Differentiation Between Phyllodes Tumors and Fibroadenomas of Breast Using Mammography-based Machine Learning Methods: A Preliminary Study.

OBJECTIVE: To investigate the diagnostic performance of a mammography-based radiomics model for distinguishing phyllodes tumors (PTs) from fibroadenomas (FAs) of the breast. MATERIALS AND METHODS: A total of 156 patients were retrospectively included (75 with PTs, 81 with FAs) and divided into training and validation groups at a ratio of 7:3. Radiomics features were extracted from craniocaudal and mediolateral oblique images. The least absolute shrinkage and selection operator (LASSO) algorithm and principal component analysis (PCA) were performed to select features. Three machine learning classifiers, including logistic regression (LR), K-nearest neighbor classifier (KNN) and support vector machine (SVM), were implemented in the radiomics model, imaging model and combined model. Receiver operating characteristic curves, area under the curve (AUC), sensitivity and specificity were computed. RESULTS: Among 1084 features, the LASSO algorithm selected 17 features, and PCA further selected 6 features. Three machine learning classifiers yielded the same AUC of 0.935 in the validation group for the radiomics model. In the imaging model, KNN yielded the highest accuracy rate of 89.4% and AUC of 0.947 in the validation set. For the combined model, the SVM classifier reached the highest AUC of 0.918 with an accuracy rate of 86.2%, sensitivity of 83.9%, and specificity of 89.4% in the training group. In the validation group, LR yielded the highest AUC of 0.973. The combined model had a relatively higher AUC than the radiomics model or imaging model, especially in the validation group. CONCLUSIONS: Mammography-based radiomics features demonstrate good diagnostic performance for discriminating PTs from FAs.

[Concurrent Collection of Ammonia Gas and Aerosol Ammonium in Urban Beijing During National Celebration Days Utilizing an Acid-Coated Honeycomb Denuder in Combination with a Filter System].

Since 2013, the Chinese government implemented the Air Pollution Prevention and Control Action Plan. As a result, the atmospheric concentrations of sulfate reduced significantly, whereas the nitrate concentrations remain relatively high due to the excess of ammonia (NH3). To date, there is no official observation network monitoring NH3 concentrations in China. Previous studies have focused on NH3 or ammonium (NH4+) separately. These limitations hinder a complete understanding of their dynamic changes due to the rapid gas-to-particle conversion. In this study, the concentrations of NH3 and NH4+ were measured concurrently in urban Beijing during autumn 2019 utilizing an acid-coated denuder-filter combination with a time resolution from 2 h (PM2.5>35 µg·m-3) to 5 h (PM2.5<35 µg·m-3). The mean concentrations of NH3 and NH4+ during the study were (4.1±2.9)µg·m-3 and (1.7±1.4) µg·m-3, respectively. The temporal patterns of NH3 or NH4+ coincided with that of PM2.5, CO, and NO2 throughout the sampling period. The diurnal distributions of NH3 were bimodal, both on polluted (PM2.5>75 µg·m-3) and non-polluted (PM2.5<75 µg·m-3) days, peaking at 21:30-05:30 and 05:30-08:30, respectively. The NH3 concentrations on polluted days were relatively lower during 17:30-21:30, which may be related to higher wind speeds. In contrast to NH3, NH4+ had an obvious peak during 17:30-21:30 due to the formation of ammonium nitrate. The meteorological conditions favor the gas-to-particle conversion on polluted days, resulting in a lower NH3/NH4+ ratio of 0.8. However, this value may reach 2.8 on non-polluted days. The concentrations of NH3, CO, NO2, SO2, and PM2.5 in the emission control period showed a significant increase greater than or comparable to those in the non-control period by 54.2%, 40.4%, 33.3%, 0%, and 49.4%, respectively. This result shows that the stagnant conditions offset the benefit of emission control actions implemented during and before the National Celebration Day.

[Source Apportionment of Atmospheric Ammonia: Sensitivity Test Based on Stable Isotope Analysis in R Language].

Ammonia (NH3) is an important precursor of fine particles and nitrogen deposition. It is critical to identify and quantify the sources of NH3 before the implementation of a mitigation strategy. Stable isotope analysis in R (SIAR) has potential with regard to the source apportionment of NH3, but its reliability is closely related to the signatures (δ15N-NH3) of emission sources. Based on SIAR, we found that the agricultural contribution varied significantly with mean δ15N-NH3 values of endmember input. In contrast, both the contributions of fossil fuel and NH3 slip showed low sensitivity against the change of endmember input. Moreover, the agricultural contribution changed by about 20% due to the variations in agricultural endmember mean values. Such a change is five times that due to the variations in endmember standard deviation values. Notably, regardless of the number of input sources tested, "non-agricultural source" was the dominant source of NH3 during hazy days in January 2013 in Beijing. Since various agricultural sources showed large variations in δ15N-NH3, future studies should focus on the endmember signatures of agricultural sources to further reduce the uncertainty in SIAR-based NH3 source apportionment.

[Concurrent Measurement of Wet and Bulk Deposition of Trace Metals in Urban Beijing].

To characterize the dry and wet deposition of atmospheric trace elements in urban Beijing, both active and passive samplers were used to collect bulk and wet sedimentation samples between May 2014 and April 2015.The concentrations of 19 trace elements (Na, Mg, Al, K, Ca, V, Cr, Mn, Fe, Cu, Zn, As, Se, Mo, Cd, Sb, Tl, Th, and U) in the samples were analyzed by inductively coupled plasma mass spectrometry (ICP-MS). The results show that the concentrations of metals in bulk deposition samples[7160.68 µg·L-1 (Ca)-0.02 µg·L-1 (Th)] were generally higher than those in wet deposition samples[4237.74 µg·L-1 (Ca)-0.01 µg·L-1 (Th)], but the enrichment factors of each metal in the two kinds of samples were less different. Of note, the enrichment factors of Cu, As, Tl, Zn, Cd, Se, and Sb were all larger than 100, thus indicating that these heavy metals were mainly from anthropogenic sources. The statistical analysis of the air mass trajectory shows that the precipitation chemistry in urban Beijing is mainly affected by southward air flows. The air mass originating from the southwest region always had higher concentrations of Ca, Mg, Fe, Al, Cu, Mo, U, and Th, whereas the air mass from the south had higher concentrations of K, Zn, Mn, Sb, Cd, and Tl. During the observation period, the bulk deposition fluxes of metals varied from 3591.35 mg·(m2·a)-1 (Ca)-0.01 mg·(m2·a)-1 (Th), and wet deposition fluxes varied from 1847.78 mg·(m2·a)-1 (Ca)-0.01 mg·(m2·a)-1 (Th). The dry deposition fluxes of the 19 metals varied from 1743.57 mg·(m2·a)-1 (Ca)-0.01 mg·(m2·a)-1 (Th). The particle size has important implications in the evaluation of the relative importance of dry deposition versus wet deposition during the scavenging of trace elements in air.

[Atmospheric Dry Deposition Fluxes and Seasonal Variations of Particulate Matter and Lead in Urban Beijing].

To investigate the dry deposition pattern of particulate matter and lead in urban Beijing, dry deposited particles were collected based on a surrogate surface between December 2013 and November 2014 and subjected to elemental analysis by using microwave digestion and inductively coupled plasma mass spectrometry. The results showed that dry deposition fluxes of particle mass and total lead were 0.84 t·(hm2·a)-1 and 7.56 mg·(m2·a)-1, respectively, with a seasonal trend of spring > winter > autumn~summer, exhibiting a similar temporal pattern to that of coarse particles. In addition, dry deposition flux of water-soluble lead was 3.14 mg·(m2·a)-1. It was found that the ratio of soluble fraction to total lead in dry deposited particles reached up to 43%, coinciding with a seasonal variation of ambient sulfur dioxide. The findings highlighted the anthropogenic imprints on the dry deposition flux of particles and heavy metals as well as the solubility of lead. This study also provides a first-hand dataset that can be used to assess ecological and environmental impacts of dry deposited lead, and an up-to-date scientific basis for the crafting abatement strategies to further reduce emission of heavy metals in China.

[Dry and Bulk Nitrogen Deposition in Suburbs of Xining City].

This study conducted a full two-year (2014 and 2015) measurements of dry and bulk deposition of atmospheric inorganic nitrogen (N) at the suburb of Xining city, Qinghai province. Dry N deposition fluxes were calculated by multiplying the atmospheric concentrations of NH3 and NO2 measured using passive samplers with the modeled dry deposition velocities provided by the GEOS-Chem global chemical transport model, while bulk N deposition fluxes were measured using precipitation gauge. Annual mean concentrations of gaseous NH3 and NO2 averaged 8.8 and 19.6 µg·m-3, respectively, with significantly higher values in 2015 than in 2014. Seasonal mean NH3 concentrations were higher in spring and summer than in autumn and winter, but the concentrations of NO2 changed little from season to season, with a small peak in autumn. Annual mean concentrations of NH4+-N and NO3--N in precipitation averaged 2.2 and 1.8 mg·L-1, respectively. Concentrations of NH4+-N in autumn were~55% lower than those in other seasons, whereas those of NO3--N in autumn or winter were~26% higher than those in spring and summer. Dry deposition of NH3 and NO2 was 9.0 and 2.8 kg·(hm2·a)-1, respectively. Bulk deposition of NH4+-N and NO3--N in precipitation was 7.6 and 6.2 kg·(hm2·a)-1, respectively. Reduced N (gaseous NH3 and NH4+-N in precipitation) was the dominant form of N deposition. The total dry and wet N deposition was 25.6 kg·(hm2·a)-1, which represented significant nutrient input from the environment to the suburban farmland, but this amount of deposited N exceeded the critical loads[10-20 kg·(hm2·a)-1] of terrestrial ecosystems, suggesting a risk of "N saturation" in the local natural environment.

[Observations of Reactive Nitrogen and Sulfur Compounds During Haze Episodes Using a Denuder-based System].

Reactive nitrogen and sulfur compounds are chemically active in the atmosphere and play an important role in secondary particle formation. Among them, sulfate, nitrate, and ammonium (SNA) are important components of particulate matter (PM) that account for approximately one-third of fine particles. The precursors of SNA including HNO3, SO2, and NH3 are all involved in haze formation in China. To date, the concurrent measurements of SNA and their precursors have been limited to single sites and short terms because of the high cost of the instruments. This study aimed to use DELTA (Denuder for Long-Term Atmospheric sampling) to characterize the daily concentrations of reactive nitrogen and sulfur species during haze episodes. The results showed that the background interface of NH4+, NO3-, and SO42- in the denuder was minor and could be used to determine 24-48 h concentrations of NH3, HNO3, SO2, NH4+, and NO3-. However, the SO42- concentrations in the blank filter was so high that they could only be used for sampling weekly or for longer periods of time. During the campaign between May 9 and June 7, 2016 in urban Beijing, the concentrations of NH3, HNO3, NH4+, and NO3- showed distinct daily variations at different wind directions, i.e., higher values were observed during southerly winds and lower values during northerly winds. The time series of these reactive nitrogen compounds coincided with that of PM2.5, CO, SO2, and NO2, indicating the combustion of fossil fuels. The mean concentrations of NH3, HNO3, NH4+, and NO3- were twice the concentrations during clean days, further highlighting the effect of local emissions on the urban environment. The ratios of HNO3/NO3- and NH3/NH4+ were 1.2 and 4.5, respectively, without significant differences between hazy and clean days. The findings demonstrated that both the reduced and oxidized nitrogen were preferred in gaseous phase rather than particulate phase due to relatively high temperatures during the transition season of spring/summer.

[On-line measurement of water-soluble composition of particulate matter in Beijing].

Concentrations of NO3(-), SO4(2), NH4(+) and Cl(-) were measured in Beijing during March, 2011, at a time resolution of 15 minutes using the system for rapid collection of fine particles and ion chromatography (RCFP-IC). Meteorological parameters were recorded concurrently to investigate the formation processes of typical pollutants. Five pollution episodes were observed in one month during the study period. All measured ions had similar temporal distributions showing saw-tooth cycle variations with slow accumulation and relatively rapid clearing. The peak concentrations of NO3(-) and NH4(+): in typical pollution episodes were observed to be about an order of magnitude higher than the background levels while the maximum concentrations of SO4(2-) and Cl(-) were only 2-4 times as high as the background values. Two episodes were observed during and after the home heating period. Compared to those during the home heating period, concentrations of the four ions after the home heating period were 15% -60% lower, implying that the home heating affected the concentration of ions. Measurements of ionic concentrations from RCFP-IC and those from the high time resolution of flight aerosol mass spectrometry (HR-TOF-AMS) during the study period were compared. The comparison results suggested the RCFP-IC was more reliable for the measurements of water soluble species.

[Concentrations and ozone formation potentials of BTEX during 2008-2010 in urban Beijing, China].

Benzene, toluene, ethylbenzene and dimethylbenzene are typical anthropogenic emitted organics in the atmosphere, which not only endanger human health but also actively participate in photochemical reactions, generating O3 and secondary organic aerosols. In order to investigate the pollution level and its ozone formation potentials, concentrations of BTEX and O3 were parallel measured by the passive sampler and analyzed using GC-MS and ICS-90 during 2008-2010 in urban Beijing. The results show that toluene was the most abundant compound (8.7 +/- 3.1) microg x m(-3), followed by benzene, ethylbenzene and m/p-xylene, with concentrations of (7.1 +/- 3.3), (4.2 +/- 1.4) and (3.4 +/- 1.5) microg x m(-3), respectively. Concentrations of BTEX peaked in winter, followed by autumn, summer, and spring, with values of (26.8 +/- 12.1), (25.9 +/- 4.9), (24.7 +/- 2.8) and (16.8 +/- 1.4) microg x m(-3), respectively. Benzene concentrations were the highest in winter, while for toluene, the concentration was higher in summer than that in winter. Based on the maximum incremental reactivity scale, m/p-xylene was found to be the dominant contributor to ozone formation among BTEX. The ozone formation potentials of BTEX in Beijing were 65.2, 60.2, and 75.7 microg x m(-3) in 2008, 2009, and 2010, respectively, which were consistent with the measured values of 80.5, 65.0, and 101.9 microg x m(-3) during the corresponding period. Vehicular emission and solvent evaporation were the major sources of BTEX. Concentrations of benzene were affected by coal heating in winter, whereas BTEX concentration was more influenced by solvent evaporation in summer, which had an important contribution to the formation of O3 in urban Beijing.

[Chemical characteristics and sources of trace metals in precipitation collected from a typical industrial city in Northern China].

To investigate the chemical characteristics and possible sources of trace metals in precipitation of Tangshan, a typical industrial city in Northern China, precipitation samples were collected using an automated wet-only sampler for one year from December 2009 to November 2010 and subjected to chemical analysis using ICP-MS. The results showed that crustal elements (Ca, Mg, Na, K, Fe and Al) accounted for 97.72% of the total concentrations while heavy metals (Zn, Mn, Pb, Ba, Sb, Cu, Ni, As, V, Co and Se) only contributed 2.25% of the total. Zinc was the most abundant heavy metal and calcium had the highest concentration among the crustal elements, with volume-weighted mean concentrations of 88.7 microg x L(-1) and 4.5 mg x L(-1), respectively. Seasonal variations of trace metals were significant, with higher values observed in winter and spring than in summer and autumn. The calculation of crustal enrichment factors with Al as the reference element indicated that Tl, As, Ag, Zn, Pb, Cd, Se and Sb were mainly derived from anthropogenic sources and the enrichment factor values of those elements ranged from 10(2) to 10(5). In addition, the concentrations of heavy metals such as Pb, Cu and Ni in precipitation of Tangshan city were higher than those reported for Mt. Tai. The results of principle component analysis demonstrated that steel smelting emissions, fugitive dusts and coal combustion were probably the major contributors of trace metals in the precipitation of Tangshan City.

[Chemical composition and mass closure of particulate matter in Beijing, Tianjin and Hebei megacities, Northern China].

To seek an efficient prevention and control method of regional atmospheric pollution in Jingjinji area, mass concentrations and size-resolved chemical composition of particulate matter were investigated at four urban sites and one background site from June 2009 to May 2010. The results show that the annual mean concentration of PM10 was 124, 141, 151 and 183 microg x m(-3) in Beijing, Tianjin, Tangshan and Baoding respectively, while the corresponding percentage of daily limit value exceedances was 29%, 36%, 39% and 52%, which is well above the maximum allowed limit of 150 microg x m(-3) (National Air Quality Stand II). As for PM2.5, the annual average concentration was 55, 68, 79 and 116 microg x m(-3) in Beijing, Tianjin, Tangshan and Baoding, which is higher than that in the background site with a factor of 1.5, 1.9, 2.2 and 3.2, respectively, while the corresponding proportion of daily exceedances of 75 microg x m(-3) (WHO IT-1) was 29%, 33%, 42% and 65%. Seasonal variations of PM2.5 and PM10 concentration were significant at the urban sites with the highest value being measured in winter. In order to reconstruct the particle mass, the determined components were classified into five groups as follows: secondary inorganic aerosol, sea salt, heavy metal, mineral matter and construction dust. The urban-sites-averaged contribution of these components to PM2.1 was 28.5%, 5.8%, 1.8%, 14.8% and 3.8%, whereas that to PM2.1-9 was 11.3%, 6.7%, 1.1%, 43.5% and 7.6%, respectively. The particle pollution was very severe in Baoding with the major component of secondary inorganic aerosol and mineral matter in PM2.1 and PM2.1-9, respectively. The estimated contribution of anthropogenic sources to PM2.1 in Beijing was larger than that of natural sources with a factor of 3.5 whereas the ratio of anthropogenic sources to natural ones was 0.6 for PM2.1-9. The contribution of secondary components to PM2.1 was equivalent to primary emissions, which suggests the precursors emitted from coal combustion and vehicle exhaust should be controlled in the target area. In contrast, the ratio of primary emissions to secondary particulate matters in PM2.1-9 was up to 5, indicating measures are required to reduce dust from construction areas. Finally, the reduction of human health-related heavy metals is also necessary despite its minor contribution to particles.

[Determination of trace metals in atmospheric dry deposition with a heavy matrix of PUF by inductively coupled plasma mass spectroscopy after microwave digestion].

Interest in atmospheric dry deposition results mostly from concerns about the effects of the deposited trace elements entering waterbody, soil and vegetation as well as their subsequent health effects. A microwave assisted digestion method followed by inductively coupled plasma mass spectrometric (MAD-ICP/MS) analysis was developed to determine the concentrations of a large number of trace metals in atmospheric dry deposition samples with a heavy matrix of polyurethane foam (PUF). A combination of HNO3-H2O2-HF was used for digestion. The experimental protocol for the microwave assisted digestion was established using two different SRMs (GBW 07401, Soil and GBW 08401, Coal fly ash). Subsequently, blanks and limits of detection for total trace metal concentrations were determined for PUF filter which was used for dry deposition sampling. Finally, the optimized digestion method was applied to real world atmospheric dry deposition samples collected at 10 sites in Jingjinji area in winter from Dec. 2007 to Feb. 2008. The results showed that the area-averaged total mass fluxes ranged between 85 and 912 mg x (m2 x d)(-1), and fluxes of most elements were highest at Baoding and lowest at Xinglong. In addition, the elemental fluxes in urban areas of Beijing, Tianjin and Tangshan were measured to be higher than that in suburb and rural sites. The average fluxes of crust elements (A1, Fe, Mn, K, Na, Ca and Mg) were one to three orders of magnitude higher than anthropogenic elements (Cu, Pb, Cr, Ni, V, Zn and Ba), varying from 151 to 16034 microg x (m2 x d)(-1) versus 14 to 243 microg x (m2 x d)(-1). Zinc was the most abundant heavy metal and calcium the highest of the crust elements while the elements Mo, Co, Cd, As and Be deposited less or even could not be detected. The anthropogenic and crustal contributions were estimated by employing enrichment factors (EF) calculated relative to the average crustal composition. The EF values of all elements except Pb and Zn were below 10, suggesting that local soil and/or dust generally dominate in the dry deposition flux.

[Observation on atmospheric pollution in Xianghe during Beijing 2008 Olympic Games].

There is a concern that much of the atmospheric pollution experienced in Beijing is regional in nature and not attributable to local sources. The objective of this study is to examine the contribution of sources outside Beijing to atmospheric pollution levels during Beijing 2008 Olympic Games. The observations of SO2, NO(x), O3, PM2.5 and PM10 were conducted from June 1 to September 30, 2008 in Xianghe, a rural site about 70 km southeast of Beijing. Sources and transportation of atmospheric pollution during the experiment were discussed with surface meteorology data and backward trajectories calculated using HYSPLIT model. The results showed that the daily average maximum (mean +/- standard deviation) concentrations of SO2, NO(x), O3, PM2.5, and PM10 during observation reached 84.4(13.4 +/- 15.2), 43.3 (15.9 +/- 9.1), 230 (82 +/- 38), 184 (76 +/- 42) and 248 (113 +/- 52) microg x m(-3), respectively. In particular, during the pollution episodes from July 20 to August 12, the hourly average concentration of O3 exceeded the National Ambient Air Quality Standard II for 46 h (9%), and the daily average concentration of PM10 exceeded the Standard for 11 d (46%); PM2.5 exceeded the US EPA Standard for 18 d (75%). The daily average concentrations of SO2, NO(x), O3, PM2.5 and PM10 decreased from 27.7, 18.6, 96, 90, 127 microg x m(-3) in June-July to 5.8, 13.2, 80, 60, 106 microg x m(-3) during Olympic Games (August-September), respectively. The typical diurnal variations of NO(x), PM2.5 and PM10 were similar, peaking at 07:00 and 20:00, while the maximum of O3 occurred between 14:00 to 16:00 local time. The findings also suggested that the atmospheric pollution in Xianghe is related to local emission, regional transport as well as the meteorological conditions. Northerly wind and precipitation are favorable for diffusion and wet deposition of pollutants, while sustained south flows make the atmospheric pollution more serious. The lead-lag correlation analysis during the pollution episodes from July 20 to August 12 showed that there are about 6-10 h (0.57 < r < 0.65, p = 0.01) of hourly average PM2.5 in Beijing lagging Xianghe, reaching the maximum at 8 h, which indicates that the real-time atmospheric PM2.5 database of Xianghe might provides early warning for the Beijing PM2.5 pollution events.

[Application of passive sampler to monitor and study atmospheric trace gases in Beijing-Tianjin-Hebei area].

Applying of passive sampler which is a cheap and simple method, the concentrations of SO2, NO2, O3 and NH3 were monitored in ten stations as network in Beijing-Tianjin-Hebei area from Dec. 2007 to investigate the concentration level and change of contaminants as well as the compound pollution in this region. Passive sampler method was fully evaluated to check the applicability in the area and the concentration and spatial distribution of pollutants were studied in this paper. The evaluation results indicate that passive sampler method can be used for long-term sampling and the frequency was settled as one month. The coefficients of variation of parallel samplers of SO2, NO2, O3 and NH3 were 6.4%, 7.1%, 4.2% and 3.9% respectively which can represent the good stability of this method. The concentrations monitored by passive sampler had good consistency with the monthly average concentrations calculated from active monitoring. The correlation coefficients of SO2, NO2 and O3 were 0.91, 0.88 and 0.93, slopes of fitted curve were 1.25, 0.98 and 0.93 and average relative standard deviation were 23.3%, 14.9% and 8.5% respectively which indicated that passive sampler can basically meet the atmospheric sampling requirement. Short-term monitoring of NH3 also indicated that results of passive sampler and active monitoring were comparable. Passive sampler is proven to be a reliable atmospheric monitoring method and can be used in regional pollution investigation. In the summer of 2008, average concentrations which were monitored by passive sampler calculated from 10 observation stations of SO2, NO2, O3 and NH3 were (12.3 +/- 6. 3) x 10(-9), (13.2 +/- 7.0) x 10(-9), (40.5 +/- 9.5) x 10(-9) and (24.0 +/- 13.7) x 10(-9) respectively in Beijing-Tianjin-Hebei area. The concentrations of SO2 and NO2 were relatively higher in city sites and concentrations of NH3 were higher in agricultural sites. The pollution of SO2, NO, and NH3, are obviously influenced by local emission. The concentrations of O3 were about 40 x 10(-9) in most area around Beijing and Tianjin except background station Xinglong and showed regional pollution characteristic.

[Variation of pH and chemical composition of precipitation by multi-step sampling in summer of Beijing 2007].

In order to understand variations of pH and chemical composition of precipitation in Beijing, 5 precipitation events in summer time of 2007 were sampled step by step sequentially on time, and the variations of pH, EC and the characterization of water-soluble ion such as NH4+, SO4(2-) and NO3(-) were analyzed. The results showed that pH was 5.70 +/- 0.73 in the beginning (before-35 min) and the precipitation was not acidic, but in the steady period (after-35 min) pH was 4.35 +/- 0.56 and the raining water presented acidity actually. Simultaneously, pH, EC and concentration of ions decreased rapidly with the raining continuously, 10-45 min later that became nearly constant and the concentration of water soluble ions decreased 50%-90% compare to beginning. The major pollutants in the precipitation on Aug. 1, which the raining air mass came from the northwest, was low and the concentrations of NH4+, SO4(2-) and NO3- were 65.4, 23.9 and 117.3 microeq/L, respectively; while the major pollutants in the precipitation on Aug. 6, which the air mass came from south, was high and the concentrations of NH4+, SO4(2-) and NO3- were 310.8, 95.7 and 249.8 microeq/L. But acidic precipitation was much higher when the air mass came from south than from northwest. The increasing fine particles from photo-chemical reaction in summer time of Beijing will result that the precipitation of the rain will be more and more acidic.