Air pollution has been considered as the world’s biggest environmental health risk and the major cause of environmentally related deaths  . The health impacts of air pollution are serious. The World Health Organization (WHO) found that the majority of air pollution related deaths are due to increased risks of stroke, heart disease, lung cancer, acute respiratory infections and chronic obstructive pulmonary diseases  . WHO reported a total of around 7 million air pollution related premature deaths in 2012. Low and middle income countries in the South-East Asia and Western Pacific Regions are the worst affected countries .
Vietnam is among the top countries with the worst air pollution in the world, according to the annual Environmental Performance Index (EPI) report . The air pollution was reported to reach unhealthy levels in major cities, especially the Capital city of Hanoi and Ho Chi Minh City . It was pointed out that increased number of vehicles might have been the main contribution to the increased concentration levels . Increased PM2.5 concentration level was thought to associate with the rise in the numbers of construction projects dust, vehicle emissions, smoke from manufacturing plants and factories     . Another reason is the problem of biomass burning in surrounding suburban area, especially dried rice stem burning during rice harvest seasons. Biomass burning events may have caused the rise in PM2.5 concentrations at some point in time   .
Once the air pollutants are released, they are either suspended or dispersed in the air. Previous studies found seasonal, weekly and diurnal patterns of PM2.5 concentration   . The concentration levels were found to be lower in summer and higher in winter    . Ly et al. (2018) noted that June, 2016 and July 2017 had lower concentration levels as compared to December 2016. The PM2.5 concentration was also found higher around midnight than other times of the days .
Meteorological factors were considered to play an important role in increasing or decreasing PM2.5 concentration. Although the impacts of between meteorological factors on the PM2.5 levels have been studied in several countries, especially in China  - , there has been little research conducted to determine the impacts of meteorological factors on the atmospheric PM2.5 levels in Hanoi, Vietnam. Hien et al., (2002) pointed out that meteorological contributed to 60% - 70% of day-to-day variations of PM2.5 levels during the moonsoon season. Thuy et al., (2018) indicated the relationship between wind speed, wind direction, temperature, relative humidity, surface pressure, radiation and precipitation and mass concentration of PM0.1, PM2.5 and PM10 during the rice straw opens burning episode in 2015. However, these studies were conducted in a short period and focused on specific events during a year. The long-term effects of meteorological factors on the atmospheric PM2.5 levels were not examined. The level of contribution of each meteorological factor to the variation of PM2.5 concentration has not been explained.
Our research further studies the temporal patterns of PM2.5 concentration; examines the relationship between meteorological factors and the variation of PM2.5 concentration levels in Hanoi, Vietnam; and investigates the influence of these factors on the diurnal, monthly and seasonal variation of PM2.5 concentration. The study seeks to provide governmental authorities with valuable information in order to develop early warning systems and to minimize the negative impacts of air pollution on public health.
2. Materials and Methods
Data used in this study was collected from the automatic monitoring stations (HORIBA) located on 556 Nguyen Van Cu street, Long Bien district, Hanoi of the Center for Environmental Monitoring (CEM), Vietnam Environment Administration (VEA). This is the only one out of seven stations providing real-time data with public access via http://enviinfo.cem.gov.vn/ . This monitoring station aims to monitor air quality on Nguyen Van Cu street, one of the busiest streets and the entrance to the Hanoi center with heavy traffic flow every day. The monitoring data is updated hourly. Data includes hourly measurement of PM2.5 (μg/m3), wind speed (m/s), temperature (˚C), air pressure (hPa) and solar radiation (W/m2). This is time series of hourly data collected from January 01, 2017 to December 31, 2018.
Regression models are used to examine temporal patterns of PM2.5 variation and identify factors that affect these patterns. The histogram by transformation was performed to determine appropriate regression models. The log transformation is the most suitable as it forms normal distribution and thus increases validity of the statistical analyses. This is a well-known method to deal with skewed data  .
The PM2.5 variable is transformed using logarithmic transformation and used as the dependent variables in the Loglinear regression models. The Loglinear regression model is presented in the Equation (1):
where Y is the PM2.5 concentration and is the set of explanatory variables. The terms and are constant, coefficients and error terms. The explanation for the relationship between response variable Y and an explanatory variable x is based on the elasticity or marginal effects estimated directly form the loglinear model.
If an explanatory variable is a continuous variable, an elasticity presents the percentage change in the response variable Y due to 1% change in that explanatory variable and a marginal effect measures change in the response variable Y due to a unit change in the explanatory variable x. Equations (2) and (3) show the calculation of elasticity and marginal effect for continuous variables.
In the case the explanatory variable is a dummy variable, an elasticity and a marginal effect respectively measure the percentage and absolute difference in PM2.5 concentration level between subgroups of the explanatory variable x (see calculation in Equations (4) and (5)).
All the statistical and regression models are performed by using the Stata software version 12 (StataCorp LP., College Station, TX, USA). The calculation of elasticity and marginal effect are performed by using Microsoft Excel 365 (Microsoft, Redmond, Washinton, USA).
3. Results and Discussions
3.1. Yearly, Seasonal, Monthly and Diurnal PM2.5 Variations
The regression results as presented in Table 1 suggest that there exist yearly, seasonal and monthly variations. The analytical results in Table 1 suggest that there was statistically different in PM2.5 concentration between 2017 and 2018. The mean PM2.5 was found 14.52% or 3.38 μg/m3 lower in 2017 as compared to 2018. Given the stable settlement with little construction activities in Nguyen Van Cu street, this increasing trend is expected to be caused by increased number of vehicles on this street over the years as this is one of the main corridors entering the Hanoi center. Cohen et al. (2010) pointed out that automobiles and transport are major contributors to PM2.5 concentration in Hanoi.
Table 1. Yearly, seasonal and monthly variations of PM2.5 concentration level.
Note: * p < 0.05, ** p < 0.01, *** p < 0.001.
The concentration level in Spring, Summer and Autumn is statistically differeent from the concentration level in Winter. Negative coefficients indicate that the concentration levels in these seasons are statistically lower than in Winter. This means that the concentration level is reported highest in Winter. The elasticity calculation indicates that summer has the lowest concentration level and is measured at 36.07% lower than Winter. It is followed by Spring and Autumn with the mean concentration level at 23.53% and 15.75% lower, respectively. This result agrees with observations from other studies  . Ly et al. (2018) noted that there was seasonal variation in PM2.5 concentration with lower concentration found in summer by observing 2017 data monitored at the University of Science and Technology, Hanoi, Vietnam . The study by Lasko et al. (2018) found severe air pollution in Winter period, during December 2013 and January 2014 .
The monthly analysis of PM2.5 concentration shows that except December, all other months have statistically lower concentration levels as compared to January. This means that January and December are the months with highest PM2.5 level. The concentration is found to be the lowest in May with 58.36% lower than in January. It is followed by April, July, June and February with 45.64%, 32.28%, 31.19% and 30.33% lower, respectively. The analyses on seasonal and monthly variations show the similar trend when Summer which includes June, July and August has lowest PM2.5 level, followed by Spring which is characterized by March, April and May and then Autumn which contains September, October and November. These seasonal and monthly variations were also confirmed in studies in different areas in South East Asia and East Asia      .
The diurnal analysis suggests the variation of PM2.5 concentration across different time of the day. Figure 1 indicates the concentration trend of PM2.5 level from 0:00 to 11:00 PM. Higher concentration is found early in the morning from 0:00 AM till 8:00 AM before starting to decline until it reaches the lowest peak around 2:00 PM. The concentration then starts to increase again.
Figure 1. Diurnal variation of PM2.5 concentration levels.
By comparing the concentration level between different hours of the day and the level at 0:00 AM, the regression result in Table 2 show that: 1) concentration at the time from 8:00 PM-11:00 PM, at 1:00 AM and from 9:00 AM-10:00 AM are not statistically significant difference from the concentration at 0:00 AM; 2) concentrations between 2:00 AM-8:00 AM tends to be higher than at 0:00 AM; and 3) concentrations between 11:00 AM and 6:00 PM appears to be lower than at 0:00 AM.
Taking the sample mean concentration of PM2.5 at 0:00 AM as a benchmark, Figure 2 shows the higher PM2.5 concentration occurs between 2:00 AM and 8:00 AM and reaches the highest levels at 5:00 AM-6: 00 AM with measurements of 13.05% and 13.61% higher than the measurements at 0:00 AM. The calculation of marginal effects indicates that mean PM2.5 at 6:00 AM is 3.65 μg/m3 higher than at 0:00 AM. The downtrend is recorded from 11:00 AM and reaches the lowest levels of the day at 2:00 PM-3:00 PM with 34.97% and 34.64% lower figures than at 0:00 AM. The lowest level of PM2.5 is 11.28 μg/m3 less than it is at 0:00 AM. Although the concentration level from 4:00 PM-7:00 PM is still lower
Table 2. Diurnal variations of PM2.5 concentration level.
Note: * p < 0.05, ** p < 0.01, *** p < 0.001.
Figure 2. Statistical variations of PM2.5 concentrations at different hours as compared to at 0:00 am.
at 0:00 AM, it tends to increase gradually and is only 9.01% higher at 7:00 PM than at 0:00 AM. The concentration level was recorded with no statistical difference from 8:00 PM-1:00 AM.
The above analysis provided interesting information into the temporal patterns of PM2.5 concentration levels over time. It was pointed out that the increasing number of vehicles might have been the main contribution to the increased concentration levels, accounting for 40% ± 10% . Other sources included windblown soil windblown soil (3.4 ± 2)%, secondary sulfates (7.8 ± 10)%, smoke from biomass burning (13 ± 6)%, ferrous and cement industries (19 ± 8)%, and coal combustion (17 ± 7)% . This study result is consistent with the study findings by Hai and Oanh (2013) . They found peaks of PM mass in Hanoi during morning and evening rush hours by analyzing 4 h-samples. The rush hours start from 6:00 AM to 8:00 AM and from 5:00 PM to 7:00 PM that sees more vehicles on roads which cause traffic congestion.
The variation is also affected by long-range transport of particulate matter from other geographical areas due to rainfall and wind direction . Heavy duty trucks transporting construction wastes and materials may also cause serious problems. Tran and Yanagida (2019) found operations of heavy-duty trucks generate higher dust emission and considered as a major stationary source of fugitive dust pollution . Although there is unclear how emissions from cement plants, coal power plants, industrial zones and handicraft villages, etc. located within or surrounding Hanoi areas affect air quality in Hanoi. These activities are releasing paricles into the air every day. Meteorological factors have been found to play an essential role in affecting PM2.5 variation
3.2. Correlations between PM2.5 Concentrations and Meteorological Factors
The log-linear regression model is performed to examine the relationship between PM2.5 concentration level and meteorological factors, including air pressure, wind speed, temperature and solar radiation using hourly data of 2017 and 2018.
The regression results show statistical relationship between all the meteorological factors and PM2.5 concentration level (Table 3). These factors, however, have different effects on the trend and magnitude of concentration. The results indicate positive relationship between air pressure and the concentration level, meaning that when air pressure increases, PM2.5 tends to increase. Other three factors, including windspeed, temperature and solar radiation, show negative trend. This means PM2.5 level increases when wind speed, temperature and solar radiation decrease and vice versa. The negative association between wind speed and PM2.5 concentration levels was also found in the study by Thuy et al. (2018) . It was pointed out that strong wind contributed to the dilution of pollution and therefore it reduced PM2.5 concentration in the air.
The turbulence of air as the result of interactions between these factors directly affects dispersion of air pollutants into the atmospheric boundary layer horizontally and vertically and become mixed with this layer . This refers to mixing layer height which is also one of the key characterizing air polltution together with emission source, meteorological influences  . Low wind speed and low mixing layer height were considered as the dominant factors contributing to high concentration of particulate matter in the air . As noted by Haagenson (1979), mixing depth, ventilation, and the degree of stability significantly contribute to the daily variation of particles. Of which, pollutant diversion mostly depends on mixing depth and ventilation which are largely affected by meteorological conditions . Week turbulence and radiative cooling at night lead contribute to the formation of stable boundary layer near the ground which results in higher PM2.5 concentration   .
The results of elasticity and marginal effects estimates measured at the sample mean indicate how the meteorological factors affect the levels of PM2.5 concentration. The elasticity results suggest that the PM2.5 level will increase by 6.16% if air pressure increases by 1% (other factors remain constant). Also, for each 1% increase in one of the factors, wind speed, temperature and solar radiation (other factors remain constant), PM2.5 concentration will decrease by 0.39%, 0.17% and 0.06%, respectively.
Table 3. Diurnal variations of PM2.5 concentration level.
Note: * p < 0.05, ** p < 0.01, *** p < 0.001.
The marginal effects indicate that each unit increase in air pressure would result in 0.18 μg/m3 increase of PM2.5 (other factors held constant). In case the air pressure increases from the mean value of 1010.27 to the max value of 1032.10, the PM2.5 level goes up by 3.93 μg/m3. If wind speed increases by 1 m/s, the concentration level will decrease by 8.19 μg/m3. In the case that wind reaches highest average speed of 4.88 m/s, PM2.5 can decrease by 28.26 μg/m3. The decrease levels of 3.43 μg/m3 and 17.21 μg/m3 are reached when temperature and solar radiation go up to their max value of 42.35 Celsius degree and 985.65 W/m2, respectively (other factors held constant). These factors have been determined to play important role in altering mixing layer height which lead to changes in particles concentration . Increase in these factors results in increase mixing layer height and therefore reduce PM2.5 concentration on the ground level.
Table 4 presents analysis on the cross-correlations between meteorological factors. The result reveals that wind speed, temperature, solar radiation and air pressure are statistically related. Wind speed, temperature and solar radiation have a positive relationship with each other and that all three factors have an inverse relationship with air pressure. Table 4 indicates strongest inverse relationship between air temperature and pressure with R = −0.8064 and followed by the positive relationship between temperature and solar radiation with R = 0.4228. Third is the relationship between wind and temperature with R = 0.3226.
The positive long-term cross-correlations between wind speed and solar radiation was also found in studies by Dos Anjos et al. (2015)  using 2004-2013 temporal series from Fernando de Noronha Island, Brazil and Jerez et al. (2013)  using hourly series data for the period 1959-2007 in the Iberian Peninsula. Solar radiation affects wind speed by heating soil and ocean which creates pressure and air convection movement. Wind speed, on the other hand, affects cloud coverage which has an impact on solar radiation . High correlation between surface temperature, wind speed and air pressure was also found in the study by Wooten (2011) .
Combinations of meteorological conditions have different impacts to variations of PM2.5 concentration. The positive relationship between wind speed, temperature and solar radiation means that higher temperature leads to higher wind speed and solar radiation. This phenomenon also leads to reduced air pressure. Figure 3 shows the hourly measurements of wind speed, temperature,
Table 4. Correlation between surface air pressure, wind speed, temperature and solar radiation.
Note: * p < 0.05, ** p < 0.01, *** p < 0.001.
Figure 3. Average hourly measurements of wind speed, temperature, air pressure and solar radiation in 2017-2018.
air pressure and solar radiation in 2017-2018. Declining wind speed from 0:00 AM to 7:0 AM in combination with low radiation and temperature mainly contributes to higher levels of PM2.5 during this time period due to their negative correlations as analyzed in Table 3. Although the increasing trend in air pressure occurs during this time frames, its effect on PM2.5 concentration is small as compared to other factors, especially wind speed and solar radiation. As a result, the concentration level often stays high in early morning. The rising trend of wind speed, solar radiation and temperature which starts from the morning and reaches the peak around 1:00 PM-3:00 PM is also consistent with the downward trend of PM2.5 concentration (see Figure 2).
The equations are an exception to the prescribed specifications of this template. You will need to determine whether or not your equation should be typed using either the Times New Roman or the Symbol font (please no other font). Equations should be edited by Mathtype, not in text or graphic versions. You are suggested to use Mathtype 6.0 (or above version).
3.3. Four Seasons and PM2.5 Variations
The Capital city of Hanoi, Vietnam is characterized by four seasons: Winter, Spring, Summer and Autumn. Each season has its own meteorological characteristics that make it distinguished from others. Figure 4 describes the meteorological conditions in each season measured at data point location. It is noticeable that average wind speed, temperature and solar radiation are high during Summer. This condition probably leads to higher mixing layer height, resulting in low particle concentration in Summer. The positive relationship between wind speed and temperature and mixing layer height is also confirmed in the study by Soni et al. (2014). High solar radiation and wind speed may cause dust to spread widely in the air and be blown away by the wind which results in low suspended PM2.5. Winter is characterized by low wind speed, temperature and solar and high air pressure as compared to other months. This combination of weather factors may keep fine particulate matter suspended in the air which leads to high concentration in low mixing layer height near the ground.
Figure 4. Average (a) Air pressure; (b) Wind speed; (c) Temperature; (d) Solar radiation by season.
This type of climate confirms the close relationship between meteorological factors and PM2.5 concentration as analyzed in Table 1 and Table 3. Hot climatic conditions with strong winds and solar radiation make PM2.5 concentration during the Summer fall sharply compared to other seasons. The highest PM2.5 level is reported in Winter. This season is characterized by higher air pressure but low temperature, wind speed and radiation. This combination of meteorological factors may keep fine particulate matter suspended in the air which lead to high concentration level.
Given distinguished meteorological conditions in each season, changes in meteorological factors can have different effects on PM2.5 concentrations in each season. Table 5 present regression results about the correlation between meteorological conditions and PM2.5 concentration in each season. The analytical results indicate statistically consistent negative effects of wind speed over all four seasons. Solar radiation also shows consistent trend in Winter, Summer and Autumn, except Spring where there is no statistical significance found. Air pressure is found to have positive relationship with PM2.5 concentration in Winter, Spring and Autumn, but negative effects in Summer. Increasing temperatures tend to increase PM2.5 levels in Winter and Autumn but the trend is opposite in Spring and Summer.
This study investigates the temporal patterns of PM2.5 variations and how the meteorological factors, including wind speed, surface temperature, solar radiation and air pressure, affect these temporal patterns. The time series of hourly data was collected from the automatic monitoring stations (HORIBA) located on 556 Nguyen Van Cu street, Long Bien district, Hanoi of the Center for Environmental Monitoring (CEM), Vietnam Environment Administration (VEA) from January 01, 2017 to December 31, 2018. Loglinear regression models were performed for the analysis.
Table 5. Correlation between meteorological factors to seasonal variations of PM2.5 level.
Note: * p < 0.05, ** p < 0.01, *** p < 0.001.
The regression results suggest that there were seasonal, monthly and diurnal variations of PM2.5 concentrations. The seasonal analysis indicates that the concentration levels in Spring, Summer and Autumn were statistically negative correlated with the concentration levels in Winter. Summer has the lowest concentration level and is measured at 36.07% lower than Winter. The monthly analysis suggests that January and December are the months with highest PM2.5 levels. The concentration was found the lowest in May with 58.36% lower than in January. It was followed by April, August, July and March respectively. The diurnal analysis finds that the concentration reaches the highest level around 5:00 AM-6: 00 AM and the lowest levels around 2:00 PM-3:00 PM.
Meteorological factors are found to statistically affect PM2.5 concentration. Positive relationship is found between air pressure and the concentration level, meaning that when air pressure increases, PM2.5 tends to increase. Other three factors, including windspeed, temperature and solar radiation, show negative trend. This means PM2.5 level increases when wind speed, temperature and solar radiation decrease and vice versa. Analysis on the cross-correlations between meteorological factors and PM2.5 variations reveals hot conditions with strong winds and solar radiation lead to low particulate matter while low wind speed in combination with low temperature, low solar radiation and high air pressure contributes to high concentration.
We would like to thank the Center for Environmental Monitoring (CEM), Vietnam Environment Administration (VEA) for providing the time series of hourly data for the analysis. We also thank to Nguyen Thu Quynh for English editing.
 Cohen, D.D., Crawford, J., Stelcer, E. and Bac, V.T. (2010) Characterisation and Source Apportionment of Fine Particulate Sources at Hanoi from 2001 to 2008. Atmospheric Environment, 44, 320-328.
 Pham, H. Air Pollution in Vietnam Cities Hit Unhealthy Levels: Government Study-VnExpress International.
 Saksena, S., Quang, T.N., Nguyen, T., Dang, P.N. and Flachsbart, P. (2008) Commuters’ Exposure to Particulate Matter and Carbon Monoxide in Hanoi, Vietnam. Transportation Research Part D: Transport and Environment, 13, 206-211.
 Lasko, K., Vadrevu, K.P. and Nguyen, T.T.N. (2018) Analysis of Air Pollution over Hanoi, Vietnam Using Multi-Satellite and MERRA Reanalysis Datasets. PLoS ONE, 13, e0196629.
 Ly, B.-T., Matsumi, Y., Nakayama, T., Sakamoto, Y., Kajii, Y. and Nghiem, T.-D. (2018) Characterizing PM2.5 in Hanoi with New High Temporal Resolution Sensor. Aerosol and Air Quality Research, 18, 2487-2497.
 Streets, D.G., Yarber, K.F., Woo, J.-H. and Carmichael, G.R. (2003) Biomass Burning in Asia: Annual and Seasonal Estimates and Atmospheric Emissions. Global Biogeochemical Cycles, 17.
 Hai, C.D. and Oanh, N.T.K. (2013) Effects of Local, Regional Meteorology and Emission Sources on Mass and Compositions of Particulate Matter in Hanoi. Atmospheric Environment, 78, 105-112.
 Thuy, N.T.T., Dung, N.T., Sekiguchi, K., Thuy, L.B., Hien, N.T.T. and Yamaguchi, R. (2018) Mass Concentrations and Carbonaceous Compositions of PM0, PM2.5, and PM10 at Urban Locations in Hanoi, Vietnam. Aerosol and Air Quality Research, 18, 1591-1605.
 Oanh, N.K., Upadhyay, N., Zhuang, Y.-H., Hao, Z.-P., Murthy, D.V.S., Lestari, P., Villarin, J.T., Chengchua, K., Co, H.X. and Dung, N.T. (2006) Particulate Air Pollution in Six Asian Cities: Spatial and Temporal Distributions, and Associated Sources. Atmospheric Environment, 40, 3367-3380.
 Hien, P.D., Loc, P.D. and Dao, N.V. (2011) Air Pollution Episodes Associated with East Asian Winter Monsoons. Science of the Total Environment, 409, 5063-5068.
 Li, Y., Chen, Q., Zhao, H., Wang, L. and Tao, R. (2015) Variations in PM10, PM2.5 and PM1.0 in an Urban Area of the Sichuan Basin and Their Relation to Meteorological Factors. Atmosphere, 6, 150-163.
 Zhang, Z., Zhang, X., Gong, D., Quan, W., Zhao, X., Ma, Z. and Kim, S.-J. (2015) Evolution of Surface O3 and PM2.5 Concentrations and Their Relationships with Meteorological Conditions over the Last Decade in Beijing. Atmospheric Environment, 108, 67-75.
 Zhao, C.X., Wang, Y.Q., Wang, Y.J., Zhang, H.L. and Zhao, B.-Q. (2014) Temporal and Spatial Distribution of PM2.5 and PM10 Pollution Status and the Correlation of Particulate Matters and Meteorological Factors during Winter and Spring in Beijing. Environmental Sciences, 35, 418-427.
 Huang, F., Li, X., Wang, C., Xu, Q., Wang, W., Luo, Y., Tao, L., Gao, Q., Guo, J. and Chen, S. (2015) PM2.5 Spatiotemporal Variations and the Relationship with Meteorological Factors during 2013-2014 in Beijing, China. PLoS ONE, 10, e0141642.
 Guo, J., Xia, F., Zhang, Y., Liu, H., Li, J., Lou, M., He, J., Yan, Y., Wang, F. and Min, M. (2017) Impact of Diurnal Variability and Meteorological Factors on the PM2.5-AOD Relationship: Implications for PM2.5 Remote Sensing. Environmental Pollution, 221, 94-104.
 Liu, X., Li, C., Tu, H., Wu, Y., Ying, C., Huang, Q., Wu, S., Xie, Q., Yuan, Z. and Lu, Y. (2016) Analysis of the Effect of Meteorological Factors on PM2. 5-Associated PAHs during Autumn-Winter in Urban Nanchang. Aerosol and Air Quality Research 16, 3222-3229.
 Chen, T., He, J., Lu, X., She, J. and Guan, Z. (2016) Spatial and Temporal Variations of PM2.5 and Its Relation to Meteorological Factors in the Urban Area of Nanjing, China. International Journal of Environmental Research and Public Health, 13, 921.
 Pu, W., Zhao, X., Zhang, X. and Ma, Z. (2011) Effect of Meteorological Factors on PM2. 5 during July to September of Beijing. Procedia Earth and Planetary Science, 2, 272-277.
 Chou, C.-K., Hsu, W.-C., Chang, S.-Y., Chen, W.-N., Chen, M.-J., Huang, W.-R., Huang, S.-H., Tsai, C.-Y., Chang, S.-C. and Lee, C.-T. (2017) Seasonality of the Mass Concentration and Chemical Composition of Aerosols around an Urbanized Basin in East Asia. Journal of Geophysical Research: Atmospheres, 122, 2026-2042.
 Chuang, M.-T., Chou, C.C.-K., Lin, N.-H., Takami, A., Hsiao, T.-C., Lin, T.-H., Fu, J.S., Pani, S.K., Lu, Y.-R. and Yang, T.-Y. (2017) A Simulation Study on PM2.5 Sources and Meteorological Characteristics at the Northern Tip of Taiwan in the Early Stage of the Asian Haze Period. Aerosol and Air Quality Research, 17, 3166-3178.
 Tran, C.C. and Yanagida, J.F. (2019) Environmental Impact Assessment of Banagrass-Based Cellulosic Ethanol Production on Hawaii Island: A Spatial Analysis of Re-Suspended Soil Dust and Carbon Dioxide Emission. Applied Sciences, 9, 2648.
 Seibert, P., Beyrich, F., Gryning, S.-E., Joffre, S., Rasmussen, A. and Tercier, P. (2000) Review and Intercomparison of Operational Methods for the Determination of the Mixing Height. Atmospheric Environment, 34, 1001-1027.
 Schafer, K., Emeis, S., Hoffmann, H. and Jahn, C. (2006) Influence of Mixing Layer Height upon Air Pollution in Urban and Sub-Urban Areas. Meteorologische Zeitschrift, 15, 647-658.
 Oleniacz, R., Bogacki, M., Szulecka, A., Rzeszutek, M. and Mazur, M. (2016) Assessing the Impact of Wind Speed and Mixing-Layer Height on Air Quality in Krakow (Poland) in the Years 2014-2015. Journal of Civil Engineering, Environment and Architecture, 63, 315-342.
 Wang, C., Jia, M., Xia, H., Wu, Y., Wei, T., Shang, X., Yang, C., Xue, X. and Dou, X. (2019) Relationship Analysis of PM2.5 and Boundary Layer Height Using an Aerosol and Turbulence Detection Lidar. Atmospheric Measurement Techniques, 12, 3303-3315.
 Zhang, J. and Rao, S.T. (1999) The Role of Vertical Mixing in the Temporal Evolution of Ground-Level Ozone Concentrations. Journal of Applied Meteorology, 38, 1674-1691.
 Soni, K., Singh, M., Singh, G. and Agarwal, S. (2014) Sodar Mixing Height Estimates and air Pollution Characteristics Over Delhi a Big City During Spring and Summer. Journal of Acoustical Society of India, 41, 196-199.
 dos Anjos, P.S., da Silva, A.S.A., Stosic, B. and Stosic, T. (2015) Long-Term Correlations and Cross-Correlations in Wind Speed and Solar Radiation Temporal Series from Fernando de Noronha Island, Brazil. Physica A: Statistical Mechanics and Its Applications, 424, 90-96.
 Jerez, S., Trigo, R.M., Sarsa, A., Lorente-Plazas, R., Pozo-Vázquez, D. and Montávez, J.P. (2013) Spatio-Temporal Complementarity between Solar and Wind Power in the Iberian Peninsula. Energy Procedia, 40, 48-57.