**3. Results and discussion**

#### **3.1. Spatial distribution of PM 2.5 and major water-soluble ions**

Air pollution, especially PM2.5, has become a serious issue in East Asia, and there is rising public criticism regarding its effects. Concern in Japan is also increasing as winds transfer the pollution into domestic areas. Figure 4 shows the distribution of PM2.5 monitoring data at foursampling sites (i.e., a heavy traffic area (A), a residential area (B), an industrial area (C), and a desolate area (D)).

Site C, an industrial area, showed the heighest PM2.5 (65.3 µg m-3) followed by site A (35.3 µg m-3), site B (22.0 µg m-3), and site C (11.3 µg m-3). As might be expected, the highest PM2.5 level was monitored at site C where there is a compact mass of manufacturing companies. Among the four monitoring sites, two sites (i.e., site A and C) exceeded the Japan's PM2.5 criteria (a daily average of 35 µg m-3). The reason for the relatively high PM2.5 in both heavy traffic and industrial areas might be that much of fine particles and their precursors came from the vehicle emissions and fuel combustion and manufacturing processes in factories. The PM2.5 data monitored in this study are compared with those measured in other urban and rural areas in Asia during the springtime. Zhang *et al*. [12] reported that PM2.5 was 145.3 µg m-3 in Beijing during a non-Asian dust period in the springtime. Meanwhile, in Gosan, a typical rural area in Korea, PM2.5 during non-Asian dust period was measured at 26.1 µg m-3 [13].

**Figure 4.** distribution of PM 2.5 monitoring data at 4-sampling site

Using an EDX and a computer system, information about the elemental properties of asbestos fibers can be gathered and graphed in their appropriate relative ratios. Computation of the exact ratios of the elemental compositions in asbestos fiber allows the researcher to distinguish not only one type of asbestos fiber from another but also asbestos fibers from non-asbestos

In this study, for the purpose of observing and analyzing the morphological and chemical properties of airborne asbestos and pollen grains, an SEM (JEOL JSM-5400) equipped with an EDX (Philips, EDAX DX-4) was employed. The samples were placed inside the SEM's vacuum column (10-6 Torr) through an air-tight door. Pollen species were also distinguished and

Calculation of the airborne asbestos fiber concentration on the filter sample was carried out

*Fcon* <sup>=</sup> *Aeff* <sup>×</sup> *Ntotal*

where *Fcon* is airborne fiber concentration (fiber/L, f L-1), *Aeff* is effective collecting area of filter

*nfield* is total number of fields counted on the filter, and *Vair* is total air volume (L) calculated by

Air pollution, especially PM2.5, has become a serious issue in East Asia, and there is rising public criticism regarding its effects. Concern in Japan is also increasing as winds transfer the pollution into domestic areas. Figure 4 shows the distribution of PM2.5 monitoring data at foursampling sites (i.e., a heavy traffic area (A), a residential area (B), an industrial area (C), and a

Site C, an industrial area, showed the heighest PM2.5 (65.3 µg m-3) followed by site A (35.3 µg m-3), site B (22.0 µg m-3), and site C (11.3 µg m-3). As might be expected, the highest PM2.5 level was monitored at site C where there is a compact mass of manufacturing companies. Among the four monitoring sites, two sites (i.e., site A and C) exceeded the Japan's PM2.5 criteria (a daily average of 35 µg m-3). The reason for the relatively high PM2.5 in both heavy traffic and industrial areas might be that much of fine particles and their precursors came from the vehicle emissions and fuel combustion and manufacturing processes in factories. The PM2.5 data monitored in this study are compared with those measured in other urban and rural areas in Asia during the springtime. Zhang *et al*. [12] reported that PM2.5 was 145.3 µg m-3 in Beijing during a non-Asian dust period in the springtime. Meanwhile, in Gosan, a typical rural area

in Korea, PM2.5 during non-Asian dust period was measured at 26.1 µg m-3 [13].

), *Ntotal* is the total number of fibers in an SEM field area, *afield* is an SEM field area (cm2

),

*a field* ×*n field* ×*Vair*

countered under 3000 x magnification and 15 - 20 kV working conditions.

sample collection time (min) and pump flow rate (L/min).

**3.1. Spatial distribution of PM 2.5 and major water-soluble ions**

fibers.

208 Current Air Quality Issues

(cm2

using the following formula:

**3. Results and discussion**

desolate area (D)).

It is a matter of course that a large number of factories contributed to the high PM2.5 level and caused the regional worsening air pollution in a residential area (B). However, although the "yellow dust" warning did not issue during our field measurement, there is a possibility of the long-range transport of PM2.5 and its precursors from the Asian continent to the local study site in Japan. Thus, it is required to clear the uncertainties regarding the linkage between locally high PM2.5 at site C and its long-range transport under the springtime meteological conditions. In this study, in order to determine the source region of aerosols at the site C, the atmospheric backward dispersion model was applied.

Figure 5 displays the backward aerosol dispersion simulated by the NOAA Air Resources Laboratory (ARL) HYSPLIT dispersion-trajectory model started from site C. The area scales mean the integrated mass concentration [mass m-3] at 100 - 1000 m height of site C (33.39o N; 130.26o E).

A detailed model description of HYSPLIT was given in reference [14]. According to the result of the HYSPLIT model, a high value of aerosol concentration at the present receptor (meas‐ urement location) was not driven from the Chinese continent but was generated from a local area.

Ambient concentrations of major ionic species (i.e., nitrate, sulfate, and ammonium) associated with PM2.5 collected at four-sampling sites were overlapped with a map of Fukuoka Prefecture (Figure 6). As shown in Figure 6, the concentrations of major ionic species turned out to be of

**Figure 5.** Backward aerosol concentrations simulated by the NOAA ARL HYSPLIT model. The area scales mean the integrated mass concentration [mass m-3] at 10-100 m height of site C (33.39oN; 130.26oE)

considerable variation among four sites. Sulfate was the most abundant species to record the highest concentrations in all urban areas of Fukuoka Prefecture (a heavy traffic area (14.1 µg m-3), an industrial area (30.5 µg m-3), and a residential area (2.8 µg m-3). The sum concentrations of NH4 + , NO3 - , and SO4 2- varied in a similar way, as PM 2.5 (i.e., site C (38.4 µg m-3) > site A (18.3 µg m-3) > site D (5.4 µg m-3) > site B (4.9 µg m-3)).

**Figure 6.** Ambient concentrations of major water-soluble ions associated with PM2.5 collected at 4-sampling site in Fu‐ kuoka Prefecture

This suggests that the overwhelmingly high level of site C should be associated with the local pollution emissions. In the case of site C and A, the sum of three ionic components correspond to 58.8% and 51.8% of PM2.5, respectively. Therefore, it seems reasonable to say that PM2.5 in sites C and A was mainly composed by the secondary inorganic aerosol, which was formed by a gas-to-gas reaction in the atmosphere.

The concentrations of three kinds of ionic species in PM2.5 in this study are comparable to those of urban areas in Beijing, China, and Durg, India. Gao *et al*. [10] reported that the concentrations of NH4 + , NO3 - , and SO4 2- in PM2.5 collected in Beijing were 20.5, 15.2, and 42.3 µg m-3, respec‐ tively. On the other hand, those concentrations in Durg, a heavy traffic and industrial area, were marked as 2.1, 3.16, and 6.75 µg m-3, respectively [9]. The occupation ratio of the sum concentration of three ionic components in this area (0.89%) is greatly dissimilar to those of this study.

**Figure 5.** Backward aerosol concentrations simulated by the NOAA ARL HYSPLIT model. The area scales mean the

integrated mass concentration [mass m-3] at 10-100 m height of site C (33.39oN; 130.26oE)

210 Current Air Quality Issues

#### **3.2. Assessment of airborne asbestos fiber**

Even though asbestos is closely regulated in the present, the deposited and accumulated asbestos fibers for the past several tens of years can be resuspended from the ground near the distributing and manufacturing shops of asbestos products.

Asbestos from natural geologic deposits is known as "naturally occurring asbestos" (NOA) [14]. Health risks associated with exposure to NOA are not yet fully understood. As air quality associated with asbestos in ambient outdoor air has seldom been evaluated in Fukuoka Prefecture, we evaluated airborne asbestos in the urban environment there. Six mineral types are defined by the United States Environmental Protection Agency as asbestos [15]. Among them, crocidolite and amosite are commonly known to cause negative health effects such as lung cancer and mesothelioma [4].

Figure 7 shows an example of asbestos image of SEM, the elemental wt% list, and elemental spectrum obtained from EDX analysis.

**Figure 7.** An example of asbestos image of SEM (left upper) and the elemental wt% list (right upper) and spectrum (bottom) obtained from EDX analysis

A part of results of SEM-EDX analysis for asbestos fibers and the processes of classifying asbestos types based on the SEM-EDX elemental wt% data are illustrated in Figure 8.

**3.2. Assessment of airborne asbestos fiber**

212 Current Air Quality Issues

lung cancer and mesothelioma [4].

(bottom) obtained from EDX analysis

spectrum obtained from EDX analysis.

distributing and manufacturing shops of asbestos products.

Even though asbestos is closely regulated in the present, the deposited and accumulated asbestos fibers for the past several tens of years can be resuspended from the ground near the

Asbestos from natural geologic deposits is known as "naturally occurring asbestos" (NOA) [14]. Health risks associated with exposure to NOA are not yet fully understood. As air quality associated with asbestos in ambient outdoor air has seldom been evaluated in Fukuoka Prefecture, we evaluated airborne asbestos in the urban environment there. Six mineral types are defined by the United States Environmental Protection Agency as asbestos [15]. Among them, crocidolite and amosite are commonly known to cause negative health effects such as

Figure 7 shows an example of asbestos image of SEM, the elemental wt% list, and elemental

**Figure 7.** An example of asbestos image of SEM (left upper) and the elemental wt% list (right upper) and spectrum

**Figure 8.** The raw data of SEM-EDX analysis for several asbestos fibers and the processes of classifying asbestos types based on the SEM-EDX elemental wt% data are illustrated in

Figure 9 illustrates the number concentration of asbestos fiber (a sum of crocidolite (Na(Fe,Mg)3Fe2Si8O22(OH,F)2) and amosite (Fe7Si8O22(OH)2) at four-sampling sites and distri‐ bution of point sources of asbestos fiber in Fukuoka Prefecture. The highest airborne asbestos fiber concentration was recorded at site C (14.4 f L-1), followed by site A (5.9 f L-1), site D (3.4 f L-1), and site B (2.5 f L-1). The average concentration of the airborne asbestos fiber at all sites was 6.14 f L-1. This average concentration level is similar to those (7 f L-1) measured at asbestos abatement sites in Korea [16].

In 1989, the Japanese Air Pollution Control Law and related orders were revised to classify asbestos as a "specified dust" and to set up 10 fibers/liter (for including all type of asbestos) as the regulation guideline Concentration at the Boundary of the Asbestos Dusts Generation Facilities (i.e. asbestos products manufacturing facilities) [6]. Asbestos fiber concentration of site C (14.4 f L-1) is considerably higher than the regulated levels of asbestos of the Concentra‐ tion Standard at the Boundary of the Asbestos Dusts Generation Facilities.

To explore the possible reasons for the high asbestos fiber concentration at site C, the forecast atmospheric dispersion was simulated by the NOAA ARL Gaussian model. Figure 10 shows

**Figure 9.** Number concentration of asbestos fiber (crocidolite and amosite) at 4-sampling site and distribution of point sources of asbestos fiber in Fukuoka prefecture.

sampling site C in Figure 9 having a dense asbestos point source and the area distribution of the air flume started from intensively distributed companies producing asbestos related products.

According to the result of NOAA's ARL Gaussian model, the C sampling site was in an downwind position. It is therefore suggested that the relative high asbestos concentration at site C was strongly influenced by the clumped distribution of point sources of asbestos fiber. Meanwhile, although the concentration of asbestos measured at site A (heavy traffic area) was lower than the national Concentration Standard at the Boundary of the Asbestos Dusts Generation Facilities, asbestos was detected at levels around 2 times higher than those of sites B and D.

NOAA's ARL Gaussian model shown in Figure 10 indicates that the point sources of asbestos fiber situated densely near site A did not exert a direct influence on the asbestos concentration at site A. For several years, automobile parts that needed insulation from heat and friction were manufactured from dangerous asbestos, due to its excellent heat-resistant qualities. Such parts included brake linings, clutch facings, transmission components, disc brake pads, drum brake linings, and brake blocks [5].

According to both the result of NOAA's ARL Gaussian model (Figure 10) and the information on extensive use of asbestos in automobiles, the asbestos concentration at site A was probably affected by automobiles.

**Figure 10.** Map showing sampling site C in Figure 9 having dense asbestos point source and the area distribution of air flume started from the intensively distributed companies producing asbestos

### **3.3. Spatial distribution of ambient pollen**

sampling site C in Figure 9 having a dense asbestos point source and the area distribution of the air flume started from intensively distributed companies producing asbestos related

**Figure 9.** Number concentration of asbestos fiber (crocidolite and amosite) at 4-sampling site and distribution of point

According to the result of NOAA's ARL Gaussian model, the C sampling site was in an downwind position. It is therefore suggested that the relative high asbestos concentration at site C was strongly influenced by the clumped distribution of point sources of asbestos fiber. Meanwhile, although the concentration of asbestos measured at site A (heavy traffic area) was lower than the national Concentration Standard at the Boundary of the Asbestos Dusts Generation Facilities, asbestos was detected at levels around 2 times higher than those of sites

NOAA's ARL Gaussian model shown in Figure 10 indicates that the point sources of asbestos fiber situated densely near site A did not exert a direct influence on the asbestos concentration at site A. For several years, automobile parts that needed insulation from heat and friction were manufactured from dangerous asbestos, due to its excellent heat-resistant qualities. Such parts included brake linings, clutch facings, transmission components, disc brake pads, drum

According to both the result of NOAA's ARL Gaussian model (Figure 10) and the information on extensive use of asbestos in automobiles, the asbestos concentration at site A was probably

products.

214 Current Air Quality Issues

B and D.

brake linings, and brake blocks [5].

sources of asbestos fiber in Fukuoka prefecture.

affected by automobiles.

It can be considered that pollen contributes to the organic carbon fraction in totoal suspended particles (TSP). In addition, some plant types can produce pollen in huge quantities. For example, a single ragweed plant can generate a million grains of pollen a day. Therefore, though pollen granules are small and light, they can also attribute to the mass concentration of ambient aerosol particles during the main pollen season [17]. From this point of view, to study the species of pollens, their distribution and concentration can be helpful to understand their ambient behavior and public health, etc.

SEM images of several types of pollens are shown in Figure 11. Classification of pollen grains using SEM was based on morphological characteristics such as shape, size, apertures and ornamentation. The combination of these characteristics makes some pollen grains easily identifiable. However there were also pollen grains sharing several common characteristics that makes identification difficult.

Figure 12 shows the spatial variation of the number concentration of three-kind of airborne pollen at the four sites in Fukuoka Prefecture. Three-type mainly identified pollen grains are reported here. There was a noticeable spatial difference in the concentrations among three pollen types. Cedar (Cryptomeria, also called the Sugi tree in Japanese) pollen was distributed in advance of other pollens at sites A, B, and C, with the maximum concentration at site A. This cedar pollen is the most common allergen for seasonal allergic rhinitis in Japan. Cedar showed a higher concentration than other types of pollen and it was probably generated from the Sugi tree that is the most important timber tree in Japan.

**Figure 11.** Morphologies of several types of pollen grains (Cryptomeria, Pine, Alder, Cyclobalanopsis, Chamaecyparis, and Equisetum) identified by SEM observation

The sum of the number concentration of three pollen types is still high at site A. Although the atmospheric presence of pollen grains is likely to vary depending on the local kind of plant, actual weather situations, and pollinating period, the vegetation of the surface is also important [18]. Cedar pollen easily absorbs water in the atmosphere and then settles down on surface. However, they can be easily resuspended into the atmosphere by the urban surface covered with asphalt and cement. Both hot-island and building wind can also promote the resuspen‐ sion of pollen [19]. Therefore, site A, with its relatively high pollen concentrations, was probably affected by the typical urban surface and local plant.

**Figure 12.** Pollens number concentration at each sampling site

#### **4. Conclusion**

identifiable. However there were also pollen grains sharing several common characteristics

Figure 12 shows the spatial variation of the number concentration of three-kind of airborne pollen at the four sites in Fukuoka Prefecture. Three-type mainly identified pollen grains are reported here. There was a noticeable spatial difference in the concentrations among three pollen types. Cedar (Cryptomeria, also called the Sugi tree in Japanese) pollen was distributed in advance of other pollens at sites A, B, and C, with the maximum concentration at site A. This cedar pollen is the most common allergen for seasonal allergic rhinitis in Japan. Cedar showed a higher concentration than other types of pollen and it was probably generated from

Cyclobalanopsis Chamaecyparis Equisetum

**Figure 11.** Morphologies of several types of pollen grains (Cryptomeria, Pine, Alder, Cyclobalanopsis, Chamaecyparis,

The sum of the number concentration of three pollen types is still high at site A. Although the atmospheric presence of pollen grains is likely to vary depending on the local kind of plant, actual weather situations, and pollinating period, the vegetation of the surface is also important [18]. Cedar pollen easily absorbs water in the atmosphere and then settles down on surface. However, they can be easily resuspended into the atmosphere by the urban surface covered with asphalt and cement. Both hot-island and building wind can also promote the resuspen‐ sion of pollen [19]. Therefore, site A, with its relatively high pollen concentrations, was

Cryptomeria Pine Alder

that makes identification difficult.

216 Current Air Quality Issues

and Equisetum) identified by SEM observation

probably affected by the typical urban surface and local plant.

the Sugi tree that is the most important timber tree in Japan.

A better knowledge of the impact of both artificial and biological PM on urban atmosphere can help establish improved the management strategies of urban air quality. This study focused on a comprehensive and detailed interpretation for the springtime air quality influ‐ enced by both artificial (particulate matter (PM) and asbestos) and biological (pollen) sources in Fukuoka Prefecture, Japan. An intensive measurement of PM was conducted at four characteristic sites (i.e., a heavy traffic area, a residential area, an industrial area, and a desolate area) in the Fukuoka Prefecture during spring of 2007. Analysis of major ionic species in PM2.5 was performed by an Ion Chromatography, and asbestos and pollen were identified by Scanning Electron Microscopy with an energy dispersive X-ray spectrometer (EDX). PM2.5 concentration (65.3 µg m-3) measured in an industrial area (site C) was extraordinarily high compared to those monitored in other areas; it greatly exceeded the Japan's PM 2.5 criteria (a daily average of 35 µg m-3). NOAA's HYSPLIT dispersion model suggests that this high level of PM 2.5 monitored at site C is unlikely to affect the Asian continent. The ambient concentra‐ tions of PM2.5-related anions (NH4 + , NO3 - , and SO4 2-) and their relative contributions to PM2.5 were also investigated in four study areas. The concentrations of these major water-soluble ions exhibit not only strong spatial dependence but also different ratios to each other. Asbestos fiber (crocidolite and amosite) concentration values changed in the range of 2.5 to 14.4 f per liter of air. The number of pollen grains showed that Cedar ranked higher in concentration than other types of pollen, with the maximum concentration at site A.

The results of our intensive field measurement suggested that synergic biological effects induced by ambient allergenic pollen and urban fine PM in atmosphere are associated with a peculiar springtime air quality in the Fukuoka Prefecture. Our study also indicates that the specific artificial and natural sources are regionally distributed to influence local air quality and public health. We should be mindful of the fact that in order to improve the understanding of urban air quality regarding the environmental load and repercussions on human health, it will be necessary to continue the monitoring of not only particulate matters but also gaseous materials.
