, Brazil, Europe, India, China) targeting 2 GtCO₂ yr⁻¹ removal, increasing soil pH and reducing N₂O; applied across 400 Mha, reducing direct agricultural soil N₂O from 5.19 to 4.69 Tg N₂O yr⁻¹ (−0.50 Tg). (2) Nitrification inhibitors applied to agricultural grid cells without ERW in the Global North assuming 50% reduction in soil N₂O; applied to ~600 Mha, further lowering crop soil N₂O from 4.69 to 3.84 Tg N₂O yr⁻¹ (−0.85 Tg). Combined mitigation: −1.35 Tg N₂O yr⁻¹ (~25% of global direct agricultural N₂O; 40% in primary regions). Economic parameters: nitrification inhibitor cost $28–45 ha⁻¹ to 600 Mha ($17–27B yr⁻¹), abatement cost $70–113 per tCO₂e; ERW N₂O reduction (0.47 Tg N₂O yr⁻¹) is a no-cost co-benefit that reduces abatement costs by 2.3–9% depending on region. Earth system modeling: Used UKESM1.0 fully coupled (1.25°×1.9°, 85 levels to ~85 km) with interactive stratosphere–troposphere chemistry (UKCA StratTrop) and GLOMAP-mode aerosols (no nitrate aerosol). Well-mixed GHGs (CO₂, CH₄, N₂O) prescribed via lower boundary conditions (LBCs) per SSP1-2.6 and SSP3-7.0. Anthropogenic and biomass burning emissions, nitrogen deposition, and land-use fractions from Input4MIPs followed SSPs. Experiment design: Compared control SSP126 (16-member ensemble) and SSP370 (15-member ensemble) with perturbed SSP126_low_N₂O and SSP370_low_N₂O (each 3 members) run 2025–2075, identical to controls except adjusted N₂O LBC reflecting −1.35 Tg N₂O yr⁻¹ sustained emissions. Additional single-member SSP126_flux and SSP370_flux runs replicated controls but output catalytic O₃ loss reaction fluxes to diagnose mechanisms. N₂O LBC adjustment: Reduced emissions E(t) by 1.35 Tg N₂O yr⁻¹ and solved d[N₂O]/dt = E(t) − N₂O/τ(N₂O), with τ(N₂O) = 139 (C_PI/C(t))⁻⁰·⁰⁴, then scaled mixing ratios by 1.033 to match UKESM1 LBC bias (~3.3%). CO₂ scenario perturbations: Used FAIR v2.1.0 (AR6-calibrated) to estimate atmospheric CO₂ concentration response to sustained −2 GtCO₂ yr⁻¹ removal from 2025 in SSP126 and SSP370; explored AR6-assessed uncertainties. Applied differences in CO₂ to radiative forcing calculations. Diagnostics and forcing: Reaction fluxes (e.g., NO₂+O, ClO+O, HO₂+O/O₃) computed online and normalized by grid-cell volume. Radiative forcing from N₂O and CO₂ changes estimated using Etminan et al. kernels with scenario-appropriate backgrounds; O₃ RF estimated with a whole-atmosphere kernel (Rap et al., updated by Iglesias-Suarez et al.) applied to mean O₃ differences (low_N₂O minus control) for 2040–2050 and 2065–2075. Analyses focused on zonal-mean O₃ responses (5–20 hPa mid-latitudes/tropics), polar regions, TCO time series (global, and 75–90°S October), and attribution via spatial (anti)correlation with catalytic loss flux changes.

• The imposed −1.35 Tg N₂O yr⁻¹ mitigation lowers N₂O throughout the stratosphere, exceeding −5 ppb by 2065–2075 in the low- and mid-stratosphere under both SSP1-2.6 and SSP3-7.0, with associated NOy decreases.
• O₃ increases regionally (notably 5–20 hPa across tropics and mid-latitudes), with annual mean increases >100 ppb (~1–1.5%) in SSP370 by 2065–2075 (95% significant), and smaller, seasonally consistent increases under SSP126. The response is stronger in 2065–2075 than 2040–2050 due to larger N₂O differences.
• Attribution: Spatial anti-correlation between O₃ changes and reductions in NOx-driven O₃ loss (NO₂ + O) indicates NOx decreases are the main driver. Cross-family coupling impacts via ClOx and HOx are weak and generally not robust relative to ensemble variability.
• Wider chemistry: Global vertical changes in NOx, ClOx, HOx, and ClONO₂ are <2% and mostly within ±1σ of control ensembles; HNO₃ occasionally exceeds 1σ but remains <2%, implying minimal perturbation to broader stratospheric composition.
• Total column ozone (TCO): No statistically significant differences between low_N₂O and controls in global annual TCO or in 75–90°S October TCO for either SSP; TCO evolution is dominated by background climate scenario (e.g., greater stratospheric cooling and higher tropospheric O₃ in SSP3-7.0 yield higher TCO than SSP1-2.6).
• Radiative forcing: N₂O reductions yield RF of −10 (−18) mW m⁻² at 2050 (2075) in SSP370 and −12 (−22) mW m⁻² in SSP126, equivalent to ~11–13% of historical (PI to present) N₂O forcing by 2075. ERW-driven −2 GtCO₂ yr⁻¹ removal lowers atmospheric CO₂ by ~4.3 (8.6) ppm in SSP370 and ~4.1 (7.4) ppm in SSP126 by 2050 (2075). Combined N₂O+CO₂ effects correspond to an effective CO₂ reduction of ~10.9 ppm (SSP370) and ~9.4 ppm (SSP126) by 2075. O₃ RF is not robust under either scenario.
• Economics: Nitrification inhibitors over ~600 Mha cost ~$17–27B yr⁻¹ with abatement cost $70–113 per tCO₂e; ERW’s ~0.47 Tg N₂O yr⁻¹ reduction is a cost-free co-benefit that reduces abatement costs by 2.3–9% regionally.

The results show that sustained, spatially targeted reductions in agricultural N₂O emissions increase stratospheric O₃ regionally via decreased NOx-driven destruction, while leaving total column ozone recovery essentially unaffected. This addresses the key concern that N₂O mitigation might hinder ozone recovery amid declining stratospheric chlorine: despite some competing effects (CO₂-driven stratospheric cooling and higher CH₄ increase HOx and reduce NOx efficiency; declining chlorine can enhance N₂O’s relative O₃ destruction efficiency), the net impact on TCO is minimal across divergent climate scenarios (SSP3-7.0 and SSP1-2.6). The absence of significant TCO changes, including at high southern latitudes in spring, implies compatibility of N₂O mitigation with Montreal Protocol-driven ozone recovery timelines. Concurrently, the N₂O reductions and ERW-associated CO₂ removal deliver clear climate benefits (negative radiative forcing, effective multi-ppm CO₂-equivalent reductions by 2075) and potential ecological co-benefits (e.g., reduced nitrate leaching with inhibitors). Observed high-latitude O₃ variability, especially in the Arctic, appears dominated by dynamics rather than direct N₂O effects within this mitigation magnitude, consistent with the lack of robust attribution in those regions.

This study demonstrates that a realistic global agricultural N₂O mitigation portfolio—combining regionally targeted nitrification inhibitors and enhanced rock weathering—can reduce N₂O emissions by ~1.35 Tg N₂O yr⁻¹ (~5% of global, ~25% of direct agricultural emissions), increase stratospheric O₃ regionally via reduced NOx-driven loss, and yield meaningful negative radiative forcing, all without compromising total column ozone recovery under diverse future climate pathways. The ERW strategy additionally delivers substantial CO₂ removal and provides a no-cost co-benefit for N₂O abatement, whereas nitrification inhibitors entail costs but confer environmental co-benefits. These findings support inclusion of N₂O mitigation in national net-zero strategies. Future work could broaden robustness by multi-model assessments, exploring larger or alternative spatial deployments and additional mitigation measures, and quantifying potential interactions with aerosol–chemistry processes not included here (e.g., nitrate aerosol), while continuing to evaluate regional dynamical influences on polar ozone.

• Model scope: Results are from a single Earth System model (UKESM1); while sensitivities align with multi-model studies, broader model intercomparison would strengthen confidence.
• Diagnostics availability: Full control ensembles lacked reaction flux outputs; attribution relied on comparison to single-member flux runs, introducing some uncertainty in mechanistic attribution where single vs. full comparisons diverged.
• Ensemble size: Perturbed scenarios used three ensemble members, limiting sampling of internal variability, particularly in dynamically variable polar regions.
• Chemistry–aerosol representation: Nitrate aerosol chemistry was not included in the UKESM1 configuration used (consistent with CMIP6 version), potentially omitting interactions that may subtly affect ozone/NOy.
• Implementation approach: N₂O mitigation was imposed via adjusted lower boundary conditions calibrated to an emissions reduction; while appropriate for comparability to ScenarioMIP, it abstracts from explicit emission–transport processes.
• Attribution in polar regions: High-latitude (especially Arctic) O₃ changes were not statistically significant and appear dominated by dynamical variability, limiting robust attribution to N₂O changes.