Chloride (HCl ∕ Cl⁻) dominates inorganic aerosol formation from ammonia in the Indo-Gangetic Plain during winter: modeling and comparison with observations

The Winter Fog Experiment (WiFEX) was an intensive field campaign conducted at Indira Gandhi International Airport (IGIA) Delhi, India, in the Indo-Gangetic Plain (IGP) during the winter of 2017–2018. Here, we report the first comparison in South Asia of high-temporal-resolution simulation of ammonia (NH₃) along with ammonium (NH${}_{\mathrm{4}}^{+}$) and total NH_x (i.e., NH₃+ NH${}_{\mathrm{4}}^{+}$) using the Weather Research and Forecasting model coupled with chemistry (WRF-Chem) and measurements made using the Monitor for AeRosols and Gases in Ambient Air (MARGA) at the WiFEX research site. In the present study, we incorporated the Model for Simulating Aerosol Interactions and Chemistry (MOSAIC) aerosol scheme into WRF-Chem. Despite simulated total NH_x values and variability often agreeing well with the observations, the model frequently simulated higher NH₃ and lower NH${}_{\mathrm{4}}^{+}$ concentrations than the observations. Under the winter conditions of high relative humidity (RH) in Delhi, hydrogen chloride (HCl) was found to promote the increase in the particle fraction of NH${}_{\mathrm{4}}^{+}$ (which accounted for 49.5 % of the resolved aerosol in equivalent units), with chloride (Cl⁻) (29.7 %) as the primary anion. By contrast, the absence of chloride (HCl $/$ Cl⁻) chemistry in the standard WRF-Chem model results in the prediction of sulfate (SO${}_{\mathrm{4}}^{\mathrm{2}-}$) as the dominant inorganic aerosol anion. To understand the mismatch associated with the fraction of NH_x in the particulate phase (NH${}_{\mathrm{4}}^{+}$ $/$ NH_x), we added HCl $/$ Cl⁻ to the model and evaluated the influence of its chemistry by conducting three sensitivity experiments using the model: no HCl, base case HCl (using a published waste burning inventory), and 3 × base HCl run. We found that 3 × base HCl increased the simulated average NH${}_{\mathrm{4}}^{+}$ by 13.1 µg m⁻³ and NH_x by 9.8 µg m⁻³ concentration while reducing the average NH₃ by 3.2 µg m⁻³, which is more in accord with the measurements. Thus HCl $/$ Cl⁻ chemistry in the model increases total NH_x concentration, which was further demonstrated by reducing NH₃ emissions by a factor of 3 (−3 × NH₃_EMI) in the 3 × base HCl simulation. Reducing NH₃ emissions in the 3 × base HCl simulation successfully addressed the discrepancy between measured and modeled total NH_x. We conclude that modeling the fate of NH₃ in Delhi requires a correct chemistry mechanism accounting for chloride dynamics with accurate inventories of both NH₃ and HCl emissions.

1 Introduction

The Indo-Gangetic Plain (IGP) is one of the global hotspots of atmospheric ammonia (NH₃) and faces a range of environmental challenges, particularly during the winter season, including adverse air pollution episodes, especially as NH₃ plays a substantial role in secondary aerosol formation (Ghude et al., 2020, 2008a, b; Kumar et al., 2021; Saraswati et al., 2019; Sharma et al., 2020; Singh et al., 2021). Atmospheric NH₃, along with oxides of nitrogen (NO_x), together accounts for the largest source of reactive nitrogen (N_r), which is primarily emitted by agricultural activities, livestock population, industrial activities, and transportation (Ghude et al., 2009, 2010, 2012, 2013; Móring et al., 2021; Pawar et al., 2021). Ammonia in the environment plays a crucial role in atmospheric chemistry and the eutrophication and acidification of ecosystems (Datta et al., 2012; Pawar et al., 2021; Sharma et al., 2008, 2012, 2014b). Control of NH₃ becomes a key priority in an emerging international strategy to manage the global nitrogen cycle (Gu et al., 2021; Sutton et al., 2020). Ammonia is one of the important aerosol precursor gases, and ammonium (NH${}_{\mathrm{4}}^{+}$) is a major counter ion for the three anions such as chloride (Cl⁻), nitrate (NO${}_{\mathrm{3}}^{-}$), and sulfate (SO${}_{\mathrm{4}}^{\mathrm{2}-}$) contributing to PM_2.5 composition (Seinfeld et al., 2016). In addition, as the dominant alkaline gas in the atmosphere, NH₃ has attracted the interest of scientific researchers since it has been known to promote new aerosol formation, both in the initial homogeneous nucleation and in the subsequent growth, especially during wintertime (Acharja et al., 2020, 2021; Ali et al., 2019; Duan et al., 2021; Wagh et al., 2021).

In this study, we focus on wintertime since this season is characterized by low-to-dense fog events, lower temperature (T), and variability of relative humidity (RH), which fluctuates from 40 % to 100 % (Ghude et al., 2017; Kumar et al., 2020). Ammonia acts as a neutralization agent for determining the acidity of aerosol particles (Acharja et al., 2020; Ali et al., 2019; Ghude et al., 2017). It also affects PM_2.5, the acidity of clouds, and the wet deposition of nitrogen by neutralizing acidic species (Gu et al., 2021; Xu et al., 2020). Increasing NH₃ concentration over Delhi compared with the surrounding area leads to an increase in PM_2.5 concentrations (Ghude et al., 2022; Sharma et al., 2008, 2012, 2014a), which in turn affects air quality, human health, and climate (Behera et al., 2013; Ghude, 2016; Ghude et al., 2008b; Nivdange et al., 2022; Sutton et al., 2017; Sutton and Howard, 2018).

Satellite observations (Van Damme et al., 2018; Warner et al., 2017), chemical transport models (CTMs) (Clarisse et al., 2009, 2010; T. Wang et al., 2020), and ground-based observations (Pawar et al., 2021) revealed that the IGP is the largest regional hotspot of NH₃ concentrations on Earth. Previous studies have identified various sources of NH₃, for example, agricultural activities, industrial sectors, motor vehicles, garbage, sewage, and urine from rural populations at the global scale (Behera et al., 2013; Huang et al., 2012; Sutton et al., 2008). However, in Delhi, agricultural activity (including surrounding arable and suburban livestock farming) is estimated to be the dominant source of NH₃, along with traffic emissions (Kuttippurath et al., 2020; Móring et al., 2021; Sharma et al., 2020), but its emissions are subject to large uncertainty. Globally, various modeling efforts have investigated the relative effectiveness of reducing NH₃ emissions in curtailing PM_2.5 formation (Gu et al., 2021; Pinder et al., 2007, 2008; Zhang et al., 2020). However, over India, the impact on reducing PM_2.5 might be limited because NH₃ emission reductions may be more challenging due to its diverse and area-wide sources. Ianniello et al. (2010) and Lan et al. (2021) have investigated the variation of atmospheric NH₃ at an urban and suburban site of Beijing with respect to meteorological factors, where RH was found to be a strong factor in influencing the NH₃ mixing ratio. A few studies over Asia have highlighted the gas-to-particle conversion of NH₃ in Delhi (Acharja et al., 2021; Saraswati et al., 2019) and China and its subsequent impact on the aerosol formation (Wang et al., 2015; Xu et al., 2020). Furthermore, excess NH₃ during fog can also enhance secondary aerosol formation in Delhi during winter (Acharja et al., 2021). However, the wintertime behavior of NH₃ in Delhi in CTMs has not yet been investigated and remains poorly understood (Ellis et al., 2011; Metzger et al., 2006). In a recent study, Pawar et al. (2021) highlighted uncertainties associated with gas-to-particle partitioning of NH₃ in a global model (Model for Ozone And Related chemical Tracers, MOZART-4) and found a significant overestimation of NH₃ in the model compared with the measurements. The overestimation of NH₃ in the model led the authors to hypothesize that a source-specific NH₃ emission inventory in India, considering agricultural statistics on fertilizer use and animal distribution, was missing. Also, there was a need for a high-resolution regional model with advanced chemistry to resolve the NH₃ emissions on the local scale.

The present study utilizes the regional Weather Research and Forecasting model coupled with chemistry (WRF-Chem) interpreted using measurements from the Winter fog Experiment (WiFEX), including NH₃, water-soluble ions in PM_2.5, other trace gases, and meteorological parameters during December–January 2017–2018. For the first time in South Asia, we discuss and compare the modeled and observed temporal variation in gaseous NH₃, particulate NH${}_{\mathrm{4}}^{+}$, and total NH_x (i.e., NH₃+ NH${}_{\mathrm{4}}^{+}$). Since we found that the total modeled NH_x matches well with the observations, we investigate the ability of the model to accurately describe the gas-to-particle partitioning of the measurements (Monitor for AeRosols and Gases in Ambient Air, MARGA) by evaluating the fraction of NH_x in the particulate phase (NH${}_{\mathrm{4}}^{+}$ $/$ NH_x). We conducted several sensitivity experiments with and without adding anthropogenic waste burning emissions of hydrochloric acid (HCl) in the model. The updated model with HCl $/$ Cl⁻ chemistry was used to analyze and compare the temporal variation of NH₃, NH${}_{\mathrm{4}}^{+}$, and total NH_x from the WiFEX measurements.

2 Data and methodology

2.1 Observational datasets

2.1.1 Description of MARGA

In the present study, we used the same dataset which was previously published by Acharja et al. (2020, 2021), which described the aerosol time series and chemistry measured with a Monitor for AeRosols and Gases in Ambient Air, model 2S instrument (MARGA). The MARGA system has two channels, one for sampling PM₁ and the other for sampling PM_2.5 for ground-based observations. The MARGA system (two sampling boxes, analytical box, and connected pumps) was located inside Indira Gandhi International Airport (IGIA), Delhi (28.56^∘ N, 77.09^∘ E), with the inlet PM₁ and PM_2.5 impactors fixed on the terrace with 2 m long inlet lines sampling outdoor air at 8 m above ground and 2 m above the rooftop. Measurements covered a winter period (19 December 2017 to 21 January 2018) with frequent moderate to dense fog events. Following intake through the PM₁ and PM_2.5 impactors, the air was passed through two parallel inlet tubes 2 m long and 14 mm inner diameter polytetrafluoroethylene (PTFE) to the PM₁ and PM_2.5 sampling channels of the MARGA. The air flow rate in each MARGA sampling box is regulated to a volumetric flow of 1 m³ h⁻¹. The measurements are close to being real-time, as two sets of syringes are employed to collect the samples in which a set of syringes collects the sample, and another set sends the collected samples from the previous hour for analysis. Each MARGA sampling system consists of a steam jet aerosol collector (SJAC) and a wet rotating denuder (WRD) for collecting and measuring water-soluble inorganic particulate species and gases in the ambient air. The continuous coating of the WRD by a thin film of absorption solution (10 ppm hydrogen peroxide (H₂O₂)) allows for the diffusion of gases into the absorption solution. By contrast, the low diffusion velocity of submicron particles restricts the ability of water-soluble aerosols to diffuse into the absorption solution. The absorption solution is continually changed to replace that abstracted for ion chromatography (IC) analysis of the dissolved gases. The air stream, depleted of gases by the WRD, subsequently enters the SJAC, where the steam enhances water-soluble aerosols to grow, allowing for their mechanical capture in a cyclone. The aqueous solutions deriving from two cyclones (for PM₁ and PM_2.5, respectively) are then supplied to the IC for chemical analysis (Acharja et al., 2020).

Ambient surface concentrations of NH₃ along with other trace gases (HCl, nitrous acid (HONO), nitric acid (HNO₃), and sulfur dioxide (SO₂)) and water-soluble inorganic components of PM₁ and PM_2.5 (Cl⁻, nitrate (NO${}_{\mathrm{3}}^{-}$), SO${}_{\mathrm{4}}^{\mathrm{2}-}$, NH${}_{\mathrm{4}}^{+}$, sodium (Na⁺), potassium (K⁺), magnesium (Mg²⁺), and calcium (Ca²⁺) were then quantified online by anion and cation chromatography in the analytical box at an hourly resolution. We have only used PM_2.5 inorganic water-soluble components and the gaseous measurements (available from both the PM₁ and PM_2.5 MARGA collection systems). Since NH${}_{\mathrm{4}}^{+}$ with the three major anions, Cl⁻, NO${}_{\mathrm{3}}^{-}$, and SO${}_{\mathrm{4}}^{\mathrm{2}-}$, constituted 97.3 % of the total measured ions in PM_2.5 (Acharja et al., 2020), we consider these four significant ions in our present study. In contrast, the remaining ionic species (i.e., Na⁺, K⁺, Mg²⁺, and Ca²⁺) contributed only about 3 % of the total measured ions and were neglected as it would not impact our present study significantly (Acharja et al., 2020). Anions are separated in a Metrosep A Supp 10 ($\mathrm{75}/\mathrm{4.0}$) column with sodium carbonate (Na₂CO₃) and sodium bicarbonate (NaHCO₃) ($\mathrm{7}/\mathrm{8}$ mmol L⁻¹) eluent, whereas for cation separation, a Metrosep C4 ($\mathrm{100}/\mathrm{4.0}$) cation column with 3.2 mmol L⁻¹ HNO₃ eluent was used (Acharja et al., 2020). To suppress the eluent background conductivity of anion chromatographs, three ion exchange units were used to ensure that the ion exchange unit is regenerated in each analysis. For this purpose, 1 M phosphoric acid (H₃PO₄) was used. This was performed to improve the signal-to-noise (S $/$ N) of the anion chromatographs. Details of the MARGA instrument can be found in Makkonen et al. (2012), Thomas et al. (2009), and Twigg et al. (2015).

2.1.2 Quality assurance and quality control (QA/QC) of MARGA

To ensure the observation's accuracy and check the data's quality, we have followed best practices during the study. The eluents, absorption, and regenerant solutions were prepared with minimum manual intervention. The operational parameters like anion and cation conductivity, SJAC heater temperature, column oven temperature, and airflow were regularly monitored to keep them within the safe limit. In addition to these, before injection of each sample into the anion and cation IC columns, the lithium bromide (LiBr) internal standard solution containing 320 µg L⁻¹ lithium (Li⁺) and 3680 µg L⁻¹ bromide (Br⁻) was mixed with each sample to provide calibration of each analysis. This ensures that each analysis is calibrated, and the concentration of gaseous and ionic samples is measured accurately. The PM₁ and PM_2.5 impactors were typically cleaned fortnightly to remove any material that may have stuck on the surface and inlets of the impactors. The lower detection limits (LODs) of the species monitored by MARGA were mentioned in Acharja et al. (2021). It shows that concentrations of species like Cl⁻, NO${}_{\mathrm{3}}^{-}$, SO${}_{\mathrm{4}}^{\mathrm{2}-}$, NH${}_{\mathrm{4}}^{+}$, SO₂, and NH₃ were always higher than LODs during the winter period. Concentrations of species like Na⁺, K⁺, Ca²⁺, Mg²⁺, HCl, HONO, and HNO₃ were sometimes below LODs, but the fraction of them was less than ∼ 10 % of the total observation period. We have omitted these values and treated them as not available (NA). As the fraction of observational hours is small, and these species contribute much less to the PM₁ and PM_2.5 mass concentrations, we believe values that are below the LODs would not significantly deviate our results. The quality of the data obtained was then checked using the ion-balance method. As an additional quality check, the ratio of the sum of cations to anions (nano equivalent per cubic meter – neq m⁻³) was used as an indicator for the viable data. We have checked the cation-to-anion ratio of each hourly sample expressed in units of neq m⁻³. We only accepted those values near to unity and rejected those not within the 10 % error bar limit. Based on this evaluation method, overall, for the campaign, the ratio was near unity (1.06 for PM₁ and 0.96 for PM_2.5). Excellent charge balance between anions and cations measured by the system also confirms that there are no significant contamination issues associated with the aerosol measurements. Values in slight excess of unity may indicate the presence of formate and acetate in the aerosol, which MARGA does not measure. Further details on the quality control of MARGA can be found in Acharja et al. (2020).

2.1.3 Other ground-based measurements

Hourly NO_x measurements were made by the chemiluminescence method, and hourly ozone (O₃) measurements were made by the UV photometric method (CPCB, 2011) at the nearest air quality monitoring station (AQMS) of IGIA, operated by the Central Pollution Control Board (CPCB). CPCB follows the United States Environmental Protection Agency (USEPA)-approved AC32M NO_x and 42M O₃ analyzer manufactured by Environment S.A. India Private Limited. We used 1 h monitored NO_x and O₃ values in our study. These air quality monitoring stations' quality control and assurance processes were followed as outlined in CPCB (2014, 2020). For data quality of CPCB, we omitted all those observed values which fell below the LOD of the instrument (2 µg m⁻³ for NO_x and 4 µg m⁻³ for O₃) (Technical specifications for CAAQM station, 2019) and above 500 µg m⁻³ for NO_x and 140 µg m⁻³ for O₃ and treated them as NA at a given site. For the NO_x and O₃ datasets, only a small fraction of data (2 %) were outside the instrument operating ranges specified. This step aims to remove any short-term local influence that the models cannot capture and retain the regional-scale variability because the nearest sites are located in the urban environment. We removed a single spike represented by a change of more than 100 µg m⁻³ in just 1 h for all the data in CPCB monitoring stations to filter out random fluctuations in the observations. We removed some very high NO_x and O₃ values that appeared in the time series right after measurement gaps. Meteorological parameters, including air temperature (T), relative humidity (RH), wind speed, and wind direction, were measured with the automatic weather station (AWS) platform on a 20 m flux tower (Ghude et al., 2017). For detailed information on the measurement site and its meteorological parameters, refer to Ali et al. (2019).

2.2 WRF-Chem v 3.9.1 model

The Weather Research and Forecasting model coupled with chemistry (WRF-Chem v3.9.1) was employed in this study to simulate atmospheric gases and aerosols over Delhi during the peak winter period, starting from 19 December 2017 to 21 January 2018. We recently used a similar model configuration to simulate the air quality over Delhi (Ghude et al., 2020; Kulkarni et al., 2020). This study used the Model for Ozone And Related chemical Tracers (MOZART-4) gas-phase chemical mechanism coupled with the Model for Simulating Aerosol Interactions and Chemistry (MOSAIC) aerosol scheme, which simulates SO${}_{\mathrm{4}}^{\mathrm{2}-}$, NH${}_{\mathrm{4}}^{+}$, NO${}_{\mathrm{3}}^{-}$, methanesulfonate, Na⁺, Ca²⁺, Cl⁻, carbonate, black carbon, and primary organic mass. Other inert minerals, trace elements, and inorganic species are lumped together as different inorganic masses. MOSAIC allows for gas-to-particle formation, which includes NH₃, HCl, sulfuric acid (H₂SO₄), HNO₃, and methane sulfonic acid (MSA) and also includes secondary organic aerosols (SOAs). Aerosol size distributions are represented by a sectional aerosol bin approach with four size bins (Georgiou et al., 2018). MOSAIC incorporates the thermodynamic and gas–particle partitioning module described by Zaveri et al. (2008). To reduce the computational cost, we selected a four-bin MOSAIC mechanism that simulates thermodynamic equilibrium and other aerosol processes such as condensation, coagulation, and nucleation. The same mechanism has been widely used with WRF-Chem for simulations outside India (Bucaram and Bowman, 2021; Sha et al., 2019; Yang et al., 2018), but only a limited number of studies have applied it to the Indian domain to include more detailed chemistry and species (Gupta and Mohan, 2015; Jena et al., 2020; Kumar et al., 2018). The SOA formation in MOSAIC is simulated using the volatility basis set approach (Knote et al., 2015). For consistency with the PM_2.5 MARGA measurements, we have chosen three bins according to simulated aerosol size (0.04–0.156, 0.156–0.625, and 0.625–2.5 µm) in accordance with the WRF-Chem aerosol size distribution.

The model domain covers the entire northern region of India, but here model simulations are compared with the observations at IGIA, Delhi (28.56^∘ N, 77.09^∘ E). The domain was set with a horizontal grid spacing of 10 km in both the latitudinal and longitudinal directions. The model top vertical grid included 47 vertical levels, with the model top set to 10 hPa. The physical parameterization schemes of model configuration are the same as those described by Ghude et al. (2020) and Jena et al. (2021). EDGAR-HTAP (Emission Database for Global Atmospheric Research for Hemispheric Transport of Air Pollution) for the year 2010 at $\mathrm{0.1}{}^{\circ }\times \mathrm{0.1}{}^{\circ }$ grid resolution was used in this study for anthropogenic emissions of aerosols and trace gases (PM_2.5, PM₁₀, organic carbon (OC), black carbon (BC), CO, and NO_x, etc.) and is scaled to 2018 as per Jena et al. (2021). Biogenic emissions are calculated online using the Model of Emissions of Gases and Aerosols from Nature version 2.1 (MEGAN2.1) (Guenther et al., 2006), and dust emissions are based on the traditional Goddard Global Ozone Chemistry Aerosol Radiation and Transport (GOCART) dust scheme that works with MOSAIC (Ginoux et al., 2001). Fire INventory from NCAR (FINNv1.5, 2017) was used in this study for daily open biomass burning emissions that are vertically distributed within the model using Freitas et al. (2007). The chemical initial and lateral boundary conditions come from the global model simulations from the Model for Ozone and Related Chemical Tracers (MOZART-4), and the meteorological initial and lateral boundary conditions are provided from the fifth-generation European Centre for Medium-Range Weather Forecasts (ECMWF) atmospheric reanalysis of the global climate (ERA5) with 6-hourly temporal resolution. The simulations were reinitialized every fifth day to limit the growth of meteorological errors in our simulations, but the chemical fields were carried forward from the previous simulation.

3 Results and discussion

3.1 Comparison of temporal variation in NH₃, NH${}_{\mathrm{4}}^{+}$, and total NH_x using WRF-Chem and MARGA

3.1.1 Diurnal variation

To investigate how well a state-of-the-art chemical transport model performs in capturing the diurnal behavior of NH₃ and NH${}_{\mathrm{4}}^{+}$, we compared observed and model-simulated diurnal profiles of NH₃ and NH${}_{\mathrm{4}}^{+}$. Figure 1 displays the comparison of diurnal variation (00:00 to 23:00 Indian standard time (IST)) in meteorological parameters (T and RH) at the IGIA site in Delhi (Fig. 1a) along with NH₃ and NH${}_{\mathrm{4}}^{+}$ averaged over the study period (Fig. 1b) between observations and the model. We adopted diurnal variation in emissions from a recent study by Jena et al. (2021). Note that diurnal variability in the model simulations is primarily controlled by the planetary boundary layer mixing. We first investigated the ability of WRF-Chem to accurately predict the meteorological parameters of RH and T, which are important determinants of the gas-to-aerosol partitioning of (semi-)volatile compounds. As shown in Fig. 1a, simulated T and RH are in reasonable agreement with the observations, with the simulated RH values falling in the range of 50 %–90 %. Overall, it can be seen that the model shows cold and wet bias compared to the observations but shows warm bias (about 2–3 ^∘C) and dry bias (about 10 %–12 %) in the afternoon hours. In spite of the small change in the amplitude of the diurnal cycle of RH, the phase characteristics of the diurnal cycle of both T and RH are reasonably well captured by the model. Figure 1b shows that simulated NH₃ and NH${}_{\mathrm{4}}^{+}$ are very different compared with the MARGA measurements. The model predicts an average NH₃ and NH${}_{\mathrm{4}}^{+}$ ± 1σ mass loading of 56.7 ± 14.3 and 14.7 ± 4.9 µg m⁻³, respectively, while MARGA measurements indicate an average NH₃ and NH${}_{\mathrm{4}}^{+}$ ± 1σ mass loading of 28.2 ± 12.4 and 36.9 ± 15.1 µg m⁻³, respectively. We find the diurnal variation of gas-phase NH₃ is significantly overestimated by the model (normalized mean bias (NMB) = 1.02). Conversely, NH${}_{\mathrm{4}}^{+}$ is underestimated by about 60 % (NMB = −0.60). Simulated NH₃ concentrations peak between 07:00–09:00 and 22:00–23:00 with bimodal variation, whilst MARGA shows a single peak around 12:00–13:00. Conversely, a nearly flat diurnal profile of NH${}_{\mathrm{4}}^{+}$ is predicted by the model, whereas the average MARGA NH${}_{\mathrm{4}}^{+}$ concentration maxima and minima were observed during nighttime (16:00–03:00) and daytime (03:00–08:00 and 09:00–16:00), respectively.

Figure 1 Comparison of observed and simulated average diurnal variation in (a) meteorological parameters such as temperature (T in ^∘C) and relative humidity (RH in %) and (b) NH₃ and NH${}_{\mathrm{4}}^{+}$ concentration (µg m⁻³) during the sampling period (bar indicates mean standard deviation of each hour).

We also looked into the average diurnal profile of NO_x and NH₃ during dense fog events, and the details can be found in the Supplement (Figs. S1 and S2). It is evident that the observed daytime peak of NH₃ did not coincide with NO_x peaks, suggesting that traffic emissions do not contribute significantly to the observed NH₃ rise. The observed correlation between fog water and enhanced NH₃ pulses is consistent with what would also be expected from the evaporation of dew (Sutton et al., 1998; Wentworth et al., 2014, 2016) (Fig. S2 in the Supplement) but is not sufficient to identify whether it is the main cause of the daytime increase of NH₃. In the future, measurements of the dew water NH${}_{\mathrm{4}}^{+}$ and the accumulation of dew water would be ideal for illuminating the contributing processes. The daytime increase in NH₃ concentration could be associated with NH${}_{\mathrm{4}}^{+}$ aerosol volatilization driven by an associated sharp change in T and RH (∼ 11:00–12:00) (Sutton et al., 2009a, 2013) off-ground surfaces. The fastest increase in T is 12:00, which is indeed when NH₃ was at maximum concentration, indicating that gas-to-particle partitioning may impact the diurnal behavior of NH₃ at Delhi during winter (Sutton et al., 2009a, b). However, in the model, because the largest increase in simulated NH₃ also precedes the large changes in simulated meteorological parameters, and because the simulated particulate NH${}_{\mathrm{4}}^{+}$ is flat compared to observations, simulated meteorology is ruled out as a significant contribution to high bias in simulated NH₃. Also, the current model does not include the bidirectional exchange of NH₃ with surfaces such as dew and fog water.

3.1.2 Daily mean variation

To assess the validity of the model, the ratio between simulated and observed (model $/$ obs) was tested. Figure 2 displays the model $/$ obs ratio of daily mean variations in the NH₃, NH${}_{\mathrm{4}}^{+}$, and total NH_x concentrations. The model shows large differences in NH₃ and NH${}_{\mathrm{4}}^{+}$ compared with observations. We find a model $/$ obs higher than 1 (1.5–4.5) in simulated NH₃, indicating the model is biased high (NMB = 1.02), while there is a poor agreement for NH${}_{\mathrm{4}}^{+}$ (model $/$ obs less than 0.5), indicating the model is biased low (NMB = −0.62). There is good agreement between the modeled total NH_x, which is mostly consistent with the observation (model $/$ obs close to 1) with a small bias (NMB = 0.08). Despite the adequate ability of the model to reproduce the accurate total NH_x, the model is biased low for NH${}_{\mathrm{4}}^{+}$ and high for NH₃, indicating that the model's representation of the gas-to-particle partitioning is not correct. It is, therefore, necessary to understand missing chemical processes in gas-to-particle partitioning responsible for the overestimation of NH₃ and underestimation of NH${}_{\mathrm{4}}^{+}$ in the model.

Figure 2 Ratio of model $/$ obs of the daily mean NH₃, NH${}_{\mathrm{4}}^{+}$, and total NH_x concentration.

Figure 3 Share of major components of gases and particulate matter (PM_2.5) based on the mean concentrations during WiFEX (share according to µeq m⁻³).

3.2 Gas-to-particle partitioning

We investigated the ability of the model to accurately describe the gas-to-particle partitioning of the measurements (MARGA) by evaluating the fraction of total NH_x in the particulate phase (NH${}_{\mathrm{4}}^{+}$ $/$ NH_x) (Ellis et al., 2011; Wang et al., 2015) for which statistical values are summarized in Table 1. The correlation coefficient (r) indicates an inverse relationship of NH${}_{\mathrm{4}}^{+}$ $/$ NH_x with NH₃ for both MARGA and the model ($r=-\mathrm{0.57}$, −0.58, respectively). A strong correlation of the MARGA ratio NH${}_{\mathrm{4}}^{+}$ $/$ NH_x with the dominant anion concentration (Cl⁻: r=0.79) was observed. However, the measurement shows a poor relationship between SO${}_{\mathrm{4}}^{\mathrm{2}-}$ and NH${}_{\mathrm{4}}^{+}$ $/$ NH_x followed by NO${}_{\mathrm{3}}^{-}$, which is probably due to very low concentrations that do not change NH${}_{\mathrm{4}}^{+}$ $/$ NH_x significantly, even when SO${}_{\mathrm{4}}^{\mathrm{2}-}$ and NO${}_{\mathrm{3}}^{-}$ are neutralized (see Fig. 6). By contrast, the model shows a strong correlation between NH${}_{\mathrm{4}}^{+}$ $/$ NH_x with SO${}_{\mathrm{4}}^{\mathrm{2}-}$ concentration (r=0.77). MARGA indicates high particulate fractions of NH${}_{\mathrm{4}}^{+}$ and Cl⁻, while the modeled composition is dominated by NH${}_{\mathrm{4}}^{+}$ and SO${}_{\mathrm{4}}^{\mathrm{2}-}$. This mismatch is due to the complete absence of Cl⁻ chemistry in the standard model. The measured NH${}_{\mathrm{4}}^{+}$ $/$ NH_x suggests that anthropogenic HCl may be promoting this increase in particle fraction of NH${}_{\mathrm{4}}^{+}$ and Cl⁻ via partitioning into the aerosol, deprotonating in the aerosol water, followed by NH₃ partitioning and being protonated by the ionization of the strong electrolyte HCl (Chen et al., 2022; Gunthe et al., 2021).

Table 1 Performance statistics of correlation coefficient (r) of NH${}_{\mathrm{4}}^{+}$ $/$ NH_x with NH₃ and aerosols (NH${}_{\mathrm{4}}^{+}$, Cl⁻, SO${}_{\mathrm{4}}^{\mathrm{2}-}$, and NO${}_{\mathrm{3}}^{-}$).

Figure 3 shows the percentage contribution of gases (NH₃, SO₂, HCl, HNO_3, and HONO) and PM_2.5 aerosol (NH${}_{\mathrm{4}}^{+}$, SO${}_{\mathrm{4}}^{\mathrm{2}-}$, NO${}_{\mathrm{3}}^{-}$, and Cl⁻) during the WiFEX measurements. The pie charts for the gases show that NH₃ (accounting for 53.3 % of the measured total gas concentration) dominates the gas phase, followed by sulfur dioxide (SO₂) (35.61 %), whereas PM_2.5 aerosol shows NH${}_{\mathrm{4}}^{+}$ (49.5 %) as a major cation and Cl⁻ (29.7 %) as a significant anion followed by NO${}_{\mathrm{3}}^{-}$ (11.7 %) and SO${}_{\mathrm{4}}^{\mathrm{2}-}$ (9 %). There is also a very high amount of SO₂ reaching the site from the nearby industrial area, which is not converted to SO${}_{\mathrm{4}}^{\mathrm{2}-}$ very quickly (Acharja et al., 2021). In a normally NH₃-rich atmosphere, gas-phase oxidation of SO₂ is much slower than the aqueous-phase oxidation by O₃, and due to nearby sources, much of the sulfur is present as SO₂ (Li et al., 2007). This appears to be because of the slow rate of gas-phase oxidation of SO₂. Although the atmosphere is rich in NH₃, in principle favoring aqueous-phase oxidation via O₃, it appears that O₃ concentrations are often insufficient (mean = 36.3, median = 33.8, minimum = 26.5, and maximum = 53.9 µg m⁻³ respectively) at the IGIA site (Fig. S3 in the Supplement). Hence for many periods during the WiFEX campaign, SO${}_{\mathrm{4}}^{\mathrm{2}}$ and NO${}_{\mathrm{3}}^{-}$ are very low, with the result that the NH${}_{\mathrm{4}}^{+}$ $/$ NH_x ratio does not change appreciably when SO${}_{\mathrm{4}}^{\mathrm{2}-}$ is neutralized (Table 1).

According to thermodynamic equilibrium theory, an aqueous solution maintains charge neutralization initially by balancing NH₃ uptake with the uptake of sulfuric acid (H₂SO₄) before HNO₃ and HCl can partition into the aqueous aerosol; hence all SO${}_{\mathrm{4}}^{\mathrm{2}-}$ in the condensed phase will be fully neutralized before any HNO₃ or HCl can partition (Behera et al., 2013). Typical Delhi winter conditions of excess NH₃, high RH, and low T favor gas-to-particle partitioning of NH₃. The principal inorganic chemical reactions that occur in aqueous atmospheric aerosols form pairs of non-volatile NH${}_{\mathrm{4}}^{+}$ and acid anions (SO${}_{\mathrm{4}}^{\mathrm{2}-}$, NO${}_{\mathrm{3}}^{-}$, and Cl⁻) and are summarized in Reactions (R1) to (R3) (Seinfeld et al., 1998).

$$\begin{array}{}\text{(R1)}& {\displaystyle }\mathrm{2}{\mathrm{NH}}_{{\mathrm{3}}_{\left(\mathrm{g}\right)}}+{\mathrm{H}}_{\mathrm{2}}{\mathrm{SO}}_{{\mathrm{4}}_{\left(\mathrm{g}\right)}}\rightleftarrows {\mathrm{NH}}_{\mathrm{4}}^{+}+{\mathrm{SO}}_{\mathrm{4}}^{\mathrm{2}-}\text{(R2)}& {\displaystyle }{\mathrm{NH}}_{{\mathrm{3}}_{\left(\mathrm{g}\right)}}+{\mathrm{HNO}}_{{\mathrm{3}}_{\left(\mathrm{g}\right)}}\rightleftarrows {\mathrm{NH}}_{\mathrm{4}}^{+}+{\mathrm{NO}}_{\mathrm{3}}^{-}\text{(R3)}& {\displaystyle }{\mathrm{NH}}_{\mathrm{3}\left(\mathrm{g}\right)}+{\mathrm{HCl}}_{\left(\mathrm{g}\right)}\rightleftarrows {\mathrm{NH}}_{\mathrm{4}}^{+}+{\mathrm{Cl}}^{-}\end{array}$$

NH${}_{\mathrm{4}}^{+}$ and Cl⁻ (Reaction R3), which are favored by low T and high RH, form a reversible equilibrium with NH₃ and HCl (Ianniello et al., 2011; Seinfeld and Pandis, 2016), which was the case during WiFEX. It is likely that high Cl⁻ in Delhi resulted from gas–to-particle partitioning of HCl into aerosol water in the presence of excess NH₃ (Reaction R3), with aqueous-phase Cl⁻ stimulating further water uptake and jointly driving aerosol mass composition and growth through co-condensation (Chen et al., 2022; Gunthe et al., 2021). Hence, to understand the driver of the measured NH${}_{\mathrm{4}}^{+}$ and the role of aqueous chemistry, we plotted the fraction of the ratio of HCl to Cl⁻ (HCl $/$ Cl⁻) as a function of NH${}_{\mathrm{4}}^{+}$ concentration and RH in Fig. 4. The decrease in the fraction of HCl $/$ Cl⁻ is associated with an increase in NH${}_{\mathrm{4}}^{+}$ concentration at high RH between 70 %–100 %. The HCl $/$ Cl⁻ is highly anticorrelated (r = −0.53) with NH${}_{\mathrm{4}}^{+}$ concentration in the presence of high RH (70 %–100 %), further supporting the view that HCl promotes the increase in the particle fraction of NH${}_{\mathrm{4}}^{+}$ (49.5 %), with Cl⁻ (29.7 %) the primary anion.

We investigated the directions of local emission sources associated with concentration increases of NH₃, NH${}_{\mathrm{4}}^{+}$, Cl⁻, and NH_x through bivariate polar graphs using the OpenAir software (Carslaw and Ropkins, 2012) at the IGIA site. Figure 5 shows the bivariate polar plots of mean NH₃ (Fig. 5a), NH${}_{\mathrm{4}}^{+}$ (Fig. 5b), Cl⁻ (Fig. 5c), and total NH_x (Fig. 5d) concentration for the observation period in relation to wind speed and wind direction. The 270–300^∘ sector dominated the wind direction at IGIA (Acharja et al., 2021). Figure 5a shows that the highest NH₃ concentration was associated with the winds coming from the east and southeast of the site, where it could have been emitted from dairy farms, including animal houses, yards, and manure storage, as well as by the application to the farmland of urea and other ammoniacal fertilizers, ammoniacal wastes, and ruminant urine located at this region (Hindustan Times, 2021; Leytem et al., 2018; Sherlock et al., 1994). Such sources of NH₃ volatilization (Hristov et al., 2011; Laubach et al., 2013) can also explain the higher concentrations of total NH${}_{\mathrm{4}}^{+}$ (and, by definition, NH_x) for air coming from the southeast of the measurement site (Fig. 5b and d). This enhancement in the southeast region is not only affected by emissions, but also by meteorology and chemistry. Thus higher NH₃ concentration may also be due to the lack of turbulent mixing, which restricts the dilution of plumes from local point sources at lower wind speeds (Ianniello et al., 2010). The bivariate polar plots of NH${}_{\mathrm{4}}^{+}$ (Fig. 5b) and Cl⁻ (Fig. 5c) concentration point to the west direction as a principal source for thermodynamic partitioning of NH₃ and HCl to the condensed phase to form NH${}_{\mathrm{4}}^{+}$ and Cl⁻. Two industrial sources are located in this direction: the site is impacted by a cluster in northwest Delhi of industrial processes, such as steel pickling industries, and others include metal finishing and electroplating, which are known to be vital HCl emitters (Acharja et al., 2021; Jaiprakash et al., 2017). Near the source, abundant quantities of NH₃ may drive the partitioning of HCl to the condensed phase, resulting in high concentrations of NH${}_{\mathrm{4}}^{+}$ and Cl⁻ towards the west at lower wind speeds. Thus, high NH${}_{\mathrm{4}}^{+}$ and Cl⁻ correspond to the lowest NH₃ concentration region (inverse relation), which can be observed in Fig. 5a, b, and c, highlighting the importance of nearby HCl industrial sources in driving the particle fraction of NH${}_{\mathrm{4}}^{+}$ and Cl⁻.

Figure 4 Fraction of HCl $/$ Cl⁻ ratio as a function of NH${}_{\mathrm{4}}^{+}$ concentration (µgm⁻³) and relative humidity (RH).

Figure 5 Bivariate plots of mean (a) NH₃ concentration, (b) NH${}_{\mathrm{4}}^{+}$ concentration, (c) Cl⁻ concentration, and (d) total NH_x concentration in relation to wind speed (m s⁻¹) and direction.

To gain insight into the role of NH${}_{\mathrm{4}}^{+}$ in the neutralization of anions (SO${}_{\mathrm{4}}^{\mathrm{2}-}$, NO${}_{\mathrm{3}}^{-}$, and Cl⁻), the aerosol neutralization ratio (ANR) was calculated using the observed data. The ANR is defined as the equivalent ratio of NH${}_{\mathrm{4}}^{+}$ to the sum of SO${}_{\mathrm{4}}^{\mathrm{2}-}$, NO${}_{\mathrm{3}}^{-}$, and Cl⁻ because these species represent the dominant cations and anions in PM_2.5, respectively. Figure 6 demonstrates, on average, how well the charge balance works between Cl⁻, NO${}_{\mathrm{3}}^{-}$, and SO${}_{\mathrm{4}}^{\mathrm{2}-}$ (in µeq m⁻³) as the anions and NH${}_{\mathrm{4}}^{+}$ as the major cation (ANR close to unity), with Cl⁻ as the most significant anion followed by NO${}_{\mathrm{3}}^{-}$ and SO${}_{\mathrm{4}}^{\mathrm{2}-}$. The mean ± 1σ ANR value for PM_2.5 during the observed period was 0.96 ± 0.14. It ranges from a minimum of 0.35 ± 0.04 to a maximum of 2.31 ± 0.08. Higher values than unity may indicate the presence of organic acids in the aerosol, which MARGA does not measure (Acharja et al., 2020). Also, high standard error in Fig. 6 indicates the possibility of uncertainties associated with the breakthrough of NH₃ spikes on the denuder at high concentration (∼ 1 %) (Stieger et al., 2019). However, the good charge balance indicates this was not a major issue. There also were certain periods where low concentrations were observed of Cl⁻ and NO${}_{\mathrm{3}}^{-}$ (3–6 January 2018 and 16–17 January 2018) in Fig.6. Comparing the model $/$ obs for NH₃, NH${}_{\mathrm{4}}^{+}$, and total NH_x during these periods provides some degree of validation of the model where sulfur chemistry dominates the reaction with NH₃. Figure S4 (in the Supplement) shows that model $/$ obs indicates substantial variability, which appears to be overestimating NH₃ (model $/$ obs >1) while underestimating total NH${}_{\mathrm{4}}^{+}$ (model $/$ obs <1) on average in the model.

Figure 6 Neutralizing effect between Cl⁻, NO${}_{\mathrm{3}}^{-}$, and SO${}_{\mathrm{4}}^{\mathrm{2}-}$ as the anions (µeq m⁻³) and aerosol neutralization ratio (ANR), where ANR >1 indicates over-neutralized (alkaline), and ANR <1 indicates under-neutralized (acid) (orange bar indicates daily mean standard error).

3.3 Influence of HCl $/$ Cl⁻ chemistry in WRF-Chem

We further conducted three scenario simulations for the period 7–16 January 2018 (10 d) to explore the potential impacts of the addition of anthropogenic chloride (HCl $/$ Cl⁻) emissions in the concentrations of NH₃, NH${}_{\mathrm{4}}^{+}$, and total NH_x. We employ the HCl emissions from trash-burning activities in Delhi, as predicted by Sharma et al. (2019) in our model setup. We tested the three sensitivity experiments, named no HCl (0 mol km⁻² h⁻¹), base case HCl (3× Sharma et al., 2019; 24.8 mol km⁻² h⁻¹), and 3 × base HCl (74 mol km⁻² h⁻¹) scenarios, reflecting adjustments which are consistent with the more recent upward adjustments in the amount of waste burned in landfills by Chaudhary et al. (2021) and also reflect additional industrial HCl sources not accounted for in the inventory. Figure 7 presents the box-and-whisker plots for secondary inorganic aerosols and trace gases from the observations (MARGA) and those simulated by the model for the three sensitivity experiments. Daily mean ± 1σ values are summarized in Table 2 for three different model scenarios. As can be observed from Fig. 7a–c, increasing the HCl emissions (Fig. 7g) in the model partitions more NH₃ to the condensed phase due to its high concentrations, reaching maximum mass loadings of NH${}_{\mathrm{4}}^{+}$ and Cl⁻ of 70 and 110 µg m⁻³, respectively, in the 3 × base HCl scenario while increasing the total mean NH_x concentration by 15 µg m⁻³ compared to the no HCl run, presumably reflecting the longer residence time of NH${}_{\mathrm{4}}^{+}$ for near-surface air measurements.

Figure 7 Box-and-whisker plot for trace gases and secondary inorganic aerosols from the observations (MARGA) and simulated in sensitivity test with changes in HCl emissions in no HCl (0 mol km⁻² h⁻¹), base case HCl (24.8 mol km⁻² h⁻¹), and 3 × base HCl (74 mol km⁻² h⁻¹) runs at IGIA, Delhi.

Table 2 Daily mean ± 1σ in gases and inorganic aerosol concentration observed (MARGA) and simulated in sensitivity test with changes in total HCl emissions in no HCl (0 mol km⁻² h⁻¹), base case HCl (24.8 mol km⁻² h⁻¹), and 3 × base HCl (74 mol km⁻² h⁻¹).

The simulated NO${}_{\mathrm{3}}^{-}$ concentration (Fig. 7e) generally exceeds the measurements in all three experiments; since the main neutralizing species for NO${}_{\mathrm{3}}^{-}$ is NH${}_{\mathrm{4}}^{+}$, it is controlled via the equilibrium between NO${}_{\mathrm{3}}^{-}$, HNO₃, and NH₃, but also the competition with HCl for free NH₃. Simulated HNO₃ is significantly underestimated (by ∼ 3 µg m⁻³) (Fig. 7h) by the model compared to the observations. As a consequence, the model suggests that NO${}_{\mathrm{3}}^{-}$ formation from gaseous NH₃ and HNO₃ cannot occur. The gas fraction of observed HNO₃ will be determined by aerosol pH and liquid water content based on NH₃ and NO${}_{\mathrm{3}}^{-}$ availability (Nenes et al., 2020). The over-prediction of NH₃ concentration in the model compared with the observations generates more NO${}_{\mathrm{3}}^{-}$ (and simultaneously reduces HNO₃), with the total fraction of HNO₃+ NO${}_{\mathrm{3}}^{-}$ (THNO₃) concentration in the model also exceeding the observed THNO₃, which is more strongly affected by reducing the NH₃ emissions in the model (Fig. S5 in the Supplement). On average, THNO₃ reduced by only 0.38 µg m⁻³ in 3 × base HCl compared to the no HCl run. But reducing NH₃ emissions by a factor of 3 (−3 × NH₃_EMI) in the 3 × base HCl scenario reduced mean THNO₃ by a further 4.71 µg m⁻³. The extent of partitioning and accumulation of NH₄NO₃ depends on T, aerosol water, and pH, as well as NH₃ availability (Nenes et al., 2020). Our model simulations find that the presence of HCl $/$ Cl⁻ does not significantly alter THNO₃ but that the excess NH₃ with missing chloride chemistry is a major contributor and will lead to mismatches in the model between measured simulated gas and particulate matter concentrations.

The simulated SO${}_{\mathrm{4}}^{\mathrm{2}-}$ concentration (Fig. 7f) was underestimated (by ∼ 7.5 µg m⁻³), while gas-phase SO₂ (Fig. 7i) was found to be overestimated by about 16 µg m⁻³ in all three experiments compared with the observations. This may be caused by the fact that the drivers for typical sulfate production via OH or aqueous H₂O₂ oxidation pathway are likely to be wrong in the model. The missing chemistry may underlie this mismatch and requires further sensitivity studies considering different SO₂ oxidation pathways. This requires further study, such as scenario evaluation of altered SO₂ emissions in the model, to examine the main pathway(s) for SO₂ to SO${}_{\mathrm{4}}^{\mathrm{2}-}$ conversion. Measurements of OH and other radicals in Delhi are currently lacking, making it difficult to constrain the associated chemical schemes. To investigate the further impact of 3 × base HCl in the model, uptake of gaseous NH₃ to form NH${}_{\mathrm{4}}^{+}$ and Cl⁻ was analyzed via a strong correlation coefficient values of r = 0.84 for NH${}_{\mathrm{4}}^{+}$ $/$ NH_x with Cl⁻ concentration, indicating that a fraction of gas-to-particle conversion in the model correlates well with the Cl⁻ concentration and was reasonably well simulated in the 3 × base HCl run.

Figure 8 Top: average diurnal cycles of NH₃ and NH${}_{\mathrm{4}}^{+}$ concentration (µg m⁻³) with mole equivalents of Cl⁻, NO${}_{\mathrm{3}}^{-}$, SO${}_{\mathrm{4}}^{\mathrm{2}-}$, NH${}_{\mathrm{4}}^{+}$, SO₂, HCl, and HNO₃ (µeq m⁻³) of (a) measured (MARGA) and (b) modeled (3 × base HCl run), along with meteorological parameters (bottom).

3.4 Comparison of the temporal variation in NH₃, NH${}_{\mathrm{4}}^{+}$, and NH_x using WRF-Chem (HCl $/$ Cl⁻) and MARGA

3.4.1 Diurnal variation

Here, diurnal variations of monitored aerosol compounds and gases were analyzed to investigate the gas-to-particle conversion of NH₃ in the model. We analyzed the simulation results of the 3 × base HCl run. The diurnal variations for NH₃ and NH${}_{\mathrm{4}}^{+}$ are controlled mainly by thermodynamic gas-to-particle partitioning, boundary layer mixing, and emission and deposition processes, along with vertical and horizontal advection (Meng et al., 2018). Figure 8 (top) presents the diurnal variations of NH₃ and NH${}_{\mathrm{4}}^{+}$ (in µg m⁻³), along with particulate NH${}_{\mathrm{4}}^{+}$, Cl⁻, NO${}_{\mathrm{3}}^{-}$, SO${}_{\mathrm{4}}^{\mathrm{2}}$, SO₂, HCl, and HNO₃ concentrations (in µeq m⁻³) measured (Fig. 8a (top)) and modeled (Fig. 8b (top)), along with meteorological parameters such as T and RH (Fig. 8 (bottom)). We adopted diurnal variation in emissions from Jena et al. (2021) based on boundary layer mixing. It can be seen in Fig. 8a (top and bottom) that a much bigger peak in NH₃ concentration is observed in the daytime than is modeled (despite turbulence differences), indeed suggesting a much stronger NH₃ in the middle of the day (11:00–13:00). As evaporation proceeds mainly in the morning (08:00–12:00), getting warmer, the peak is near midday (11:00–13:00), rather than in the afternoon (13:00–14:00), when warmest, similar to what was also observed in Sutton et al. (1998). Indeed, the decreasing NH${}_{\mathrm{4}}^{+}$ and Cl⁻ during the late morning (10:00) correspond to the increasing NH₃ peak, which reflects the fact that warming promotes the shift of aerosols to the gas phase. Ammonium decreases more than NH₃ during the day, as this also evaporates to form NH₃. Similarly, Cl⁻ evaporates during the day since the HCl concentration increases. However, it can be seen that NO${}_{\mathrm{3}}^{-}$ and SO${}_{\mathrm{4}}^{\mathrm{2}-}$ are slightly changed diurnally, inferring longer-range transport perhaps, whereas HCl and Cl⁻are from more local sources. The diurnal variability in gases and aerosols in 3 × base HCl simulations in Fig. 8b (top) is primarily controlled by the planetary boundary layer mixing, meteorology/dispersion, environment (T and RH in Fig. 8b (bottom)), and transport. So presumably, maximum NH₃ at 08:00 is due to limited turbulence/boundary layer, with dilution by mixing after 08:00. However, the model is able to represent the diurnal variation of NH${}_{\mathrm{4}}^{+}$ and Cl⁻ well, both in terms of amount and pattern, which was not the case in the no HCl run, where NH${}_{\mathrm{4}}^{+}$ was observed to be flat in Sect. 1. During the hours of 09:00 and 11:00, when measured NH₃ rises, the model predicts a large decrease in NH₃, while during 19:00–23:00, when measured NH₃ decreases, the model predicts a large increase. Furthermore, the modeled HCl and HNO₃ are very low compared to the measurements, whereas SO₂ concentration matches well with the observations. It can be seen that NO${}_{\mathrm{3}}^{-}$ and SO${}_{\mathrm{4}}^{\mathrm{2}-}$ are flat in the model. This highlights the need to develop accurate diurnal variability in NH₃ emissions over this region.

Figure 9 presents the differences in diurnal variation of mean NH₃ (Fig. 9a), NH${}_{\mathrm{4}}^{+}$ (Fig. 9b), and total NH_x (Fig. 9c) concentration for the three sensitivity experiments. While the simulated NH₃ concentrations decrease in the 3 × base HCl compared to the no HCl and base case HCl run (Table 2), none of the model experiments capture the diurnal cycle of NH₃. Higher levels of observed NH₃ during daytime and modeled NH₃ during nighttime highlight the need to improve diurnal variability in NH₃ emissions over this region based on the nature and strength of the actual sources. Between the no HCl and the 3 × base HCl run, the NMB for NH₃ reduced from 1.38 to 1.13, and NMB for NH${}_{\mathrm{4}}^{+}$ systematically improved from −0.61 to −0.03. In contrast, NMB for total NH_x increased from 0.12 to 0.39. Table 3 summarizes the statistical indicators for the three sensitivity experiments. An increase in HCl emissions in the 3 × base HCl leads to a higher mass concentration of NH${}_{\mathrm{4}}^{+}$ and Cl⁻, which also increases total mean NH_x concentration by 22.4 µg m⁻³, presumably reflecting the longer atmospheric lifetime of NH${}_{\mathrm{4}}^{+}$ compared with NH₃. We find consistent high bias in all the simulations for NH₃, which is highest during the early morning and at nighttime. In order to better understand the relationship between NH₃, NH${}_{\mathrm{4}}^{+}$, and NH_x concentrations in the diurnal profile of model, one sensitivity study is conducted in the best case HCl experiment to simulate the response of NH_x concentrations by changing NH₃ emissions. In these simulations, only NH₃ emissions were reduced further by a factor of 3 (−3 × NH₃_EMI) in the 3 × base HCl experiment, while all other processes and chemical schemes were unchanged. Figure S6 in the Supplement shows the diel profile of model $/$ obs ratio for NH₃ (Fig. S6a), NH${}_{\mathrm{4}}^{+}$ (Fig. S6b), and total NH_x (Fig. S6c) concentration simulated with the 3 × base HCl and −3 × NH₃_EMI scenario. Reducing NH₃ emissions in the model (−3 × NH₃_EMI) significantly improves model–measurement agreement for NH₃ (mean model $/$ obs = 1.9), NH${}_{\mathrm{4}}^{+}$ (mean model $/$ obs = 0.9), and total NH_x concentration(mean model $/$ obs = 1.2) compared to the 3 × base HCl run, further suggesting that the longer lifetime of NH${}_{\mathrm{4}}^{+}$ may be the controlling driver for the total NH_x concentration in the model.

Figure 9 Diurnal variation in the mean (a) NH₃ concentration (b) NH${}_{\mathrm{4}}^{+}$ concentration and (c) total NH_x concentration observed (black), simulated in no HCl (dotted red), base case HCl (dashed red), and 3 × base HCl (solid red) runs.

Table 3 Model performance statistics for NH₃, NH${}_{\mathrm{4}}^{+}$, and total NH_x concentration at IGIA, Delhi, from three sensitivity experiments in no HCl (0 mol km⁻² h⁻¹), base case HCl (24.8 mol km⁻² h⁻¹), and 3 × base HCl (74 mol km⁻² h⁻¹) runs and the MARGA.

3.4.2 Variation of daily means

Figure S7 in the Supplement illustrates a time-series graph that compares daily mean NH₃ (Fig. S7a), NH${}_{\mathrm{4}}^{+}$ (Fig. S7b), and total NH_x concentrations (Fig. S7c) for the three sensitivity experiments, and Table 2 shows the mean ± 1σ of these variables. The results show that compared to the no HCl run, NH₃ mean concentrations decreased by 2 µg m⁻³ in the base case HCl and decreased by a further 3.2 µg m⁻³ in the 3 × base HCl run. Conversely, NH${}_{\mathrm{4}}^{+}$ mean concentration increases in the base case HCl by 7.5 µg m⁻³ and further increases by 13.1 µg m⁻³ (3 × base HCl). This decrease in NH₃ is associated with the enhanced gas-to-particle conversion of NH₃ to NH${}_{\mathrm{4}}^{+}$. Associated with these changes, total mean NH_x also increased by 5.5 and 9.8 µg m⁻³ in the base case HCl and 3 × base HCl, respectively, compared to the no HCl. This is likely due to associated increases in the atmospheric lifetime of NH_x with respect to deposition as the partitioning shifted from the faster depositing gas phase to the aerosol phase. The lifetime of NH₃ is very short, a few hours, while that of NH${}_{\mathrm{4}}^{+}$ is 1 to 15 d (Aneja et al., 1998; Nair and Yu, 2020; Pawar et al., 2021; Q. Wang et al., 2020).

Figure 10 Comparison of ratio of model $/$ obs in the daily mean (a) NH₃ concentration, (b) NH${}_{\mathrm{4}}^{+}$ concentration, and (c) total NH_x concentration in the 3 × base HCl and -3 × NH₃_EMI scenarios.

To understand the overestimation of total NH_x in the daily mean variation by the model further, we compared the 3 × base HCl and −3 × NH₃_EMI sensitivity experiment. Figure 10 shows the ratio of model $/$ obs for NH₃ (Fig. 10a), NH${}_{\mathrm{4}}^{+}$ (Fig. 10b), and total NH_x (Fig. 10c) concentration. It can be seen that the model–measurement agreement improves significantly (model $/$ obs closer to 1) after reducing NH₃ emissions for all three metrics. The −3 × NH₃_EMI run would reduce the mean NH₃, NH${}_{\mathrm{4}}^{+}$, and total NH_x concentration by ∼ 8.1, 3.2, and 11.3 µg m⁻³, respectively, compared to the 3 × base HCl run. Even though NH₃ emissions are reduced, they are still sufficient to react rapidly with the varying HCl in the sensitivity experiments, contributing to an increase in NH${}_{\mathrm{4}}^{+}$. As can be seen in Fig. 10b, initially, NH${}_{\mathrm{4}}^{+}$ is somewhat lower, but it increases later and matches the 3 × base HCl run. This suggests that NH${}_{\mathrm{4}}^{+}$ formation in the model is more sensitive to changes in HCl than changes in NH₃ emission, while total NH_x agrees well by reducing the NH₃ emissions. In general, CTMs have higher NH₃ concentration than observations, further supporting models having too much NH₃. A few factors might contribute to the model discrepancies for NH₃: there are uncertainties in the emission inventory of the bottom-up approach of NH₃, and the model does not currently include the bidirectional exchange of NH₃ with surfaces, such as dew and fog water. Also, the model does not have accurate industrial sources of HCl emission. Diurnal emission profiles are uncertain for both NH₃ and HCl. Furthermore, gas-to-particle partitioning associated with SO₂ oxidation pathways in the model is not correct at present.

4 Conclusions

In this study, for the first time in South Asia we have evaluated the performance of a chemical transport model (WRF-Chem) in modeling NH₃, NH${}_{\mathrm{4}}^{+}$, and total NH_x, by comparing them against the WiFEX measurements (MARGA). In daily means, we find NH₃ is significantly overestimated by the model, and NH${}_{\mathrm{4}}^{+}$ was underestimated, while simulated total NH_x agreed well with the measurement, indicating that incorrect gas-to-particle partitioning, along with missing chemical processes, may impact this mismatch in the model. The ability of the model to accurately describe the gas-to-particle partitioning of the MARGA was evaluated by the fraction of total NH_x (i.e., NH₃+ NH${}_{\mathrm{4}}^{+}$) in the particulate phase (NH${}_{\mathrm{4}}^{+}$ $/$ NH_x). A strong relation of MARGA NH${}_{\mathrm{4}}^{+}$ $/$ NH_x was observed with the dominant anion (Cl⁻) (r = 0.79), whereas the standard model showed a strong correlation between NH${}_{\mathrm{4}}^{+}$ $/$ NH_x with the dominant anion (SO${}_{\mathrm{4}}^{\mathrm{2}-}$) (r = 0.77), pointing to the missing chloride (HCl $/$ Cl⁻) chemistry in the model.

We incorporated HCl $/$ Cl⁻ emissions in the model and conducted three sensitivity experiments of varying HCl emissions, named the no HCl (0 mol km⁻² h⁻¹), base case HCl (3× Sharma et al., 2019; 24.8 mol km⁻² h⁻¹), and 3 × base HCl (74 mol km⁻² h⁻¹) run. The revised model shows that by adding HCl emissions, more NH_x was partitioned to the condensed phase, improving agreement with the observations. The 3 × base HCl run was able to represent the diurnal variation of NH${}_{\mathrm{4}}^{+}$ and Cl⁻ well, both in terms of amount and pattern with improved NMB for NH₃. Additional sensitivity tests in changing NH₃ emissions (reduction by a factor of 3) in the 3 × base HCl also improved NH₃, NH${}_{\mathrm{4}}^{+}$, and NH_x concentrations. We find excess NH₃ along with longer lifetime of NH${}_{\mathrm{4}}^{+}$ may act as a controlling driver for NH_x overestimation in the model. These results highlight the need to include correct industrial sources of HCl emissions, along with appropriate emissions of NH₃, to reduce biases in NH_x. Appropriate NH₃ emissions are developed using country-specific emission inventories, which are currently under development as part of the Global Challenges Research Fund (GCRF), South Asian Nitrogen Hub (SANH). Also, there is potential to develop top-down constraints on NH₃ emissions by taking inference from the satellite-, model-, and ground-based observations. Challenges remain in simulating NH₃ as a contributor to particulate matter due to temporal factors in ammonia peaks, including the role of fog and dew, where more work is needed. This work also suggests that model improvements to SO₂ oxidation pathways could improve NH_x partitioning.