Distribution Characteristics and Influencing Factors of PM2.5 in Northern-Taiwan

This paper analyzes the concentration data of PM2.5, SO2, NO2, CO and O3 recorded at 25 monitoring points in northern Taiwan in 2015. The geographical distribution and seasonal changes of PM2.5 are assessed. The geographical distribution is higher in cities than in rural areas, and higher in the west and the middle than the east. In the seasonal changes, the pollution is the most serious in spring and lightest in autumn. The author uses Pearson correlation coefficient to analyze the correlation between PM2.5 and gaseous pollutants (SO2, NO2, CO and O3). The PM2.5 concentration in Northern-Taiwan Province has a strong correlation with SO2, NO2 and CO, and a weak correlation with O3.


Introduction
In China, with the rapid development and weak regulations, the problem of atmospheric pollution has gradually become severe. Ambient fine particulate matter air pollution (PM 2.5 ), as one of the core issues of the atmospheric environment, had received international attention. Epidemiological studies have found a strong causal relationship between long-term exposure to PM 2. 5 and premature death from various diseases [1]. The 2010 Global Burden of Disease Comparative Risk Assessment (GBD) pointed out that PM 2.5 is the sixth largest overall risk factor for premature death worldwide. In 2010, global PM 2.5 caused 9-3.2 million deaths. In order to better control the problem of air pollution, this paper provides some scientific basis. When controlling PM 2.5 , it is necessary to consider the influencing factors of other pollutants. In this paper, the concentration data of PM 2.5 , SO 2 , NO 2 , CO and O 3 recorded at 25 monitoring points in north Taiwan throughout 2015 were analyzed. Based on the data, the seasonal variation of PM 2.5 was evaluated, and the Pearson correlation coefficient was used to analyze the correlation between PM 2.5 and gaseous pollutants.  5 analysis is for some large inland cities like Beijing. In this paper, the author analyzes the Northern-Taiwan area in order to provide data support for the analysis of pollutants in the southern coastal area of China and some coastal areas in Southeast Asia.

Method
For comprehensive analysis for 25 sites, two classification methods are used. One is dividing the selected area into western, middle and eastern areas, according to geographical location; and the other is to divide the area into urban and rural areas. In the process of PM 2.5 seasonal analysis, all monitoring points in each group are selected, and the average value of the monitoring concentration is taken to represent the overall concentration level of the area. What is more, Pearson coefficient and Spearman coefficient are used to analyze the correlation between the two variables PM 2.5 and the other pollutants. When describing the correlation between two variables, the Pearson correlation coefficient is usually used. The calculation formula is as follows: (1) The Spearman correlation coefficient is different from the Pearson correlation coefficient. It does not need to assume the frequency distribution of variables, does not need to assume that the relationship between variables is linear, and does not need to measure variables with interval scales [3].
The calculation process could be simplified by observing the rank difference between the two variables (denoted as equation (2)). Here, n represents the number of data, and di represents the level difference between two variables. (2) The value range of the two correlation coefficients is from -1 to 1. The greater the absolute value of the value is, the higher the correlation between the two variables is; otherwise, the correlation would be much lower [3,4].  As shown in figure 1, the annual average PM 2.5 concentration can be found to vary with geographical distribution. According to the World Health Organization (WHO) air quality guidelines levels, the annual mean value of fine particulate matter (PM 2.5 ) is recommended under 10 μg/m 3 [5]. As can be seen from figure 1, the values of all classified areas exceed the standard value. Among them, the urban index is 0.68 higher than that of the rural area, reaching 18.79 μg/m 3 .
In the geographical location classification, western area is up to 19.06 μg/m 3 , followed by the eastern area is 18.63μg/m 3 , and the middle area is the lowest which is 18.06 μg/m 3 . It can be seen that the annual average PM 2.5 concentration of coastal area is higher than the inland.  concentration in some areas is mainly affected by CO and SO 2 , which shows that CO and SO 2 have gradually become the main factors to affect the increase in PM 2.5 concentration. However, as a result, the PM 2.5 concentration in northern-Taiwan province has a strong correlation with SO 2 , NO 2 and CO, and a weak correlation with O 3 .

Discussion
From figure 1, it can be found that pollutants in northern Taiwan are concentrated in cities and the western region. In urban area, pollution is caused by traffic exhaust emissions, while the geographical gap may be caused by different industrial distributions. The seasonal distribution of PM 2.5 concentration in northern Taiwan is different from that in Beijing. The highest concentration turns up in spring, while Beijing has the highest concentration in winter. This may be caused by the fact that Beijing provides heating in winter, therefore the source of pollution is mainly coal combustion [6].
In northern Taiwan, such a high concentration of pollutants appears in the spring may be due to some external factors in addition to some local pollution factors. The reason for Taiwan's high spring pollutants is the long-distance transportation of pollutants from biomass burning in Southeast Asia which is caused by the climate [7]. In March, the westward expansion of the East Asian High and the Northwest Pacific Subtropical High merged and formed an organizational convergence center on the Indochina Peninsula and further development in April. The air pollutants emitted from the burning of biomass in Southeast Asia can be lifted and transported to the east immediately.
In terms of PM 2.5 prevention and control, the influencing factors of other pollutants should be considered at the same time. Among them, SO 2 , NO 2 and CO should be carefully controlled. Most of CO, SO 2 and NO 2 come from four sources, including power plants, industry, residential and transportation. For different pollutants, the main sources are also different. For CO, its main source is from residential usage, accounting for about 75% of the total pollutants. The two sources that contribute most to SO 2 are the industrial and power sectors, with approximately 53% and 27% of emissions respectively, as power plants use coal with higher sulfur content to generate electricity. Industrial emissions are the largest source of NO 2 emissions, accounting for about 42%. What is more, transportation cannot be underestimated. It is the fastest growing source, increased by 62% from 1990 to 1995, which is also important to control the pollutant emissions [8].
In terms of measures to reduce pollutant emissions, the methods to reduce SO 2 emission could include using clean energy as much as possible with electricity generation, and using coal with lower sulfur content by carrying out coal washing operations when generating electricity with coal. There are also some measures to reduce NO 2 emissions. For industrial aspects, different absorption and conversion treatments are used according to the exhaust gas of different industries. In terms of transportation, the annual inspection mechanism of automobiles shall be strengthened to install more effective catalysts for automobiles. For reducing CO emissions in agriculture, the crop residues could be used for plow instead of burning them.
From the above seasonal analysis, it can be seen that not only local factors cause pollution, but also external pollution migrates with the wind. Therefore, in order to control air pollution, it is necessary to control local sources of pollution as well as work with other surrounding regions to jointly control it.

Conclusion
The geographical distribution is higher in cities than in rural areas, and higher in the west and the middle than the east. In the seasonal changes, the pollution is most serious in spring and lightest in autumn. Pearson correlation coefficient is used in this paper to analyze the correlation between PM 2.5 and gaseous pollutants (SO 2 , NO 2 , CO and O 3 ). Overall, the PM 2.5 concentration in northern-Taiwan province has a strong correlation with SO 2 , NO 2 and CO, and a weak correlation with O 3 . However, the data and methods in this article cannot analyze the source of pollutants, and it is hoped that it can be analyzed in the subsequent research.