The Impact of Farming Mitigation Measures on Ammonia Concentrations and Nitrogen Deposition in the UK

Abstract

Ammonia (NH₃) is an important precursor to airborne fine particulate matter (PM_2.5) which causes significant health issues and can significantly impact terrestrial and aquatic ecosystems through deposition. The largest source of NH₃ emissions in the UK is agriculture, including animal husbandry and NH₃-based fertilizer applications. This study investigates the impact of mitigation measures targeting UK NH₃ emissions from farming activities, focusing on their implications for air quality and nitrogen deposition in 2030. A series of mitigation scenarios—low2030, medium2030, and high2030—were developed through engagement with stakeholders, including farmers, advisers, and researchers, and their impact was modelled using the CMAQ air quality model. These scenarios represent varying levels of the uptake of mitigation measures compared to a baseline (base2030). The results indicate that reductions in total NH₃ emissions across the UK could reach up to 13% under the high2030 scenario (but reaching nearly 20% for some regions). These reductions can lead to significant decreases in NH₃ concentrations in some parts of the UK (up to 22%, ~1.2 µg/m³) but with a mean reduction of 8% across the UK. However, the reductions have a limited effect on fine ammonium particulate (${\mathrm{N}\mathrm{H}}_4^{+}$) concentrations, achieving only modest reductions of up to 4%, with mean reductions of 1.6–1.9% due to a NH₃-rich atmosphere. Consequently, the mitigation measures have minimal impact on secondary inorganic aerosol formation and PM_2.5 concentrations, aligning with findings from other studies in Europe and beyond. These results suggest that addressing the primary sources of PM_2.5 or other PM_2.5 precursors, either alone or in combination with NH₃, may be necessary for more substantial air quality improvements. In terms of nitrogen (N) deposition, reductions in NH₃ emissions primarily affect NH₃ dry deposition, which constitutes approximately two-thirds of reduced nitrogen deposition. Total N deposition declines by 15–18% in source regions depending on the scenario, but national average reductions remain modest (~4%). While the study emphasizes annual estimates, further analyses focusing on finer temporal scales (e.g., daily or seasonal) could provide additional insights into exposure impacts. This research highlights the need for integrated mitigation strategies addressing multiple pollutants to achieve meaningful reductions in air pollution and nitrogen deposition.

1. Introduction

Air pollution is a global concern, affecting the environment, human health, and ecosystems. Among the atmospheric pollutants, ammonia (NH₃) stands out as a potent yet often overlooked contributor.

NH₃ is a gas that is emitted into the atmosphere. Most NH₃ emissions in the UK come from agriculture through the spreading of manures, slurries (semi-liquid manure), and fertilisers [1]. A smaller portion of NH₃ emissions stem from waste and a range of diffuse sources like synthetic fertilizers (particularly urea), and non-agricultural sources such as sewage, catalytic converters, wild animals, and industrial processes.

NH₃ only stays in the atmosphere for a few hours once emitted [2,3,4,5]. However, when NH₃ mixes with other gases in the atmosphere, such as nitrogen oxides (NO_x) and sulphur dioxide (SO₂), it can form particulate matter (PM) which can persist for several days and be transported large distances. There are major health concerns linked to exposure to fine airborne particulate matter (PM_2.5), to which NH₃ is an important contributor [6,7]. This PM_2.5 comprises particles with a diameter lower than 2.5 µm that infiltrate our lungs and bloodstream [8,9]. Indeed, secondary PM_2.5 aerosol consists of secondary organic aerosols (SOAs) related to anthropogenic and biogenic volatile organic compounds (VOCs), and also secondary inorganic aerosols (SIAs) which gather sulphate ${(\mathrm{S}\mathrm{O}}_4^{2+})$, nitrate (${\mathrm{N}\mathrm{O}}_3^{-}$), and ammonium (${\mathrm{N}\mathrm{H}}_4^{+}$). This ${\mathrm{N}\mathrm{H}}_4^{+}$ is formed through chemical reactions from NH₃. NH₃ also reacts with acidic compounds such as sulfuric acid (H₂SO₄) and nitric acid (HNO₃), forming ammonium sulphate ((NH₄)₂SO₄) and ammonium nitrate (NH₄NO₃), which are significant constituents of PM_2.5.

Thus, NH₃ plays a dual role. On the one hand, it sustains food production by providing essential nitrogen (N) for crops. On the other, it impacts the atmosphere, setting off a chain of chemical reactions that lead to the formation of PM_2.5.

NH₃ is not normally directly harmful to human health at typical outdoor air concentrations but is an important contributor to PM_2.5 and can also have significant impacts on the natural environment, both directly and by contributing to acid and nitrogen deposition. Indeed, NH₃ is known to contribute significantly to total N deposition into the environment [10,11,12]. In addition, reduced N gases and aerosols (NH₃ and ${\mathrm{N}\mathrm{H}}_4^{+}$, abbreviated as RDN), act as powerful plant and microorganism nutrients when deposited into aquatic and terrestrial ecosystems. Excessive N can lead to an exceedance of N critical loads, which in turn can lead to eutrophication and the loss of ecosystem productivity and biodiversity [13,14].

Mitigating NH₃ emissions is particularly important in the UK since the UK is a signatory of the United Nations Economic Commission for Europe (UNECE) Gothenburg protocol, legislated through the UK National Emission Ceilings Regulations adopted in 2018 [15]. In addition, the UK, with its diverse agricultural activities, transportation emissions, and industrial processes, is a notable region for studying the impact of NH₃ on PM_2.5 concentrations.

A previous review of agricultural interventions in the UK also identified a gap in the understanding of the level of the implementation of mitigation actions, the potential for future implementation, and the consequent impacts on public health [16]. This review highlighted the importance of evaluating the drivers and challenges facing farmers in introducing NH₃ mitigation measures, in terms of understanding the limitations to uptake.

Given the complexity of atmospheric chemistry, numerical models, such as the atmospheric chemistry transport models (CTMs), are commonly used to simulate the processes involved and estimate the outcomes of plausible control strategies. The Community Multiscale Air Quality (CMAQ) model [17,18], developed and distributed by the US Environmental Protection Agency (EPA) is a state-of-the-science numerical air quality model with comprehensive representations of the emission, transport, formation, destruction, and deposition of many air pollutants, including NH₃. Thus, the CMAQ model can be used to study the relationships between emissions such as NO_x, NH₃, and the wet and dry deposition of N since they simulate the main processes influencing the fate of atmospheric pollutants (turbulent dispersion, atmospheric chemistry, cloud processes, long-range transport, wet and dry deposition, etc.). Although they are no substitute for observations, CTM simulations have the advantage of estimating deposition rates for locations where there are no measurements and for processes for which measurements are difficult and/or sparse (e.g., dry deposition). They can also be used for simulating hypothetical scenarios, such as the effect of emission reduction strategies as carried out in this work.

This study considers a series of mitigation measures to understand the impact of emissions changes from farms across the UK on pollutant concentrations. The mitigation measures were modelled through scenarios which represented various levels of uptake (low–high) on farms across the UK in 2030 and were developed based on stakeholder engagement with farmers and stakeholders (i.e., farm advisers, academics, and farmer representatives) [19]. These measures are related to reducing emissions from the housing, storage, and spread of manure and slurry from the dairy, pig, and poultry sectors. This study has focused on the analysis of annual changes in NH₃ concentrations and their role in PM_2.5 formation and N deposition due to the uptake of different levels of intervention implementation of these farming activities. The methodology is described in Section 2. An evaluation of the NH₃ and ${\mathrm{N}\mathrm{H}}_4^{+}$ concentrations is presented in Section 3. Section 4 describes the changes in NH₃ concentrations and N deposition due to the emissions reductions predicted in the different scenarios. Section 5 provides the conclusions and perspectives.

2. Methodology

2.1. Model Set-Up

To undertake the study, the CMAQ model version 5.4 [20] was used in this work. The chemical mechanism used for the gaseous species is the carbon bond mechanism (CB06r5) [21] combined with the aerosol mechanism using the 7th generation aerosol module (AERO7) [22].

Meteorological fields have been generated with the Weather Research and Forecast (WRF) model version 4.5. The following configuration for WRF was used: the Rapid Radiation Transfer Model Global (RRTMG) for longwave [23] and Dudhia for shortwave radiation [24], ACM2 for the PBL scheme [25,26] with the Pleim-Xiu surface layer scheme [27], the Rapid Update Cycle Land-Surface Model [28], the Kain–Fritsch (KF) cumulus parametrization scheme [29], and the Noah-modified 21-category IGBP-MODIS land use.

In the CMAQ configuration used in this study, the dry deposition of gaseous species is simulated utilizing deposition velocity and the M3Dry aerosol deposition parameterization has been used [30]. As explained in [30,31], the M3Dry dry deposition calculations performed in CMAQ are designed to maintain maximum consistency with the flux calculations performed in the WRF Pleim-Xiu land surface model. The M3Dry calculations are performed on a grid-scale basis. Dry deposition is computed by electrical resistance analogy where concentration gradients are analogous to voltage, flux is analogous to current, and deposition resistance is analogous to electrical resistance [32]. It is important to note the bidirectional NH₃ surface flux was not enabled in the configuration since no development in this parameterization has been undertaken for this study.

The calculation of biogenic emissions has used an online module incorporated in CMAQ. This uses the Model of Emissions of Gases and Aerosols from Nature (MEGAN) (version 3.2) [33].

The domain design follows a nested approach involving dividing a large geographic domain into smaller, nested grids, with each grid representing a progressively finer level of detail. This hierarchical structure allows us to capture more detailed variations in emissions and meteorology. A 50 km horizontal resolution over Europe (EU50) has been selected as the outer domain while the studied domain over the UK is characterized by a 10 km horizontal resolution (UK10), as shown in Figure 1.

The selected meteorological year feeding the air quality simulation was 2019. This 2019 simulation was used for model performance evaluation. It is worth noting that 2019 can be seen as a typical meteorological year in the UK since it was not a stormier year compared to recent decades, there was no abnormal wind speed record, and it was not drastically wetter than the previous 20-year period (1% wetter than 1981–2010) [34]. Even if 2019 was the 12th warmest year for the UK in a series from 1884; it is worth noting that all the top 10 warmest years for the UK in this series are recent since they have occurred since 2002. This shows that 2019 was not an abnormal year in terms of temperature for this recent decade. The year 2019 was also the most recent UK emissions year at the beginning of the project. Since the objective of the study is to analyse the impact of mitigation measures in agriculture and not future climate conditions, no climate projections have been tested in this work.

The air quality simulation started with a spin-up period of 2 weeks. The simulations were defined as “forecast-cycling experiments”; i.e., the calculated fields have been used to initialize the successive (following-day) simulation [35] and the full year 2019 is calculated for both domains.

The initial and boundary conditions for the outer domain (EU50) are based on prescribed concentrations representative of annual background conditions provided by the US EPA [36]. Then, the CMAQ concentrations calculated within this EU50 domain were used as boundary conditions for the UK10 domain.

2.2. Emission Scenarios

The study focused on three future scenarios in 2030 in the UK, in addition to the baseline; thus, only the emissions used for the UK scenarios vary and all use common boundary conditions corresponding to the baseline from the EU50 domain.

The baseline 2030 future scenario for the EU50 domain was based on the European Monitoring and Evaluation Programme (EMEP) [37] gridded emissions for 2019 and scaled with the factors provided by the GAINS ECLIPSE (Greenhouse Gas and Air Pollution INteractions and Synergies—Evaluating the Climate and Air Quality Impacts of Short-Lived Pollutants) V6b Baseline CLE scenario [38]. All the UK10 emissions for the scenarios were built using the Scenario Modelling Tool [39].

The 0.1 × 0.1-degree-resolution anthropogenic data from EMEP were post-processed into 50 × 50 km to populate the EU50 domain in CMAQ. The sectoral splits defined by EMEP were used and temporally allocated to the emissions by sector. The speciation profiles required by CMAQ were also used.

Anthropogenic emissions across the UK, including from agriculture, are based on the gridded emissions from the UK National Atmospheric Emission Inventory (NAEI) for 2019 [40]. The 1 km × 1 km grids were resampled into 10 × 10 km emission grids for the UK10 domain. The 2019 large-point-source emission inventory was also used to vertically allocate the emissions in the CMAQ grid.

Except for the UK 2030 baseline, all UK scenarios included the same measures and increasing the uptake of these measures reflects an increased ambition to mitigate air pollutants across the low, medium, and high uptake scenarios. To simplify the terminology, these scenarios are hereinafter named base2030, low2030, medium2030, and high2030.

The scenarios focus on implementing mitigation measures on dairy, pig, and poultry farms. The uptake scenarios were developed through stakeholder engagement with farmers and stakeholders and the details on the design of these scenarios are described in [19]. These scenarios included 19 mitigation measures, and 2030 was selected as a realistic timeline for the practical implementation of new activities on farms. The selection of mitigation actions focused on controlling NH₃ emissions and included actions related to livestock diet, livestock housing, air filtration, and the improved storage and spread of solid and liquid livestock manures as shown in [41].

The 10 × 10 km gridded NH₃ emissions for base2030 used in CMAQ are shown in Figure 2a. The NH₃ area sources are well highlighted by the map in Figure 2a. These source regions are mainly located in Northern Ireland, Staffordshire–West Midlands, Suffolk, Cornwall, and Devon to cite a few in the UK.

The scenarios mainly target the NH₃ emissions, which are characterized by a larger decrease (in the total emissions) compared to the base2030, by 8.5%, 9.3%, and 12.9% for the low2030, medium2030, and high2030 scenario, respectively (see Figure 2b). It is worth noting that for some locations, the decreases in NH₃ emissions reach up to 20%, 22%, and 24% for the low2030, medium2030, and high2030 scenarios, respectively (Figure S1). Carbon monoxide (CO) emissions are predicted to be constant for all of the scenarios and are 6.5% lower than the base2030 emissions. The other emissions (VOCs, PM₁₀, PM_2.5, SO₂, and NO_x) do not vary significantly. Slightly larger decreases in emissions are calculated for the high2030 scenario for VOCs and PM₁₀, while the changes in NO_x and PM_2.5 remain marginal, and are null for SO₂.

3. Evaluation

For the evaluation of the modelled concentrations, only NH₃ measurement data from rural background measurement sites with at least 90% data capture in the year were used to avoid bias. The observations are taken from the UK AIR platform including observations of the UK Eutrophying and Acidifying Atmospheric Pollutants (UKEAP) network and from Monitor for AeRosols and Gases in Air (MARGA) instruments [42]. This represents a total of 45 stations. The CMAQ annual map and comparison with observations at the measurement sites are shown in Figure 3.

This comparison shows CMAQ fairly represents the annual mean distribution of the NH₃ concentrations in 2019 (r ~0.4) with a small underestimation (MB ~0.2 µg/m³; NMB ~10%), a low mean relative error (MRE < 0.9%), and a reasonable root mean square error (RMSE ~1.6 µg/m³). These statistics are explained in Appendix A.

The result of the RMSE needs to be kept in perspective since most of the annual mean measured concentrations are relatively small except for Brompton station (with a measured concentration of 9.3 µg/m³), as shown on the map and the scatterplot. This comparison is relatively good with a somewhat coarse horizontal resolution and satisfactory especially knowing that a previous study [43] showed larger discrepancies with the same observations but using a global model with a coarser horizontal resolution (~25 km latitude × ∼31 km longitude). It is worth noting that a sensitivity simulation was undertaken by increasing the NH₃ emissions by 50% since [43] suggested these emissions were underestimated by about half in comparison to satellite-derived estimates. This 50% increase was also applied by [44] for their calculation of UK PM_2.5 concentrations. However, this recommendation leads to an overestimation of our modelled NH₃ concentrations by ~38% (see Figure S2) and does not improve the PM_2.5 simulations, as shown in [41].

The NH₄⁺ concentrations were also evaluated. However, with the stringent criteria used for NH₃, only two stations are available. Thus, a data capture of 75% was used and, for comparison, a similar evaluation of the NH₃ concentrations using this new criterion is also shown. These results are presented in Figure S3, showing two groups of points, leading to an overall disagreement. However, the NH₄⁺ concentrations from the model and the measurements remain low.

The meteorological conditions in the UK, as used in CMAQ, were well represented, as shown in the comparison with the measurements of the NOAA Integrated Surface Database (https://www.ncei.noaa.gov/products/land-based-station/integrated-surface-database, accessed on 31 January 2025) (Table S1). Despite the lower correlation in the wind direction (0.34), all parameters (wind direction, speed at 10 m, and temperature at 2 m) have a minimal bias and good IOA (0.5–0.7) (Table S1). These statistics might be further improved for future studies by using Newtonian nudging in the WRF model.

4. Analysis on Predictions

4.1. NH₃ Concentrations

Emissions reductions in the three scenarios are effective at reducing NH₃ concentrations, as shown in Figure 4, especially across the source regions. The three scenarios (low, medium, and high) present similar maximum reductions in the annual mean NH₃ concentration, ranging from 18% to 22% (up to 1 and 1.2 µg/m³ reduction, respectively) in the UK, but it is clear the high2030 scenario shows the larger reduction in terms of both the mean value and the median value. The mean decrease in the annual NH₃ concentration across the whole UK10 domain is always close to 4% and the median reduction is between 2 and 3%, but reaches 7–8% (mean) and nearly 7% (median) across UK land grid cells.

However, the emissions scenarios have only a minor impact in terms of reducing NH₄⁺ (within PM_2.5), as shown in Figure S4 (always up to a 4% reduction, with a mean reduction ranging only from 1.6 to 1.9% in the UK). To assess the role of NH₃ in PM_2.5 formation, Figure 5a shows the free ammonia (F-NH_x) for the base2030 scenario.

This F-NH_x is the amount of NH₃ available, after neutralizing ${\mathrm{S}\mathrm{O}}_4^{2+}$, for NH₄NO₃ formation mainly. This indicator reflects that the (NH₄)₂SO₄ aerosol is the favoured form for ${\mathrm{S}\mathrm{O}}_4^{2+}$ [45]. It is expressed in a molar basis (µmole/m³) and defined as:

F-NH_x = NH₃ + NH₄ − 2 × SO₄

(1)

A positive value of this F-NH_x indicates an excess of NH₃ and so NH₃ is not the limiting species.

To complete the analysis, G_ratio is also used (Figure 5b) [46], indicating whether fine-particle nitrate formation is limited by the availability of HNO₃ or NH₃ and expressed in a molar basis as follows:

$${\mathrm{G}}_{\mathrm{ratio}}={\displaystyle \raisebox{1ex}{${\mathrm{F}\text{-}\mathrm{N}\mathrm{H}}_{\mathrm{x}}$}\!\left/ \!\raisebox{-1ex}{${(\mathrm{N}\mathrm{O}}_3+{\mathrm{H}\mathrm{N}\mathrm{O}}_3)$}\right.}$$

(2)

A G_ratio larger than 1 indicates that HNO₃ is limiting for the formation of ${\mathrm{N}\mathrm{O}}_3^{-}$. A value lower than 0 indicates that NH₃ is severely limiting, and a G_ratio between 0 and 1 indicates NH₃ is the limiting species and available for reaction with HNO₃.

Similar results are calculated for the three scenarios (low, medium, and high) and are shown in Figures S5 and S6, for F-NH_x and G_ratio, respectively.

It is noticeable that high F-NH_x concentrations (>0) and G_ratio (>1) are present in the UK, especially across the source regions showing a NH₃-rich chemical regime calculated within the domain (see Figure 5a,b). This confirms the findings of [14]. Their study showed that a decrease in NH₃ emissions only has limited effects on mitigating SIA formation. As a consequence, the PM_2.5 concentrations will only be slightly impacted by the mitigation of agricultural activities implemented in the different scenarios. Ref. [47] also found a shift in the SIA formation regime in the rural United States, making rural regions less sensitive to changes in NH₃. It was also found that the potential to curb PM_2.5 levels in Europe through reductions in NH₃ emissions is decreasing due to an increase in NH₃ compared to SO_x and NO_x [48]. Ref. [49] also concluded that NH₃ emissions reduction is more efficient in Europe where NH₃ is less abundant than where it is abundant.

Since most parts of the UK are characterized as “NH₃ rich”; this indicates that whilst reducing NH₃ emissions will decrease the Reduced N (RDN = NH₃ + NH₄⁺) concentration, it will have little effect on mitigating SIA as also suggested by the study conducted by [14].

This limited impact on the secondary formation of PM_2.5, also suggests that PM_2.5 precursors other than NH₃ emissions, or even primary PM_2.5 emissions, may need to be tackled to reduce annual mean PM₂.₅ concentrations due to farming activities. Indeed, ref. [50] also concluded that targeting acidic aerosol ${(\mathrm{S}\mathrm{O}}_4^{2+}$ and ${\mathrm{N}\mathrm{O}}_3^{-}$) precursor emissions of SO₂ and NO_x is more efficient in terms of reducing PM_2.5 concentrations in the UK. Ref. [51], also focusing on the impact of UK NH₃ emissions from agriculture, found that other precursors of PM_2.5 should potentially reduce the overall PM_2.5 load.

The role of NO_x emissions in the results has not been analysed, but the minor changes in NO_x emissions in the scenarios (Figure 2b) likely have a marginal impact on NH₃ and NH₄⁺. It is also worth noting that this analysis has focused on annual changes; the impact of reductions in NH₃ emissions might provide a different picture if another temporal resolution, such as seasonal or monthly averages, is considered e.g., [49].

4.2. N Deposition

In addition to the reduction in NH₃ concentrations, the scenarios show decreases in total N deposition (see Figure 6) in areas within the NH₃ emission hotspots (see Figure 2) and which are impacted by the reductions in emissions. This reduction in total N deposition reaches a value between roughly 15 and 18%, but with a mean reduction of only 4% in the UK and around 1.3 and 1.5% in the whole domain, depending on the scenario.

This change in total N deposition is mainly due to a change in RDN deposition (Figure 7) and not in oxidised nitrogen (OXN). Indeed, this RDN is dominant in the fraction of total N deposition (Figure 8) and this RDN deposition in the UK is mainly composed of NH₃ dry deposition in the different scenarios (Figure 9).

It is worth noting that CMAQ assumes that gaseous NH₃ is transformed into ${\mathrm{N}\mathrm{H}}_4^{+}$ when it dissolves in water, explaining why the NH₃ wet deposition is not represented in the distribution of the RDN deposition in Figure 9. Figure 9 also shows that the different scenarios have a limited impact on the type of deposition since the percentages slightly vary. The scenarios show a small decrease in the NH₃ dry deposition (with a ratio close to 67%) while the ${\mathrm{N}\mathrm{H}}_4^{+}$ slightly increases, up to 10% for the dry deposition and 24% for the wet deposition.

The RDN deposition follows the same trend as the total N deposition, with larger decreases from the low2030 scenario to the medium2030 and high2030 scenarios, with a decrease reaching up to 17–20% depending on the scenario (with a mean between 5 and 6% in the UK and nearly 2% in the whole UK10 domain) (Figure 7). It also shows RDN is deposited closer to its sources.

Therefore, as shown by [14], gaseous NH₃ dominates RDN. This indicates that, although reducing NH₃ emissions over the source regions will effectively decrease RDN for the three scenarios, it will have little effect on mitigating SIA formation.

This change in total N deposition impacts biodiversity, as shown by [52]. Using these CMAQ calculations on the deposition, ref. [52] studied the economic impact of the change in biodiversity in terms of the appreciation of biodiversity, and focusing on bog, acid grassland, heathland, and dune grassland habitats. They found a larger impact for bogs in Northern Ireland for the high2030 scenario (1.20%). For acid grassland, the most impact occurred in Wales (3.12%) and for heathland, the greatest increases in biodiversity occurred in Northern Ireland (4.25%). Dune grassland experienced the greatest increase in biodiversity in England and Wales (~1.9–2%) due to this change in N deposition. A health impact assessment based on PM_2.5 exposures and an economic analysis, capturing the valuation of health, productivity, and costs to the National Health Service (NHS) completed their study. Overall, ref. [52] found the mitigation measures in the different scenarios would have small positive effects on health, the economy, and ecosystems.

5. Conclusions

To study the impact of mitigation measures on farming activities in the UK, mainly tackling NH₃ emissions, a series of mitigation measures were developed and the state-of-the-science air quality numerical model, CMAQ, was used.

These scenarios represented various levels of uptake (low to high) on farms across the UK in 2030 and were developed based on stakeholder engagement with farmers and stakeholders (i.e., farm advisers, academics, and farmer representatives). These scenarios were named low2030, medium2030, and high2030; compared to the baseline (base2030).

Considering this series of measures impacting emissions from farms across the UK, this study has shown a clear decline in NH₃ emissions in the different scenarios in 2030 (up to ~13% reduction for the total emissions, and up to 24% for some source regions in the high2030 scenario).

Using the CMAQ model, this modelling study has highlighted reductions in NH₃ emissions are effective at reducing NH₃ concentrations (up to 22% for the high2030 scenario, but with a mean reduction of 8% in the UK), but are considerably less effective at reducing ${\mathrm{N}\mathrm{H}}_4^{+}$ concentrations (only up to 4% reduction in all scenarios, with a mean reduction ranging only from 1.6 to 1.9% in the UK). This limited impact of NH₃ emissions on ${\mathrm{N}\mathrm{H}}_4^{+}$ concentrations is explained by a HNO₃-limited regime in the UK, shown by high F-NH_x and G_ratio values over the whole UK. Since NH₃ is not ‘limiting’, these NH₃ emissions reductions have a small or no impact on mitigating secondary inorganic aerosol formation and thus on the PM_2.5 concentrations, confirming results from other studies in Europe or other countries [14,47,48]. This also suggests that tackling other PM_2.5 precursors (solely or combined) or potentially primary PM_2.5 might be required to reduce their airborne concentrations [48,49]. However, this study focused on annual changes and conclusions might differ if other periods (e.g., daily, monthly, seasonal) are selected for exposure analysis.

Similarly, the changes in emissions also impact the N deposition which is mainly dominated by NH₃ dry deposition (almost 2/3 of the RDN deposition). The reduction in total N deposition ranges between 15 and 18%, but with a mean reduction of only around 4% (3.5 and 4.2% from the low2030 to high2030 scenarios). This reduction in the N deposition is mainly calculated near the source regions.

Finally, further model development should be conducted to assess the results, as this study does not incorporate the bi-directional NH₃ flux representation in CMAQ. This bi-directional treatment of NH₃ fluxes should improve the prediction of NH₃ in the air [31,53]. The modelling might also benefit from a finer spatial resolution, especially in analysing the deposition.