Since 1991, air pollution has gained special attention in Taiwan after a petrochemical complex was constructed in Mailiao Township, Yunlin County. We explored the association between the magnitude of PM 2.5 and meteorological factors during 2012-2016. Our findings revealed that 1) mean PM 2.5 levels gradually decreased from 30.70 μg/m 3 in 2013 to 25.36 μg/m 3 in 2016; 2) wind speed is the main determinant of air quality—air quality significantly improved when it was faster than 4 m/sec; and 3) wind direction is another determinant of air quality—when the wind direction was southerly, air quality improved. Elevated PM 2.5 levels were defined as those hourly levels higher than the third quartile (36 μg/m 3). The significantly negative predictive factors for elevated PM 2.5 levels were the summer or autumn seasons, rainfall, increased wind speed, and wind direction from 150 ° to 230 ° from the north. The significantly positive predictive factors for elevated PM 2.5 levels were working hours from 6 a.m. to 2 p.m., a temperature between 11 °C and 25 °C, relative humidity between 40% and 68%, and wind direction (e.g., northerly wind, northeasterly wind, and easterly wind). The predictive formula is attached in the Appendix. Therefore, people should protect themselves on these high-risk days.
Air pollution is a major health concern and has been extensively studied. Exposure to pollutants such as fine particulate matter (PM2.5) and ozone has been associated with an increase in mortality and hospital admissions due to respiratory and cardiovascular diseases [
Located in Yunlin County in western Taiwan, Mailiao Township has a world-class petrochemical industry zone and a coal-fired power plant. Within the township, the complex (area around 26 square kilometers) is located in the north-western coastal area, and an air monitoring station is located 6 km away from the complex in the populated south-east area of Mailiao. This chief industry includes oyster oil plants, ethylene light oil pyrolysis plants and their related petrochemical plants, heavy machinery plants, steam power plants, Mailiao Industrial Port, and the third largest coal-fired power plant in Taiwan. Furthermore, there are many small industrial areas in Yunlin County.
Another study reported that ambient PM2.5 pollution is a major mortality risk factor in Taiwan. The same study also reported that substantial geographic variations in PM2.5 attributable to the mortality fraction were found, and Yunlin County had the highest percentage (21.8%) of deaths attributable to PM2.5 [
Air quality is severely influenced by weather conditions. A 19-year observational study reported that 14,700 excess deaths from PM2.5 were attributable to weather-related increases in air quality in the United States [
In this study, we explored the effects of meteorological variables, such as temperature, wind speed and direction, RH, and daily rainfall on elevated PM2.5 levels (higher than the third quartile) of all hourly records at the air monitoring station in Mailiao during 2012-2016. The novelty and contributions of this study were meteorological factors such as wind speed and direction, and daily rainfall on the effect of elevated PM2.5 levels near the huge petrochemical complex.
First, the hourly mean data for temperature, wind speed and direction, RH, rainfall, and PM2.5 from the air quality monitoring station were downloaded from the website of the Environmental Protection Administration; these data were open for public use. Second, we downloaded data from the nine air quality monitoring stations within 50 km for hourly spatial changes [
Concentration of PM2.5 (μg/m3): Since 2012, hourly mean concentrations have been continually measured at the monitoring station using a β-ray attenuation method for PM2.5. The instrument used for PM2.5 analysis was VEREWA F701 [
Wind speed and direction: The hourly mean wind speed was measured using cup anemometers (Model 014A; Met One Instruments, Inc., Grants Pass, USA), and the mean wind direction was measured using a wind vane (Model 024A; Met One Instruments, Inc., Grants Pass, USA) [
Seasons: We defined seasons as spring (February to April), summer (May to July), autumn (August to October), and winter (November to January).
The study protocol was reviewed and approved by the Research Ethics Committee of Buddhist Dalin Tzu Chi Hospital, Taiwan (No. B10601004).
All statistical operations were performed using the R 3.0.2 software (R Foundation for Statistical Computing, Vienna, Austria). The openair, sp, and gstat packages of R were applied to analyze the data.
Continuous variables were examined using analysis of variance and are expressed as mean ± standard deviation. Categorical variables were examined using a trend test and are presented as the frequency and percentage. The mean wind speed and direction were calculated using a vector method for days, weeks, and months.
The multiple logistic regression model with a stepwise variable selection procedure was conducted to determine the vital predictors of PM2.5 level above the third quartile (36 μg/m3). The quartile method was used in other study (Rumchev et al., 2018). Furthermore, generalized additive models were fitted to detect the potential nonlinear effects of continuous covariates and determine the appropriate cutoff points for discretizing the covariates, if necessary, during stepwise variable selection.
We assessed the goodness of fit of the final logistic regression model according to the estimated area under the receiver operating characteristic curve (AUC). Statistical tools of regression diagnostics were applied to discover any problems associated with the regression model or data. Two-sided p ≤ 0.05 was considered statistically significant. Hourly mean data of the nine air monitoring stations within 50 km were downloaded and stratified by seasons. The Kriging method was applied to estimate the spatial data to explore changes by hour.
In total, 43,847 hourly records from 2012 to 2016 from the air quality monitoring station in Mailiao were used; 42,499 (96.9%) records were enrolled for analysis after excluding the records with missing data. The third quartile for PM2.5 was set at 36 μg/m3, which was defined as elevated PM2.5 levels in this study. The mean PM2.5 levels at this station had a lower trend according to Theil–Sen analysis (p < 0.001) (
The daily means of PM2.5 were higher at the end of the year and beginning of the following year and lower during the summer (
| Year | PM2.5 (μg/m3) | ||||
|---|---|---|---|---|---|
| Mean | Standard deviation (SD) | Minimum | Maximum | Median | |
| 2012 | 25.47 | 17.39 | 0 | 132 | 22 |
| 2013 | 30.70 | 20.25 | 0 | 179 | 27 |
| 2014 | 27.75 | 20.54 | 1 | 145 | 24 |
| 2015 | 24.44 | 19.83 | 0 | 145 | 20 |
| 2016 | 25.36 | 16.73 | 3 | 153 | 21 |
The data were stratified and compared by season and daytime or nighttime. The means of PM2.5 were significantly lower during summer as well as daytime and nighttime (p < 0.001) than during other seasons. Spring and summer are the rainy seasons in Taiwan (from 7.0% to 8.6%) (p < 0.001). The hottest season is summer (28.7˚C ± 2.1˚C), followed by autumn (25.7˚C ± 3.1˚C) and the coldest season is winter (18.0˚C ± 3.0˚C) (p < 0.001). The percentage of higher PM2.5 levels was lowest in the summer. The mean RH (77% - 87%) is the highest during summer. Regarding wind direction, the mean wind directions were southerly or southeasterly during summer. During autumn and winter, the mean wind directions were easterly to northeasterly during both daytime and nighttime. During spring, the mean wind direction was southerly during the daytime and easterly during the nighttime. The highest mean wind speed was noted at nighttime during winter (1.81 m/s), and the lowest mean wind speed was noted in the daytime during spring (0.37 m/s) (
The significant factors associated with elevated PM2.5 levels included work hours from 6 a.m. to 2 p.m. [6 a.m.: odds ratio (OR): 1.20, 95% confidence interval (CI): 1.06 - 1.35; 7 - 9 a.m., OR: 1.66, 95% CI: 1.56 - 1.78; 11 a.m. to 2 p.m.:
| Variables | Total | Spring | Summer | Autumn | Winter | p value |
|---|---|---|---|---|---|---|
| PM2.5, mean ± SD (μg/m3) | ||||||
| whole day (24 hours) | 26.7 ± 19.2 | 30.7 ± 19.3 | 12.7 ± 10.0 | 29.7 ± 17.5 | 33.7 ± 20.5 | <0.001 |
| Daytime | 27.5 ± 19.4 | 30.6 ± 19.5 | 13.0 ± 10.0 | 31.8 ± 17.4 | 34.6 ± 20.7 | <0.001 |
| Nighttime | 25.9 ± 18.9 | 30.8 ± 19.2 | 12.5 ± 10.0 | 27.5 ± 17.2 | 32.8 ± 20.3 | <0.001 |
| PM2.5, hourly records > Q3 | ||||||
| whole day (24 hours) | 10,580 (24.9%) | 3598 (33.7%) | 318 (3.0%) | 2903 (27.3%) | 3761 (35.2%) | <0.001 |
| Daytime | 5586 (13.1%) | 1739 (16.3%) | 166 (1.6%) | 1705 (16.0%) | 1976 (18.5%) | <0.001 |
| Nighttime | 4994 (11.8%) | 1859 (17.4%) | 152 (1.4%) | 1198 (11.3%) | 1785 (16.7%) | <0.001 |
| Rainfall (per hour records), n (%) | ||||||
| whole day (24 hours) | 2205/42,499 (5.2) | 763/10,665 (7.2) | 833/10,514 (7.9) | 257/10,633 (2.4) | 352/10,687 (3.3) | <0.001 |
| Daytime | 1145/21,028 (5.4) | 384/5272 (7.3) | 449/5204 (8.6) | 136/5262 (2.6) | 176/5290 (3.3) | <0.001 |
| Nighttime | 1060/21,471 (4.9) | 379/5393 (7.0) | 384/5310 (7.2) | 121/5371 (2.3) | 176/5397 (3.3) | <0.001 |
| Temperature, mean ± SD (degree C) | ||||||
| whole day (24 hours) | 23.9 ± 5.0 | 23.2 ± 4.0 | 28.7 ± 2.1 | 25.7 ± 3.1 | 18.0 ± 3.0 | <0.001 |
| Daytime | 24.8 ± 5.2 | 24.1 ± 4.2 | 29.7 ± 2.2 | 26.7 ± 3.1 | 18.8 ± 3.1 | <0.001 |
| Nighttime | 23.0 ± 4.7 | 22.3 ± 3.7 | 27.7 ± 1.5 | 24.8 ± 2.8 | 17.4 ± 2.8 | <0.001 |
| Relative humidity, mean ± SD (%) | ||||||
| whole day (24 hours) | 81.4 ± 10.1 | 82.0 ± 10.6 | 82.8 ± 8.7 | 80.5 ± 10.0 | 80.2 ± 10.6 | <0.001 |
| Daytime | 77.2 ± 10.3 | 77.4 ± 11.3 | 78.0 ± 9.0 | 76.6 ± 9.9 | 77.0 ± 10.9 | <0.001 |
| Nighttime | 85.4 ± 7.9 | 86.5 ± 7.5 | 87.4 ± 5.3 | 84.4 ± 8.5 | 83.3 ± 9.2 | <0.001 |
| Wind direction, mean (degree) | ||||||
| Whole day | 100.2 | 126.8 | 203.7 | 71.4 | 67.8 | <0.001 |
| Daytime | 122.5 | 184.3 | 220.4 | 76.4 | 72.4 | <0.001 |
| Nighttime | 88.9 | 94.6 | 178.6 | 67.5 | 63.4 | <0.001 |
| Wind speed, mean (m/second) | ||||||
| Whole day | 0.58 | 0.34 | 1.30 | 0.92 | 1.79 | <0.001 |
| Daytime | 0.42 | 0.37 | 1.68 | 0.83 | 1.77 | <0.001 |
| Nighttime | 0.79 | 0.57 | 1.11 | 1.02 | 1.81 | <0.001 |
Daytime: 6 am to 5 pm; nighttime: 6 pm to next day 5 am. Q3: the third quartile of PM2.5 (36 μg/m3).
OR: 1.32, 95% CI: 1.22 - 1.42], a temperature between 11˚C - 25˚C (OR: 2.22, 95% CI: 2.07 - 2.38), and RH between 40% - 68% (OR: 0.42, 95% CI: 0.36 - 0.49). Regarding wind direction, northerly winds (<30˚ and > 330˚ from north, OR: 1.57, 95% CI: 1.43 - 1.74) and northeasterly winds (30˚ - 60˚ from the north, OR: 1.23, 95% CI: 1.16 - 1.30) were associated with higher probabilities of elevated PM2.5 levels. Easterly winds (60˚ - 90˚ from the north) with a wind speed between 4 and 6 m/s (OR: 4.57, 95% CI: 2.50 - 8.24) and easterly to southeasterly winds (90˚ - 120˚ from the north) with a wind speed between 2 and 4 m/s (OR: 1.45, 95% CI: 1.22 - 1.73) were associated with higher probabilities of elevated PM2.5 levels (
The relationships between the air quality monitoring stations are shown in
| Variable | Estimate | O.R. (95% C.I.) | p value |
|---|---|---|---|
| Summer | −2.02 | 0.13 (0.12 - 0.15) | <0.001 |
| Autumn | −0.21 | 0.81 (0.76 - 0.86) | <0.001 |
| Hour on 6 o’clock | 0.18 | 1.20(1.06 - 1.35) | 0.003 |
| Hour on 7, 8, 9, 10 o’clock | 0.51 | 1.66 (1.56 - 1.78) | <0.001 |
| Hour on 11, 12, 13, 14 o’clock | 0.27 | 1.32 (1.22 - 1.42) | <0.001 |
| Temperature (11 - 25 degrees C) | 0.80 | 2.22 (2.07 - 2.38) | <0.001 |
| Relative humidity between 40% - 68% (yes vs. no) | 0.90 | 2.46 (2.28 - 2.65) | <0.001 |
| Rainfall (yes vs. no) | −0.86 | 0.42 (0.36 - 0.49) | <0.001 |
| Wind speed (per m/s) | −0.46 | 0.63 (0.61 - 0.65) | <0.001 |
| North wind (<30, >330 degree) | 0.45 | 1.57 (1.43 - 1.74) | <0.001 |
| North-east wind (30 - 60 degree) | 0.21 | 1.23 (1.16 - 1.30) | <0.001 |
| East wind (60 - 90 degree) × wind speed (4 - 6 m/s) | 1.52 | 4.57 (2.50 - 8.24) | <0.001 |
| East wind (90 - 120 degree) × wind speed (2 - 4 m/s) | 0.37 | 1.45 (1.22 - 1.73) | <0.001 |
| South west wind (150 - 230 degree) × wind speed (per m/s) | −0.20 | 0.82 (0.78 - 0.86) | <0.001 |
| Intercept | −0.82 |
Yunlin county, and higher wind speeds and southerly wind brought the lower levels of PM2.5 to Yunlin county. The adjusted generalized R squared (Nagelkerke’s R squared) value of this final model was 0.272, and the AUC (78.3%) was acceptable (
The novel finding of this study was that the annual mean PM2.5 concentration had significantly negative trend from 30.70 μg/m3 in 2013 to 25.36 μg/m3 in 2016. The contributions of this study were wind speed and direction, and daily rainfall on the effect of elevated PM2.5 levels near the huge petrochemical complex. Daytime (from 6 a.m. to 2 p.m.) was the significant factor associated with hourly PM2.5 concentrations higher than the third quartile (36 μg/m3). Another key proposed approach was the hourly spatial data of the surrounding nine monitoring stations within 50 km were stratified by seasons and analyzed using the Kriging method. Another key finding was that the significantly positive meteorological factors associated with hourly levels of PM2.5 higher than the third quartile (36 μg/m3) were cool weather (temperature between 11˚C - 25˚C or 51.8˚F - 77˚F), RH of 40% - 68%, northerly winds (330˚ - 360˚ and 0˚ - 30˚ from the north), northeasterly winds (30˚ - 60˚ from the north), easterly winds (60˚ - 90˚ from the north) with a wind speed of 4 - 6 m/s, and easterly winds (90˚ - 120˚ from the north) with a wind speed of 2 - 4 m/s. Furthermore, the significantly negative meteorological factors associated with higher PM2.5 levels were summer and autumn, daily rainfall, wind speed, and southerly winds (150˚ - 230˚ from the north) with a wind speed of m/s. These findings were further discussed in the below sections.
After decades of industrialization, air pollution has become a major environmental problem in Taiwan. Poor air quality has both acute and chronic effects on human health. In 2012, the Air Pollution Control Act (APCA) was formulated to control air pollution, maintain public health and living environments,
and improve the quality of life in Taiwan [
In this study, we observed a diurnal variation in PM2.5 with elevated levels during the daytime. This result differs from another study that reported that PM2.5 levels were higher in concentration during the nighttime and lower during the daytime in Beijing [
A relevant study reported that wind direction is a critical parameter affecting PM2.5 levels [
A study reported that elevated PM2.5 levels occurred during the winter and spring [
In the current study, we found that cool weather (11˚C - 25˚C) was associated with higher PM2.5 levels, compared with temperatures lower than 11˚C and higher than 25˚C, an inverted U-shaped effect. One study reported that temperature was positively correlated with PM2.5 concentration in four seasons in Nagasaki, Japan (Wang and Ogawa, 2015) [
In the current study, we found that RH between 40% and 68% was a positive factor for PM2.5 levels above the third quartile. A similar finding was reported in that RH had an inverted U-shaped relationship with PM2.5 concentration (peaking at an RH of 45% - 70%) [
A study reported that rainfall had washout effects on atmospheric particulate pollution and was recognized as one of the main mechanisms for reducing PM2.5 pollution [
Our study has some limitations. First, the data about the original sources of PM2.5 included atmospheric condition, emission, or mobile-source influence with multi-resources data such as pollution source information, real-time population grid data, meteorological data, and traffic data were not available in this study, and that is a limitation. Second, we find some phenomena that cannot be easily explained by atmospheric condition, emission, or mobile-source influence. For example, PM2.5 concentrations start to rise from 6:00 a.m. until 2:00 p.m. Therefore, further studies should be carried out with multi-resources data, such as pollution source information, real-time population grid data, meteorological data and traffic data, to provide reasonable interpretations for these unexplainable phenomena. Third, the result cannot apply to other places, however we provided a formula based on the final multiple logistic regression model in the Appendix to calculate the probability of elevated PM2.5 level.
Furthermore, we found that meteorological factors were associated with elevated PM2.5 levels, and provided a formula for calculating the probability of elevated PM2.5 (>36 μg/m3) levels at this key monitoring station. Similar functions based on our method could be established to calculate the probability of elevated PM2.5 levels in other cities. Further studies should investigate multi-resources data, such as PM2.5 source information, real-time population data, and traffic data, which may provide deeper interpretations for air pollution prediction.
The annual PM2.5 concentration gradually decreased from 30.70 μg/m3 in 2013 to 25.36 μg/m3 in 2016. The meteorological conditions have important effect on PM2.5 mass concentration. We found that the factors associated with elevated PM2.5 levels were daytime, cool weather (11˚C - 25˚C), winter and spring, RH between 40% - 68%, lower wind speeds, wind direction, and days with no rain. Furthermore, by the relationship with meteorological conditions, it can be depicted that daytime with cool weather, no rain, relatively dry (RH = 40% - 68%), lower wind speeds (breeze), northerly wind, and in winter and spring brought the most pollutants to Yunlin county; and daytime in summer, with rain, relatively humidity (RH > 68%), higher wind speeds, and southerly wind brought the lower probabilities of pollution to Yunlin county. If people want to know the more accurate probability of elevated PM2.5 level (>36 μg/m3), the predictive formula is attached in the Appendix. Therefore, people should protect themselves on these high-risk days.
JK Chiang received research grants from Buddhist Dalin Tzu Chi Hospital (DTCRD106(2)-E-04). This manuscript was edited by Wallace Academic Editing. Yee-Hsin Kao and Chih-Wen Lin contributed equally.
JKC and YHK designed, conducted and drafted the manuscript. JKC and CWL analyzed the data. All authors contributed to the manuscript, revised drafts critically for important intellectual content, and read and approved the final manuscript.
The authors declare no conflict of interest.
Kao, Y.-H., Lin, C.-W. and Chiang, J.-K. (2019) Predictive Meteorological Factors for Elevated PM2.5 Levels at an Air Monitoring Station Near a Petrochemical Complex in Yunlin County, Taiwan. Open Journal of Air Pollution, 8, 1-17. https://doi.org/10.4236/ojap.2019.81001
Programming code in OpenOfficeCalc, Microsoft Excel, and R environment for calculating the probability of elevated PM2.5 (>36 μg/m3) based on our multiple logistic regression model.
1) In OpenOfficeCalc or Microsoft Excel:
Key in the values for summer (yes = 1, no = 0) in the A1 cell, autumn (yes = 1, no = 0) in the A2 cell, 6 o’clock (yes = 1, no = 0) in the A3 cell, 7 - 10 o’clock (yes = 1, no = 0) in the A4 cell, 11 - 14 o’clock (yes = 1, no = 0) in the A5 cell, temperature 11 - 25 degree C (yes = 1, no = 0) in the A6 cell, relative humidity 40% - 68% (yes = 1, no = 0) in the A7 cell, rainfall (yes = 1, no = 0) in the A8 cell, wind speed (m/s) in the A9 cell, wind direction, <30 degree, >330 degree from north, (yes = 1, no = 0) in the A10 cell, wind direction, 30 - 60 degree from north, (yes = 1, no = 0) in the A11 cell, wind direction, 60 - 90 degree from north, (yes = 1, no = 0) in the A12 cell, wind speed 4 - 6 m/s, (yes = 1, no = 0) in the A13 cell, wind direction, 90 - 120 degree from north, (yes = 1, no = 0) in the A14 cell, wind speed 2 - 4 m/s, (yes = 1, no = 0) in the A15 cell, wind direction 150 - 230 degree from north, (yes = 1, no = 0) in the A16 cell, and wind speed, m/s in the A17 cell.
Key in the following formula in any empty cell on the same spreadsheet to obtain the estimated probability of elevated PM2.5 (>36 μg/m3):
= 1/(EXP(−(−0.82 + (−2.02)*A1 + (−0.21)*A2 + 0.18*A3 + 0.51*A4 + 0.27*A5 + 0.80*A6 + 0.90*A7 + (−0.86)*A8 +(−0.46)*A9 + 0.45*A10 + 0.21*A11 + 1.52*A12*A13 + 0.37*A14*A15 + (−0.20)*A16*A17))+1)
2) In an R environment:
To calculate the probability of elevated PM2.5 (>36 μg/m3), substitute the values for the variables X1 to X17 in the following regression equation and execute in the R console:
yhat<- (-0.82 # constant
+ (-2.02)*X1 # X1= summer (yes=1, no=0)
+ (-0.21)*X2 # X2 = autumn, (yes=1, no=0)
+ 0.18*X3 # X3 = 6 o’clock, (yes=1, no=0)
+ 0.51*X4 # X4 = 7-10 o’clock, (yes=1, no=0)
+ 0.27*X5 # X5 = 11-14 o’clock, (yes=1, no=0)
+ 0.80*X6 # X6 = temperature 11-25 degree C, (yes=1, no=0)
+ 0.90*X7 # X7 = relative humidity 40-68%, (yes=1, no=0)
+ (-0.86)*X8 # X8 = rainfall, (yes=1, no=0)
+ (-0.46)*X9 # X9 = wind speed, m/s
+ 0.45*X10 # X10 = wind direction, <30 degree, >330 degree from north, (yes=1, no=0)
+ 0.21*X11 # X11 = wind direction, 30-60 degree from north, (yes=1, no=0)
+ 1.52*X12 *X13
# X12 = wind direction, 60-90 degree from north, (yes=1, no=0)
# X13= wind speed 4-6 m/s, (yes=1, no=0)
+ 0.37*X14 *X15
# X14 = wind direction, 90-120 degree from north, (yes=1, no=0)
# X15= wind speed 2-4 m/s, (yes=1, no=0)
+ (-0.20)*X16*X17
# X16 = wind direction 150-230 degree from north, (yes=1, no=0)
# X17 = wind speed, m/s)
phat <- 1/(1 + exp(-(yhat)))
phat