3.1. Mass Concentrations of Atmospheric Particles
Within each functional area, three or four locations were selected to sample the ambient aerosols to study their PM characteristics. To provide a general description of the PM levels, the TSP, PM
10, PM
5 and PM
2.5 mass concentrations are listed in
Table 2. The PM
10/TSP and PM
2.5/PM
10 ratios are also listed in
Table 2. The TSP concentrations in Xiamen were found to range from 63.51 μg∙m
−3 to 142.02 μg∙m
−3, with a mean of 102.50 ± 23.14 μg∙m
−3. The PM
10 concentrations varied from 51.41 μg∙m
−3 to 116.95 μg∙m
−3, with a mean of 82.79 ± 20.79 μg∙m
−3. The PM
2.5 concentrations ranged from 22.49 to 70.38 μg∙m
−3, with a mean of 43.70 ± 15.23 μg∙m
−3. The mean PM
10/TSP ratio was approximately 0.806, which indicates that most of the TSP particles were inhalable particles smaller than 10 μm. The average PM
2.5/PM
10 ratio was 0.528, which is higher than that stipulated by the Chinese Ambient Air Quality Standards (GB 3095-2012; in the new standard II, PM
2.5 should account for no more than 50% of the ambient PM
10), especially near the hospitals and parks, which were near parking lots. These observations suggest that PM
2.5 in Xiamen may come from traffic vehicles. Xiamen is among the top 10 cleanest cities in China. All of the PM
10 and PM
2.5 mass concentrations were below the 24-hr mass-based standards of 150 μg∙m
−3 and 75 μg∙m
−3 set by the Chinese Ambient Air Quality Standards (GB 3095-2012, the new standard II). However, the concentrations exceeded the 24-hr mass-based standards of 50 μg∙m
−3 (PM
10) and 25 μg∙m
−3 (PM
2.5) set by the World Health Organization’s Healthy Ambient Air quality standards. The higher ambient average concentrations of small particulates (PM
10 and PM
2.5) suggest that inhalable particles, which have been linked to comparatively higher rates of premature mortality in many epidemiological studies, were the primary pollutants affecting human health during the study period in Xiamen.
Table 2.
Mass concentrations measured from the PM samples collected at different functional areas in Xiamen.
Table 2.
Mass concentrations measured from the PM samples collected at different functional areas in Xiamen.
Sampling Sites | Mass Concentrations (μg·m−3) | Ratios (%) |
---|
TSP | PM10 | PM5 | PM2.5 | PM10/TSP | PM2.5/PM10 |
---|
Parks | ZSP | 123.16 | 106.98 | 85.50 | 66.08 | 86.86 | 61.77 |
BLZP | 79.01 | 60.79 | 40.45 | 27.64 | 76.94 | 45.48 |
ZLP | - | - | - | 52.05 | - | - |
HLP | 119.19 | 100.70 | 76.12 | 47.87 | 84.49 | 47.54 |
Means | 107.12 | 89.49 | 67.36 | 48.41 | 82.76 | 51.60 |
Hospitals | FAH | 142.02 | 116.95 | 91.64 | 70.38 | 82.35 | 60.18 |
ZSH | 90.14 | 72.13 | 50.16 | 34.60 | 80.02 | 47.97 |
TCMH | 120.48 | 109.61 | 86.41 | 68.83 | 90.98 | 62.80 |
Means | 117.55 | 99.56 | 76.07 | 57.94 | 84.45 | 56.98 |
Commercial areas | ZSRC | 126.72 | 96.61 | 69.17 | 46.84 | 76.24 | 48.48 |
RJC | 91.31 | 70.47 | 49.93 | 34.30 | 77.17 | 48.68 |
MFC | 101.12 | 84.43 | 60.37 | 41.78 | 83.49 | 49.48 |
Means | 106.38 | 83.84 | 59.82 | 40.97 | 78.97 | 48.88 |
Industrial areas | KYI | - | 66.57 | 48.49 | 33.83 | - | 50.82 |
HLI | 63.51 | 51.41 | 35.82 | 22.49 | 80.95 | 43.74 |
HSI | 97.52 | 77.65 | 53.08 | 31.66 | 79.62 | 40.77 |
Means | 80.52 | 65.21 | 45.80 | 29.33 | 80.29 | 45.11 |
Residential areas | XXR | 106.26 | 84.19 | 60.42 | 39.20 | 79.23 | 46.56 |
QPR | 65.70 | 51.73 | 36.98 | 27.10 | 78.73 | 52.39 |
JSR | 108.90 | 91.60 | 69.60 | 54.50 | 84.11 | 59.50 |
Means | 93.62 | 75.84 | 55.67 | 40.27 | 80.69 | 52.82 |
Means | 102.50 | 82.79 | 60.94 | 43.70 | 81.51 | 51.07 |
SDs | 23.14 | 20.79 | 18.22 | 15.23 | 4.16 | 6.84 |
The average PM mass concentrations for the five functional areas exhibited the following order: hospitals > parks > commercial areas > residential areas > industrial areas. Among the 16 sites, the lowest observed PM levels in Xiamen were found in the Huli industrial area, whereas the highest PM values were observed at the First Affiliated Hospital of Xiamen University (FAH). At FAH, PM
2.5 accounted for approximately 60% of the PM
10, which nearly reached 75 μg∙m
−3. The PM
2.5 concentrations in Zhongshan Park, Xiamen Hospital of Traditional Chinese Medicine (TCMH) and Jinshang residential area were also higher than in the other areas. According to the study by the authors [
23], emissions from traffic vehicles are an important source of PM
2.5. The selected sampling sites were as far from the pollution sources as possible as to ensure that the sites were representative of the corresponding functional areas. However, traffic vehicles cannot be avoided in urban areas, especially near hospitals, parks, and commercial and residential areas, which are closely related to people’s daily lives. Thus, one reason for the higher PM pollution in these areas is the emissions. There are no heavy industries on Xiamen Island. The existing light industries, such as the optoelectronics and software industries, emit small amounts of PM into the atmosphere. Meanwhile, these industrial areas are relatively large with relatively few vehicle emissions and other pollution sources. Thus, the lowest concentrations were found in the industrial areas.
To identify the impact on the ambient air quality, correlation coefficients between the PM, API and visibility were calculated; the results are listed in
Table 3. Because the API was calculated based on the concentrations of five atmospheric pollutants,
i.e., PM
10, sulfur dioxide (SO
2), nitrogen dioxide (NO
2), carbon monoxide (CO), and ozone (O
3), measured at the monitoring stations throughout the city, the API exhibited significant positive correlation coefficients with PM
5 (r = 0.599 at
p = 0.05), PM
10 (r =0.588 at
p = 0.05), and TSP (r = 0.578 at
p = 0.05), although the same correlation was not identified for PM
2.5. These observations indicate that the mass concentrations of PM collected at different areas in our study were similar to the results of the environmental protection departments. Because high visibility is common when the air quality is good, significant negative correlation coefficients were identified between the API and visibility (r = −0.441 at
p = 0.01). Significant negative correlation coefficients were also found between visibility and PM
5 (r = −0.772 at
p = 0.01), PM
10 (r = −0.778 at
p = 0.01), and TSP (r = −0.728 at
p = 0.01), although this was not true for PM
2.5. These relationships suggest that visibility was primarily reduced via the scattering light due to the presence of larger particles (particulate matter with an aerodynamic diameter greater than or equal to 5.0 μm), such as PM
5 and PM
10, during the summertime observation period in Xiamen. Non-significant correlations between PM
2.5 and both the API and visibility suggest that PM
2.5 did not play a primary role in visibility and the API during the investigated period.
Table 2 and
Figure 3 show that the air quality was good during this period.
In this study, samples were not collected simultaneously; the meteorological conditions may be slightly different. Viability in the meteorological parameters was tested using the one-way analysis of variance (ANOVA) method. Significant (
p < 0.05) differences in wind speed were found between the five functional areas (F = 3.815 at
p = 0.035) during the sampling periods, whereas no significant differences for temperature (F = 0.745 at
p = 0.581), relative humidity (F = 1.27 at
p = 0.339), and pressure (F = 2.138 at
p = 0.144) were found during the sampling periods between the five functional areas. The wind speed is usually considered to have a very significant effect on the diffusion of particles. Thus, to study the influence of the meteorological conditions on PM, we also calculated the correlations between PM and wind speed, relative humidity and temperature; the results are shown in
Table 3. However, non-significant correlation coefficients were found between wind speed and PM levels in this study.
Figure 2 shows that most of the wind speeds were less than 2.0 m∙s
−1 (approximately 55.8% of the measurements were calm) at different sampling sites during the sampling periods. This finding suggests that the low wind speeds resulted in weak diffusion of PM
2.5; therefore, the wind had little effect on the mass concentrations of PM. Precipitation usually exerts a very significant scavenging effect on the airborne particles.
Figure 3 shows that there were several precipitation events between 5 August and 12 September 2012. Although atmospheric particles were not collected on rainy days, PM (and pollutant concentrations) may be affected by the previous day’s rain. Therefore, the correlation between precipitation and PM level was also calculated. Non-significant correlation coefficients were also found between precipitation events and PM levels in this study. These findings indicate that the variations in meteorological conditions had little effect on the PM levels at the different sites, although the meteorological conditions exhibited small variations during the study period. Thus, we compared our results primarily according to the different functional areas.
Table 3.
Relationships between PM, air quality and meteorological parameters during the sampling period.
Table 3.
Relationships between PM, air quality and meteorological parameters during the sampling period.
| Precip | R.H | W.S. | T | Vis | API | PM2.5 | PM5 | PM10 | TSP |
---|
Precip | 1 | 0.199 | 0.155 | −0.188 | −0.162 | −0.155 | 0.282 | 0.444 | 0.399 | 0.277 |
R.H | | 1 | −0.328 | 0.111 | −0.220 | 0.083 | −0.226 | 0.251 | 0.249 | 0.081 |
W.S. | | | 1 | −0.927 ** | 0.138 | −0.024 | −0.049 | −0.254 | −0.221 | −0.157 |
T | | | | 1 | 0.128 | 0.039 | 0.338 | −0.088 | −0.086 | 0.100 |
Vis | | | | | 1 | −0.441 ** | −0.146 | −0.772 ** | −0.778 ** | −0.728 ** |
API | | | | | | 1 | −0.105 | 0.599 * | 0.588 * | 0.578 * |
PM2.5 | | | | | | | 1 | −0.050 | −0.052 | −0.016 |
PM5 | | | | | | | | 1 | 0.992 ** | 0.953 ** |
PM10 | | | | | | | | | 1 | 0.979 ** |
TSP | | | | | | | | | | 1 |