High NH3 deposition in the environs of a commercial fattening pig farm in central south China

Intensive livestock production has been increasing, and has resulted in the emission of more than seven teragram per year of ammonia (NH3) in China in recent years. However, little is known about the fate of the emitted NH3, especially the dry deposition of NH3 in the environs of intensive animal farms. In this study, the spatial and temporal variations of NH3 deposition in the environs of an intensive fattening pig farm were investigated in the central south of China. NH3 concentrations were measured at sites situated 50, 100, 200, 300, and 500 m in the downwind direction from the farm each month from July 2018 to June 2019. The NH3 deposition was calculated based on a bidirectional NH3 exchange model. The monthly NH3 emissions from the pig farm were estimated based on the breeding stock. The annual average NH3 concentrations ranged from 1200 to 14 μg m−3 at the downwind sites within 500 m of the pig farm, exhibiting exponential decay as distance increased. Strong seasonality in NH3 deposition was observed, with the highest season being in the summer and lowest in the winter, and air temperature was found to be an important factor affecting this seasonal variation. The estimated monthly total dry deposition within 500 m of the pig farm ranged from 92 to 1400 kg NH3–N mo−1, which accounted for 4.1%–14% of the total monthly NH3 emissions from the pig farm. The estimated total NH3 emissions and NH3 deposition from the pig farm were 63 000 kg NH3–N yr−1 and 5400 kg NH3–N yr−1, respectively, with the annual average ratio of NH3 deposition to NH3 emission being 8.6%. This study found NH3 deposition around intensive pig farms is high, and determined it as a significant fate of the NH3 emitted from pig farms.


Introduction
NH 3 is a highly reactive and alkaline gas with detrimental human health and ecological impacts (Gourley et al 2012. It originates from both natural and anthropogenic sources, with agriculture being its major source (Van Damme et al 2018, Guo et al 2020, Mueller and Lassaletta 2020).
The NH 3 emissions originating from the agricultural sector activities dominantly contribute to total anthropogenic NH 3 emissions, for example, above 80% on a global scale (Mencaroni et al 2021), 90% in Europe (Jacobsen et al 2019), and 88% in China . Other non-agricultural emission sources include industries (Cui et al 2013), urban waste (Elser et al 2018, Shao et al 2020, transport (Fenn et al 2018), residential (Bhattarai et al 2020), power plants , and biomass burning (Yu et al 2020). Anthropogenic NH 3 emissions contribute significantly to secondary aerosol formation, and thus contribute to the widespread regional haze and affect human health (Sutton et al 2008, Behera et al 2013, Bao et al 2019, Giannakis et al 2019.
The excess N input via atmospheric NH 3 deposition has noticeably detrimental effects on ecosystems, including soil acidification (Shen et al 2018), N 2 O emission enhancement (Xie et al 2018), and eutrophication and acidification of surface and ground water (Scudlark et al 2005, Zhan et al 2017. For example, the atmospheric deposition of NH 3 is a potential acid input, as recently described by Wang et al (2018). Soil acidification has been observed near feedlots owing to high local NH 3 deposition (Shen et al 2018). Xie et al (2018) reported high N 2 O emissions from a nitrogen-saturated subtropical forest in China. In addition, NH 3 deposition has become an important source of N content in surface water for the lakes, and may trigger the eutrophication and acidification of surface water (Scudlark et al 2005, Zhai et al 2009, Zhan et al 2017. Intensive animal farms are known as 'hotspots' for NH 3 emissions (Shen et al 2018). These NH 3 emissions return to the earth's surface via wet or dry deposition. NH 3 may completely dominate the overall load of reactive nitrogen (N r ) from the atmosphere near intense livestock farms (Zapletal and Mikuska 2019). Recent studies have reported NH 3 deposition from poultry facilities (Walker et al 2014, Baker et al 2020 and from typical intensive feedlots (Shen et al 2018, Zapletal and Mikuska 2019, Lassman et al 2020. Within a radius of 150-1000 m from the sources, approximately 3%-16% of NH 3 emissions deposit near the farms (Fowler et al 1998, Hao et al 2006, Walker et al 2008, Shen et al 2018, Zapletal and Mikuska 2019. Pig production is one of the largest sources of NH 3 emissions in China (Xu et al 2017). However, there are few studies on the NH 3 deposition in the environs of the commercial fattening pig farms. Furthermore, only a few studies have specifically investigated the links between NH 3 emissions from typical animal facilities and NH 3 deposition around these sources. The research objectives of this study were (a) to quantify NH 3 dry deposition within 500 m of the edge of an intensive commercial fattening pig farm in the central south of China, and to analyse the seasonal variations of NH 3 deposition; and (b) to gain insight into the relationship between NH 3 emissions and NH 3 deposition around the pig farm. Through this study, we can also know how much the emitted NH 3 or its derivative (e.g. particulate ammonium) will be transported to long distance. By quantifying the NH 3 deposition gradient around the pig farm, we can also further study the impacts of NH 3 deposition on the neighbouring natural ecosystems along a natural gradient.

Experimental site
The study was conducted at an intensive commercial fattening pig farm in Junchuan town (31 • 38 ′ 53 ′′ N, 113 • 13 ′ 48 ′′ E) located in Suizhou City, Hubei Province, China (figure 1). The study region is a hilly forested area, approximately 18 km away from the city of Suizhou. The altitude ranged from 90 to 127 m above the mean seal level. The annual precipitation was 940 mm. The average annual temperature was 15.6 • C. Winds were predominantly northerly, while relative humidity ranged between 39% and 99%, during the sampling periods. The dominant soils were Alumi-Ferric Alisols, Haplic Luvisols, and Anthraqui-umbric Gleysols, based on the Food and Agriculture Organization of the United Nations soil classification.
Land use within 500 m of the farm was divided into five categories (figure 1). The total area within 500 m of the farm covered 135 ha, which consisted of 53% forest, 17% arable land, 13% surface water, 9% shrubs, and 8% construction land (7% rural residential, and 1% traffic infrastructure).
The daily pig population in the building ranged from 1330 to 13 400 heads, with an average of 8900 heads. The fattening cycle lasted approximately 110 d. The studied farm had two pig houses set up in the south-north direction. The slurry facility is located next to the pig houses, where pig manure was piled and stored openly (figure 1). Fresh slurry from the pig house was cleared daily and added to the manure pile.

NH 3 emissions estimation 2.2.1. Pig building
In this study, the method reported by Zhu (2007) was used to estimate NH 3 emissions from a pig farm. The main sources of NH 3 emissions were manure and urine generated by the animals in the barn. The predictive models to estimate daily NH 3 emissions per pig were established by building relationships between the influencing factors (e.g. temperature, ventilation rate, nitrogen content of manure) of NH 3 emissions, and the NH 3 emissions of pig manure or urine per unit mass. These influencing factors were identified mainly based on the understanding of the processes of NH 3 emissions from pig farms. Similar studies for estimating NH 3 emissions from pig farms can be found in Aneja et al (2001), Harper et al (2004), andNi et al (1999). Because significant differences were observed in the parameter values in the abovementioned studies, we just used the parameters in the study of Zhu (2007), which was conducted in China. The NH 3 emissions for pig manure per unit mass were calculated using equation (1): where F NH3 is the NH 3 emissions of pig manure per unit mass in kg kg −1 , and T is the indoor temperature of the pig house, in • C (23 • C-28 • C). V is the ventilation rate in L min −1 , D F is the depth of pig manure in cm, and [N F ] is the nitrogen content of pig manure in g kg −1 (23 g kg −1 ). The D F is calculated using equations (2) and (3): where M Fpig is the weight of pig manure per head per day in kg head −1 , ρ F is the density of pig manure in kg m −3 (1005.9 kg m −3 ), S is the area per pig in the barn in m 2 head −1 , P barnpig is the pig population of the barn (unit: head), 100 is the conversion factor from m to cm, and 600 is the area available for pigs in the barn in m 2 . The fitted equation (4) for calculating the NH 3 emissions of pig urine per unit mass (U NH3 ) is expressed as follows: where U NH3 is the NH 3 emission of pig urine per unit mass in kg kg −1 , D U is the depth of pig urine in cm, and [N U ] is the nitrogen content of pig urine in g l −1 (2.85 g l −1 ). The D U is calculated using equation (5): where V Upig is the volume of urine per head per day in l head −1 , and 1000 is the conversion factor from l to m 3 .
The manure and urine production per pig per day were calculated using the model according to the First National Census of Pollution: Manual of Discharge Coefficient of Livestock and Poultry Industry (IEDA and NIES 2009). The total daily manure and urine production were calculated using equations (6)-(10): where M Fpig is the weight of pig manure per head per day in kg head −1 , W is the mean pig weight in kg, M F is the total daily manure production in kg, P pigbuilding is the daily pig population in the building (unit: head), V U is the volume of the total daily urine production in the building in l, V Upig is the volume of urine per head per day in l head −1 , M U is the mass of the total daily pig urine production in the building in kg, ρ U is the density of pig urine in kg m −3 (1000.3 kg m −3 ), 1.18 is the given pollution coefficient per pig in kg head −1 , 74 is the reference weight of pig (74 kg), 3.18 is the given pollution coefficient per pig in l head −1 , and 1000 is the conversion factor, from l to m 3 . The daily NH 3 emissions from the pig building (B NH3 , kg) were calculated using equation (11):

Manure pile
The cumulative NH 3 emissions of daily manure production from the open-pile storage of pig manure were calculated using equation (12) as follows: where M NH3j is the cumulative NH 3 -N emissions from manure pile on day j in kg, M Fj is the total daily manure production on day j in kg, H is the height of the manure pile in m (0.5 m), f NH3 is the emission factor of pig manure pile in g NH 3 -N m −2 d −1 (3.5 g NH 3 -N m −2 d −1 ) (Shan et al 2019), N is the number of days in a month in d, j is the jth day of the month in d, and 1000 is the conversion factor, from g to kg.

Total NH 3 emissions
Monthly and annual NH 3 emissions were extrapolated from the daily NH 3 emissions. Monthly and annual NH 3 emissions were calculated using equations (13) and (14) as follows: where E Tolmonthi is the NH 3 emissions in the ith month in kg, B NH3ij is the NH 3 emissions from the pig building on the jth day of the ith month in kg, M NH3ij is the cumulative NH 3 emissions from the pig manure pile on the jth day of the ith month in kg; N is the number of days in month i; E Tolyear is the annual NH 3 emissions in kg; and M is the number of months in a year.

NH 3 concentration monitoring
The NH 3 concentrations were measured using the active denuder for long-term atmospheric (DELTA) sampling system (Tang et al 2001, 2009, Sutton et al 2001a, 2001b, Zhu et al 2021. NH 3 samples were collected each day for five continuous days in the middle or towards the end of each month between July 2018 and June 2019. According to the prevailing direction during the sampling periods, NH 3 air concentrations were measured along one of the eight transects (north, northeast, east, southeast, south, southwest, west, and northwest) downwind of the pig farm. The DELTA systems were placed at five distances from the farm (50, 100, 200, 300, and 500 m), and one system was placed 200 m upwind of the pig farm to measure background NH 3 levels. NH 3 concentrations were measured at 1.5 m above ground level in the open areas. The methods for samples preparing, extraction and analysis were detailed in Tang et al (2009). In this study, the quality control method described in Tang et al (2009) was referred to assure the quality of the measured NH 3 concentrations.

NH 3 dry depositions flux calculation
According to Nemitz et al (2001) and Shen et al (2016), the NH 3 dry deposition around the pig building in this study was estimated using a bi-directional NH 3 exchange model. Based on equations (15)-(17), the total NH 3 dry deposition flux (F t ) can be calculated as follows: where x c is the canopy NH 3 compensation point; and R a , R b , R g , R s , R w , x g , and x s are the aerodynamic resistance, quasi-laminar boundary layer resistance, incanopy resistance to the ground, stomatal resistance, circular resistance, ground layer NH 3 compensation point, and stomatal compensation point, respectively. The seven parameters listed above were calculated according to the methods reported by Wesely (1989) for R g , R s and R w , Erisman and Draaijers (1995) for R a and R b , and Massad et al (2010) for x g and x s . x a is the measured NH 3 concentration. x(z 0 ) is the NH 3 concentration at the height d + z 0 , d is the zero-plane displacement height, z 0 is the surface roughness length, and F t is the total NH 3 dry deposition flux. More information about the bi-directional NH 3 exchange model can also be found in Zhu et al (2021). In theory, NH 3 deposition principally occurs in the downwind areas of pig farms (Shen et al 2016). In this study, the background NH 3 concentrations were relatively high (mean: 7.9 µg N m −3 , maximum: 17.3 µg N m −3 , and minimum: 2.1 µg N m −3 ), which were near the average NH 3 concentration at rural monitoring sites (8.2 µg N m −3 ) in 2018 in China (Wen et al 2020). The background NH 3 concentration was used to calculate background NH 3 deposition using equation (17). NH 3 deposition in the downwind area caused by NH 3 emissions from the pig farm was then calculated by subtracting the background NH 3 deposition from the total NH 3 deposition in the downwind area (Yi et al 2020). The area within 500 m of the pig farm was divided into eight downwind sectors based on a combination of eight major wind directions (shown in figure S1 available online at stacks.iop.org/ERL/16/125007/mmedia). Each downwind site was further divided into five sub-areas: (a) area within 50 m from the pig farm, (b) area between 50 and 100 m from the pig farm, (c) area between 100 and 200 m from the pig farm, (d) area between 200 and 300 m from the pig farm, and (e) area between 300 and 500 m from the pig farm. The monthly NH 3 deposition in the downwind area was calculated by multiplying the frequency of wind direction in a month with the accumulated NH 3 deposition in five sub-areas of the downwind area. The monthly NH 3 deposition flux within 500 m from the pig farm was then calculated using equation (18): where T Dk is the monthly NH 3 deposition (kg N mo −1 ) in the area located 500 m away from the pig farm, in month k, A ij is the size (ha) of the jth sub-area of the ith downwind area; the summation of A ij is the total downwind area (ha); D j is the NH 3 deposition rate (kg N ha −1 mo −1 ) in the jth sub-area; f i is the frequency of the ith wind direction in a year; and 1000 is the unit conversation factor. By summing the monthly NH 3 deposition, the total annual NH 3 deposition within 500 m of the pig farm (T D , kg N yr −1 ) was obtained using equation (19):

Monthly NH 3 emissions from the pig farm
In this study, the emissions from the pig building and manure storage facilities were estimated to be 63 100 kg NH 3 -N yr −1 . The pig building was the largest source of total NH 3 emissions (>90%) in the farm, as shown in figure 2. The monthly NH 3 emissions of the pig building for the period between July 2018 and June 2019 ranged from 2100 to 10 000 kg, with an average of 5210 kg. The daily NH 3 emissions ranged from 25 kg NH 3 -N d −1 to 400 kg NH 3 -N d −1 , with an average of 173 kg NH 3 -N d −1 . The mean NH 3 emissions rate in the study was calculated to be 17.9 g NH 3 -N head −1 d −1 .

Monthly mean NH 3 concentrations at downwind sites
The NH 3 concentrations in the study exhibited significant spatial-temporal variations, as shown in figure 3(a) 3(b)). High concentrations of NH 3 occurred mainly in summer, whereas NH 3 concentrations in autumn and spring declined rapidly and reached the minimum level in winter.

Monthly NH 3 dry depositions in the environs of the pig farm
The monthly NH 3 deposition fluxes also varied strongly in space and in time (table 1), ranging from 0.03 to 8.7 µg N m 2 s −1 from July 2018 to June 2019. NH 3 deposition fluxes declined significantly as distance from the farm increased. The highest NH 3 deposition fluxes generally occurred at a distance of 50 m, while the lowest NH 3 deposition fluxes were observed at a distance of 500 m. Table 1 depicts the mean monthly NH 3 deposition fluxes during the sampling periods under the land use types of forest, shrubs, paddy, and inland water. There was a large variation in the mean NH 3 deposition fluxes among the four land use types. The NH 3 deposition fluxes of forest, shrubs, paddy and inland water ranged from 0.08-8.8 µg N m 2 s −1 , 0.04-7.8 µg N m 2 s −1 , 0.12-7.7 µg N m 2 s −1 , and 0.03-3.8 µg N m 2 s −1 , respectively. NH 3 deposition flux also exhibited a decreasing trend as distance from the pig farm increased (from 50 to 500 m) along the eight transects ( figure 4). The estimated total annual NH 3 -N deposition in the areas within 500 m of the pig farm to be 5400 kg N yr −1 (table 2) or 40 kg N ha −1 yr −1 as an area-weighted mean.

Percentage of NH 3 depositions in the environs of pig farms emitting NH 3
The monthly percentage of NH 3 deposition in the 500 m of pig farm due to the NH 3 emissions from the farm to the total NH 3 emissions from the farm was calculated to indicate the fate of emitted NH 3 in the environs of pig farms. The percentage was in the range    of 4.1%-14%, with an average of 7.6% (shown in figure 5(a)). The percentage was highest in June, and lowest in February. The percentage tendency could be divided into three parts: the percentage sharply decreased from 14% in July to 6% in November; then remained steady in December and January by approximately 6%. Finally, the percentage increased from 4.1% in February to 14% in June. Moreover, the trend of the percentage was consistent with that of the temperature ( figure 5(b)).

High NH 3 deposition around the pig farm
In this study, NH 3 deposition was high within 500 m of the pig farm. The study's estimates of NH 3 deposition fluxes were higher than those reported in other studies. Walker et al (2014) estimated the NH 3 deposition nearby a large poultry facility with 4000 000 laying hens and 750 000 pullets to be 10.1 kg N ha −1 yr −1 at the refuge boundary, decreasing to 5.4 kg N ha −1 yr −1 1500 m. The results of the study conducted by Fowler et al (1998) showed that NH 3 deposition close to a large poultry unit of 120 000 broiler chickens declined from 42 kg N ha −1 yr −1 at 15 m to 5 kg N ha −1 yr −1 at 270 m, with annual emissions of 4800 kg NH 3 -N. Walker et al (2008) reported that NH 3 deposition near a swine production facility with a monthly stock of approximately 4900 pigs ranging from 145 kg N ha −1 yr −1 at 10 m from the source to 16 kg N ha −1 yr −1 at 500 m, with annual emissions of 34 000 kg NH 3 -N. McGinn et al (2016) reported a decrease in deposition with distance from the feedlot, with the average stock of 8200 cattle, with deposition decreasing by 50% over 200 m, from 519 to 260 kg N ha −1 yr −1 . The differences of deposition rates between this and other studies were mainly related to source strength (e.g. animal type, animal population, housing type) and environmental factors (e.g. climate type, terrain, and land use). There were an average stock of 8900 head of pigs in the studied farm, which caused high NH 3 emissions as well as high NH 3 deposition in the environs of the farm. Another possible explanation for the significantly higher NH 3 deposition in the study was the presence of the extensive coniferous forest in the farm environs, which may serve as a barrier to NH 3 horizontal dispersion. Previous studies have also shown that tree belts around farms could be used as an effective way of removing ammonia from the air (Bealey et al 2014(Bealey et al , 2016. The large NH 3 deposition flux gradient between 50 and 500 m is attributable to the fast dispersion and dilution of the NH 3 plume (Shen et al 2016).
In fact, NH 3 will also be wet deposited via scavenging in precipitation or the dry and wet deposition of particulate ammonium, although the component of aerosol ammonium will presumably be negligible compared with gaseous ammonia, since there is insufficient time for NH 3 emissions from the pig farm to convert to ammonium within the 500 m distance from the farm. The annual total precipitation in the study site was approximately 900 mm, thus the lack of estimate of wet deposition of NH 3 might cause the underestimation of the total NH 3 deposition around the pig farm.
Our assessment of the area-weighted mean NH 3 deposition rate (40 kg N ha −1 yr −1 ) indicated higher levels of NH 3 deposition compared with those of typical NH 3 deposition in eastern China known as the NH 3 emission 'hotspot' (deposition 8 kg N ha −1 yr −1 ) (Liu et al 2020). The dose effect of NH 3 deposition was based on critical loads (i.e. the deposition levels below which 'significant harmful effects' did not occur (Posch et al 2015)). Liu et al (2011) suggested that N critical loads for N deposition in subtropical coniferous forests in China were 15-30 kg N ha −1 yr −1 . In this study, subtropical coniferous forests (Masson pine forest) covered 53% of the study area. The annual average NH 3 -N deposition rate within 500 m of the pig farm exceeded the critical load. Excess N may lead to potential risk of soil acidification and cause increased N 2 O emissions from the Masson pine forest (Xie et al 2018), and result in a decline in forest growth rate (Huang et al 2015).

Seasonal variation of NH 3 deposition
The study showed that meteorological conditions were critical in shaping the seasonality of NH 3 concentrations, which is consistent with the study conducted by Walker et al (2014). The seasonal variation in NH 3 deposition was likely caused by environmental factors, such as temperature, precipitation, wind speed, and wind direction (shown in figures S2 and S3). The summer season exhibited the highest NH 3 deposition rate in the downwind area (2800 kg NH 3 -N), and NH 3 deposition in autumn, winter, and spring decreased by 53%, 83%, and 72%, respectively, compared with the summer deposition level. Previous results (Jones et al 2007) highlighted that NH 3 concentrations directly affect NH 3 deposition. Air temperatures affect the source intensity and soil and vegetation compensation points (Walker et al 2014), thus affecting NH 3 concentration in areas downwind of the pig farm. Accordingly, air temperature was a significant variable influencing NH 3 deposition. Previous studies (Cui et al 2011, Wen et al 2020, Deng et al 2021 have shown that precipitation leads to decreased NH 3 deposition. Cook et al (2018) suggested that precipitation is not the main driver of N deposition. One possible explanation for this is that the NH 3 depositions in the study area were sufficiently large to obscure the reduction by precipitation, especially in summer. As shown in figure 6, NH 3 deposition and NH 3 emissions were significantly and positively correlated with the monthly mean air temperature. High air temperatures usually favoured a high NH 3 emission rate and caused high NH 3 concentration as well as high NH 3 deposition.

Low percentage of NH 3 deposition in the neighbourhood to NH 3 emissions from the pig farm
The estimated annual NH 3 deposition (5400 kg N yr −1 ) in the area within 500 m from the studied pig farm accounted for 8.6% of the annual NH 3 emission (63 100 kg NH 3 -N yr −1 ). The percentage established in this study was compared with that found in other studies, as described in detail below. Fowler et al (1998) estimated that 3.8% of the total NH 3 emitted from a poultry farm with 120 000 broilers deposited to the woodland within 270 m from the farm. This study's estimated percentage was substantially lower than that reported by Yi et al (2020), which showed that NH 3 deposition in the 100 m neighbourhood of a 0.6 ha paddy field accounted for 80% of the NH 3 emitted from the paddy field. A possible explanation is that a smaller emission intensity of the emission source might lead to a higher percentage in the nearsource region. The percentage in this study was lower than that estimated by Hao et al (2006) (16%), probably owing to differences in NH 3 -emitting source strength (average 8900 heads of pig vs 50 000 heads of cattle). This study's results are slightly lower than those presented by Walker et al (2008) at 10%, whose study was conducted within 500 m of a pig farm with natural air flow. The percentage obtained in this study was close to the mean estimate reported by Shen et al (2018) and Zapletal and Mikuska (2019), who estimated that NH 3 deposition in the 400-1000 m environs of intensive feedlots accounted for 8% and 12% of the annual NH 3 emissions. Possible outcomes of additional NH 3 emitted from the farm being retained in the atmosphere without being deposited may be elevation to heights of 100-1500 m within the atmospheric mixing layer (Shen et al 2016), or spilling over into non-livestock production regions. The study region was close to cities with two small towns (Junchuan and Anju). The towns and cities produced high concentrations of acidic gas due to heating, transportation, and industry, at a distance of less than 18 km from the farm, which may favour for the formation of secondary aerosols.

Uncertainty analysis
In this study, NH 3 emissions from the pig farm were estimated using empirical models, thus the values still have some uncertainties. Based on the NH 3 emission factors (11-19 g NH 3 -N head −1 d −1 ) for pig from former studies (Balsdon et al 2000, Zahn et al 2001, Zhang et al 2010, Grant et al 2016, Ye et al 2019, the NH 3 emissions from the pig farm were 38 000-66 000 kg NH 3 -N, approximately 60%-104% of our estimation. In the Emission Database for Global Atmospheric Research database, the NH 3 emissions in China as reported by Crippa et al (2018), and Janssens-Maenhout et al (2015) were estimated based on the NH 3 inventory from Peking University (Huang et al 2012), which calculated NH 3 emissions from livestock wastes using the mass flow approach. Based on the Huang's method, as well as the updated Huang's method by Xu et al (2017), which reported total daily amount of provincial condition-specific N excretion rate for pigs), the estimated total NH 3 emissions for the studied pig farm was 45 t NH 3 -N, which was 71% of the estimated NH 3 emissions of this study. This indicates that our results are still reliable when compared with the former studies.
The uncertainties of the measured NH 3 by DELTA system was approximately 10% (Zhu et al 2021). The coefficient of variation for the daily NH 3 concentration and deposition measured at the same location in a month was 6%-19%, which showed relatively stable of NH 3 measurement. Though the bi-directional NH 3 exchange model is theoretically well established, but there are innate challenges in measuring the required parameters. The calculated NH 3 deposition is still subject to uncertainty in the model input parameters (R a , R b , R s , R w , R g , x g and x s ), because parameterization of these variables was mainly using the equations or empirical values based American or European studies. For evaluating the model, we calculated NH 3 dry deposition velocities by dividing the NH 3 deposition fluxes by NH 3 concentrations. The monthly NH 3 deposition velocities were on average 0.5-0.8, 0.3-0.8, 0.1-0.6, and 0.1-0.4 cm s −1 , for forest, shrubs, paddy and inland water, respectively. These deposition velocities are comparable with those published mean NH 3 deposition velocities for forest (0.1-3.0 cm s −1 ), farmland (0.13-0.75 cm s −1 ) and water (0.5-0.9 cm s −1 ) (Schrader andBrümmer 2014, Xu et al 2015), which indicates that the calculated NH 3 deposition fluxes in this study are in a reasonable range.

Conclusions
This study investigated NH 3 concentration measurements at 50, 100, 200, 300, and 500 m downwind of an intensive fattening pig farm with an average stock of 8900 animals in the central south of China from July 2018 to June 2019. The NH 3 deposition exhibited strong seasonality, which was mainly influenced by the temperature. The annual average NH 3 concentrations ranged from 1200 to 14 µg m −3 in the downwind direction within 500 m from the pig farm, exhibiting exponential decrease as the distance from the pig farm increased. Monthly NH 3 deposition ranged between 92 and 1400 kg NH 3 -N mo −1 , which accounted for 4.1%-14% of the total monthly NH 3 emissions from the pig farm. The estimated total NH 3 emissions and deposition from the pig farm were approximately 63 000 kg NH 3 -N yr −1 and 5400 kg NH 3 -N yr −1 , respectively, with an annual average percentage of NH 3 deposition to NH 3 emission of 8.6%. The study results suggest that NH 3 deposition around the source of NH 3 is an important result of the emitted NH 3 from pig farms and causes high N input in the pig farm environs. Further measuring and modelling studies are required to explore the effect of the emitted NH 3 from pig farms across areas in far proximity (e.g. more than 500 m).

Data availability statement
The data generated and/or analysed during the current study are not publicly available for legal/ethical reasons but are available from the corresponding author on reasonable request.