Numerical analysis of the impact of agricultural emissions on PM 2 . 5 in China using a 1 high-resolution ammonia emissions inventory

Abstract. China is one of the largest agricultural countries in the world. The NH3 emissions from agricultural activities in China significantly affect regional air quality and horizontal visibility. To reliably estimate the influence of NH3 on agriculture, a high-resolution agricultural NH3 emissions inventory, compiled with a 1 km × 1 km horizontal resolution, was applied to calculate the NH3 mass burden in China. The key emission factors of this inventory were enhanced by considering the results of many native experiments, and the activity data of spatial and temporal information were updated using statistical data from 2015. Fertilizer and husbandry, as well as farmland ecosystems, livestock waste, crop residue burning, fuel wood combustion, and other NH3 emission sources were included in the inventory. Furthermore, a source apportionment tool, ISAM (Integrated Source Apportionment Method), coupled with the air quality modeling system RAMS-CMAQ (Regional Atmospheric Modeling System and Community Multiscale Air Quality), was applied to capture the contribution of NH3 emitted from total agriculture (Tagr) in China. The aerosol mass concentration in 2015 was simulated, and the results showed that a high mass concentration of NH3, which exceeded 10 μg m−3, appeared mainly in the North China Plain (NCP), Central China (CNC), the Yangtz River Delta (YRD), and the Sichan Basin (SCB), and the annual average contribution of Tagr NH3 to PM2.5 mass burden in China was 14–18 %. Specific to the PM2.5 components, Tagr NH3 provided a major contribution to ammonium formation (87.6 %) but a tiny contribution to sulfate (2.2 %). In addition, several brute-force sensitivity tests were conducted to estimate the impact of Tagr NH3 emissions reduction on the PM2.5 mass burden. Compared with the results of ISAM, it was found that even though the Tagr NH3 only contributed 10.1 % of nitrate under current emissions scenarios, the reduction of nitrate could reach 98.8 % upon removal of the Tagr NH3 emissions. The main reason for this deviation could be that the NH3 contribution to nitrate is small under rich NH3 conditions and large in poor NH3 environments. Thus, the influence of NH3 on nitrate formation could be enhanced with the decrease of ambient NH3 mass concentration.



Introduction
60 Ammonia (NH3) is an important pollution species which principal neutralizing agent for the acid 61 aerosols, SO and NO formed from the SO2 and NOx (Chang, 1989;McMurry et al.;1983). In addition, 62 NH3 also influences the rate of particle nucleation (Ball et al.;1999;Kulmala et al.;2002) and enhances  Previous studies have investigated the influence of NH3 emissions on aerosol loading in several typical 105 areas of China. Wu et al. (2008) conducted sensitivity studies to assess the impact of livestock-produced 106 NH3 emissions on PM2.5 mass concentration in North China by using the MM5/CMAQ modeling system. 107 The results showed that the livestock-produced NH3 provided >20% contributions to nitrate and ammonium,    which indicates that the model can capture the regional variation in the measurements. The standard 192 deviations between the observations and simulations were similar in most cases, except for SO2 in January.

193
The largest deviation of the modeled mean, which was higher than that of the observation, was also between 194 the observed and modeled SO2 in January. However, the correlation coefficients reached 0.71 in January, 195 and the performance of the model in other months was relatively good, as shown in Table 1    Since NH3 concerns mainly with secondary inorganic aerosols: sulfate, nitrate, and ammonium (SNA) 226 formation, the analysis hereafter will mainly focus on the SNA.  Table 2. As shown in this table, Tagr NH3 provided the major contribution to ammonium, which reached 240 approximately 90%, and a relatively small contribution to nitrate mass burden, which was 5-10%. However, 241 the contribution to sulfate was tiny, and the main reason is that there are various methods of sulfate 242 formation from SO2 other than neutralization by NH3, such as oxidation by H2O2, O3, or peroxyaceticn acid. provided an approximately 14-18% contribution to the total PM2.5 mass concentration in these places, and 246 the largest annual average contribution appeared in CNC (17.5%).

247
In addition, the brute-force method (zero-out sensitivity test), which can capture the effect of emissions 248 changes on aerosol mass burden, was applied to investigate the impact of the removal of Tagr NH3 249 emissions. Unlike online source apportionment, the brute-force method mainly reflects the disparity of the 250 chemical balance caused by the emissions change, which could significantly alter secondary pollutant 251 formation. Several sensitivity tests were conducted, and the results are shown in Figure 7 and Table 3.  Table 3 shows the percentage of the variation of sulfate, nitrate, ammonium, SNA, and PM2.5.

256
Compared with Table 2, it can be seen that the variation percent of SNA and PM2.5, which reached 40-51% 257 and 23-35%, respectively, were approximately two times higher than those of the contribution percent, and 258 this significant distinction was mainly caused by the variation of nitrate: the contribution of Tagr NH3 to 259 nitrate was generally below 10%, as shown in Table 2, but the reduction of nitrate associated with removing 260 Tagr NH3 emissions could exceed 95%, as shown in Table 3. This difference between the results of ISAM 261 and brute-force was expected as a result of high nonlinearity in the NOx chemistry. The nitrate formation 262 could become more sensitive when the "rich NH3" environment shifts to a "poor NH3" environment, which 263 means the decrease of the nitrate mass burden would accelerate with the NH3 emissions reduction.

264
Therefore, it can be deduced that the contribution of NH3 to nitrate should be significantly lower under a 265 "rich NH3" environment than that under a "poor NH3" environment. A similar phenomenon was also respectively) is shown in Figure 8. It can be seen that the decrease in nitrate mass concentration was more 270 rapid than that of ammonium, and the trend became slightly faster with the reduction of NH3 emissions 271 (signifying the transition from a "rich NH3" to a "poor NH3" environment) in the regions with a high mass

276
The emission budget of agriculture NH3 was huge and played an important role on the regional particle 277 pollution in China. As a precursor of the secondary aerosol, reasonably estimate the nonlinear processes of 278 secondary aerosol formation should be the key point for capturing the contribution of NH3 to particle 279 pollution. In this study, the air quality modeling system RAMS-CMAQ was applied to simulate spatial-280 temporal distribution of trace gas and aerosols in 2015. In addition, the PKU-NH3 emission inventory which 281 compiled on 1km×1km horizontal resolution with monthly based data was applied to accurately capture 282 the agriculture NH3 emission features in China. Then, the source apportionment module ISAM was coupled 283 into this modeling system to quantitatively estimate the contribution of agriculture NH3 to PM2.5 mass 284 burden. The brute-force sensitivity tests were also conducted for discussing the impact of the agriculture 285 NH3 emission reduction. The meteorological factors and mass concentration of NH3, SO2, NO2, and PM2.5 286 from simulation were evaluated and showed well agreement with the observation data. Some interesting 287 results were explored and summarized as follow:

288
(1) The high mass burden of NH3 could exceeded 10 μg m -3 , and mainly appeared in the NCP, CNC, 289 YRD, and SCB. These regions were highly coincidence with the regions that heavy particle pollution 290 covered in China. Therefore, it can be deduced that the influence of agriculture NH3 on the PM2.5 mass 291 concentration should be significant.

292
(2) The results from ISAM simulation shows that the Tagr NH3 provided 14-18% contribution to the 293 PM2.5 in the most part of east China, and the largest annual average contribution appeared in CNC (17.5%).

294
Specific to the SNA components, the annually and regional average contribution of Tagr NH3 to ammonium, burden. The results indicated that the reduction percent of PM2.5 mass burden due to removal Tagr NH3 300 emission could reach 23-35% in the most part of east China, which was approximately two times higher 301 than the contribution. The reduction percent of nitrate that reached exceed 95% was the main reason caused 302 this significant different. In addition, the further analysis proved that the ambient NH3 mass burden could 303 obviously affects its contribution to the SNA formation: the NH3 contribution to nitrate should be lower 304 under "rich NH3" and higher under "poor NH3". Therefore, the influence of NH3 would enhance with the 305 decreasing of ambient NH3 mass concentration. 306 It is suggested that the influence of NH3 on the PM2.5 mass burden is complex because of the 307 nonlinearity of secondary aerosol formation. Significant deviation exists between the results from ISAM 308 and the brute-force method; therefore, these two kinds of results should be distinguished and applied to 309 explain different issues: the contribution under the current scenario and the effect due to emissions reduction, 310 respectively. The modeling system is a versatile tool that allows us to investigate this valuable information 311 to choose more efficient strategies for reducing the impact of agricultural NH3 and improving air quality.   530  531  532  533  534  535  536  537  538  539  540  541  542  543  544  545  546  547  548  549  550  551  552  553