Microsoft Word - 01_JES_E4167_10_20_2020 Journal of Engineering Science 12(1), 2021, 1-8 DOI: https://doi.org/10.3329/jes.v12i1.53095 METEOROLOGICAL INFLUENCES ON URBAN AIR QUALITY PARAMETERS IN DHAKA CITY Nafisa Islam*1, Md. Golam Saroar2 and Tanvir Ahmed1 1 Department of Civil Engineering, Bangladesh University of Engineering and Technology, Dhaka, Bangladesh 2 Clean Air and Sustainable Environment Project, Department of Environment, Dhaka, Bangladesh Received: 07 May 2019 Accepted: 10 November 2020 ABSTRACT This study aims at investigating the effect of meteorological parameters on seasonal variation of particulate matter(PM) (both PM2.5 and PM10) using a 4-year (2013-2016) monitoring data of air quality parameters from CASE project implemented by the Department of Environment (DoE). Using monthly data of the Continuous Air Monitoring Station(CAMS) of Darus-Salam, Dhaka, cross correlation analysisis performed between PM and meteorological parameters where inverse relationships of PM with temperature, rainfall and relative humidity are found. Increased biomass burning during low temperature period, washout effect of rainfall, wet deposition mechanism of higher humidity may be held responsible for these negative correlations. Significant seasonal variation is observed from daily data analysis of Darus Salam station and it is found that winter PM concentrations are 4.5-5.5 times higher than monsoon PM concentrations. Seasonal cross-correlation between PM10 and PM2.5 shows lower correlation during winter (December-February) and monsoon (June-September) seasons. Two possible effects can attribute to this seasonal difference: i) presence of biomass burning during winter which increases PM2.5 and ii) presence of rainfall during monsoon which decreases PM10.PM2.5/PM10 ratios for different months indicate the contrasting influences of different mechanisms on different sized PM particles. PM2.5/PM10 ratio is found to be higher during December-February and lower during March- September with a rise in August, which indicates the effect of 3 mechanisms: i) dilution effect of wind speed on PM2.5 during December-February, ii) re-suspension effect of wind speed on PM10 during March-September and iii) more pronounced scavenging effect of rainfall on PM10 during August. The study indicates the need for properly accounting the influence of meteorology for better understanding of PM variation in urban areas in Bangladesh. Keywords: Cross correlation; Meteorology; PM; PM2.5/PM10 ratio; Seasonal variation. 1. INTRODUCTION Particulate matter (PM) is defined as a complex mixture of different sizes of airborne particle having different chemical compositions. They are mainly classified in two categories, the finer particles ranging from 0.005 µm to 2.5 µm which is called PM2.5 and the coarser particles with aerodynamic diameter ≤ 10 µm which is called PM10. Like other countries, particulate matter (PM) in ambient air has become one of the major concerns in Bangladesh. According to the Global Air Report 2017, Dhaka city has become 2nd most air polluted city (HEI, 2017). PM concentration in the air has been found to have significant correlation with diseases such as chronic respiratory illness, cardiovascular morbidity etc. (Dockery et al., 1993). To fully understand the process responsible for this distribution of particulate matters, analysis of the meteorological condition and detailed study on their influence on PM concentration are required. Different studies have shown that particulate matters are highly dependent on specific meteorological parameters (Dayan and Levy, 2005). It has been reported that wind speed, precipitation, relative humidity, temperature, time of day, atmospheric stability etc. are the major factors to drive the PM10 concentration in Germany (Gietl and Klemm, 2009). Several studies have been performed to evaluate the extent of urban pollution in the major cities of Bangladesh (Begum et al., 2013). Although the relationship between meteorology and PM has been investigated, very little information is available on the dependence of urban aerosol on atmospheric parameter in the major cities of Bangladesh. In this study, attempt is taken to determine the inherent relation between PM and meteorological parameters in Dhaka city using common statistical techniques. The aim is to obtain a deeper understanding of the process involved in the variation of PM concentration over time. 2. METHODS AND DATA ARCHIVING 2.1 Data Collection Under the Clean Air and Sustainable Environment (CASE) project, the Department of Environment (DoE) monitors real-time PM10 (24hr), PM2.5 (24hr) as well as ambient temperature (1hr), rainfall (1hr), relative JES an international Journal *Corresponding Author: nafisa.nikita@gmail.com https://www2.kuet.ac.bd/JES/ ISSN 2075-4914 (print); ISSN 2706-6835 (online) 2 Nafisa Islam et al. Meteorological Influences on Urban Air Quality ……….. humidity (1hr), through 11 Continuous Air Monitoring Stations (CAMS) throughout Bangladesh. Air quality and meteorological data of CAMS-3 (Darus Salam, Dhaka) and CAMS-8 (Red Crescent Campus, Sylhet) for the year of 2013-2016 are collected. However, the data of CAMS-3 (Darus Salam, Dhaka) is used for the analysis. 2.2 Approach for Analysis Single linear regression model is used to quantify the correlation between PM2.5 and PM10 with meteorological parameters. The regression equation is in the form of y = β 0 + β 1 x+ ε (1) Here, y is the concentration of PM2.5 or PM10, x is the meteorological parameters (Temperature, Rainfall, Relative Humidity and Solar Radiation), β is the regression coefficients and ε is the error term, where ε= (yi-ŷ), yi =observed y values, ȳ = mean value of series y and ŷ= y values given by the equation. The coefficient of determination r2 measures how related the PM concentration is with response to these meteorological parameters. r2 = St-Sr Sr (2) Here, St is Total Sum of Squares, where St =Total Sum of Squares = yi − ȳ n i=1 2 (3) And, Sr is Error Sum of Squares, where Sr = Error Sum of Squares = yi − ŷ ² n i=1 (4) r2 continues to increase with increasing terms to the model, which can deter the goodness of fit. Hence, adjusted r2 is introduced, which modifies the r2 based on the added terms to the model: Adjusted 𝑟 = 1 − (1 − 𝑟 ) ∗ (5) Here, m is the number of elements in a series and p is the number of independent variables. Besides, cross correlation analysis is performed between PM and meteorological parameters. Since the PM variation and meteorological parameters both vary with time, time dynamic analysis of cross correlation would be the best way to represent the actual relationship between PM and weather parameters through a lead-lag relationship. Basic cross correlation formula used in the analysis is written below: For 𝑘 ≥ 0, 𝐶 = ∑ [{𝑥(𝑡) − �̅�} ∗ {𝑦(𝑡 + 𝑘) − 𝑦}] (6) For 𝑘 ≤ 0, 𝐶 = ∑ [{𝑦(𝑡) − 𝑦} ∗ {𝑥(𝑡 − 𝑘) − �̅�}] (7) Here, x(t) is the concentration of PM at time t, y(t+k) is the respective meteorological parameter at time (t+k), k is the lag between two-time series x and y, T is the total number of elements in series x and y. In order to standardize the correlation values, the cross-correlation coefficient is calculated which is given by: rxy(k) = Cxy(k) Sx*Sy (8) Here, Sx = Cxx(0) (9) And, Sy = Cyy (0) (10) 3. RESULTS AND DISCUSSIONS 3.1 Average PM Concentration and Meteorological Conditions in Bangladesh In Bangladesh, the year can be divided into four different seasons: Winter (December-February), Pre-monsoon (March-May), Monsoon (June-September), Post-monsoon (October-November) (Begum et al., 2014). The climate of Bangladesh experiences prominent variation in weather during different seasons. It endures cold and dry air in winter as well as hot and humid air during the other three seasons. However, high temperature and high humidity is observed for most of the year. Precipitation shows marked distinction between seasons, maximum rainfall occurs in the monsoon and a minimum in winter. During winter, dry soil condition, scanty rainfall and low relative humidity prevails. During pre-monsoon, rainfall becomes moderately strong and relative humidity increases. During monsoon, moist air condition and high relative humidity prevails. Besides, the amount of rainfall also remains at its highest during this season. In the post monsoon, the amount of precipitation starts to decrease and so as the relative humidity. Journal of Engineering Science 12(1), 2021, 1-8 3 Figure 1: Daily average PM concentration with corresponding Bangladesh National Ambient Air Quality Standard (BNAAQS) for (a) PM2.5 and (b) PM10 Figure 1(a) and (b) show the daily 24hr average concentration for PM2.5 and PM10, respectively, spanning for the year of 2013-2016. The average PM2.5 and PM10 concentrations during the study period are found to be 91.03µg/m3 and 161.69 µg/m3, respectively. To understand the PM variation throughout the entire year, seasonal and annual mean are calculated for the entire study period and the results are shown in Table 1. Winter PM concentration is found to be considerably higher than the Bangladesh National Ambient Air Quality PM2.5 Standard of 65 µg/m3 (daily 24hr average) and PM10 Standard of 150 µg/m3 (daily 24hr average). For the year of 2013-2016, respectively 172, 194, 174 and 161 daily PM concentrations, corresponding 48, 53, 48 and 44% of the sampling days exceeded the BNAAQS PM2.5 limit value. Similarly, for this four-year period, respectively 42, 44, 40 and 43% of the sampling days exceeded the BNAAQS PM10 limit value. From the above statistics, it is evident that PM2.5 concentration is more prone to exceed the limit value compared to PM10. Performing the analysis on seasonal basis, the exceedance is found to be highest for winter season (99.45% for PM2.5 and 95.85% for PM10) and lowest for monsoon season (4.5% for PM2.5 and 1.23% for PM10), while exceedance during other seasons are moderate. Significant monthly variation has been obtained for both PM fractions. The winter to monsoon ratio of PM2.5 and PM10 concentration during 2013-2016 were 6.09, 5.56, 5.04 and 6.2 as well as 4.22, 4.54, 4.11 and 4.95, respectively (Table 1). Comparing with other studies, our observations of the difference between PM concentration of winter and monsoon season have been found very high. For example, the winter to monsoon PM ratio has been found to be 2.9 for PM10 and 2.2 for PM2.5 in India (Kulshrestha et al., 2009) whereas winter to summer ratio of 2.14 forPM10 has been found in Egypt (Elminir, 2005). This may be because, during winter, higher atmospheric stability as well as dry weather condition favors suspension of particulate matter in the air. Along with it, brick kilns in Bangladesh remain operational during this season. Aerosol concentration in monsoon was minimum due to scavenging effect of precipitation and the higher winter to monsoon ratio for PM10 indicates that this scavenging effect is more pronounced on PM10 compared to PM2.5. 3.2 Seasonal PM Concentration Prevalence Figure 2 shows frequency distribution of PM2.5/PM10 ratio for all seasons which is divided into nine categories starting from <0.2 to <1. Here, during high PM prevailing season i.e. in winter, PM2.5/PM10 ratio curve shows a symmetric pattern with its peak at 0.6-0.7 (above 40% of the cases). It indicates that, this ratio fits normal distribution during winter and the contribution of PM2.5 remains higher. Similarly, during low PM prevailing season i.e. in monsoon, symmetric pattern is also observed with peak at 0.4-0.5 (above 35% of the cases), which indicates that this distribution too follows normal distribution. However, slightly right skewed distribution is observed for pre-monsoon and post-monsoon season with peak at 0.4-0.5, which indicates that during these seasons, contribution of PM2.5 concentration starts to decrease after winter. During pre-monsoon, monsoon and post-monsoon season, approximately 40, 35 and 45%, respectively are observed to be in the range of 0.4-0.5 4 Nafisa Islam et al. Meteorological Influences on Urban Air Quality ……….. whereas it is only 5% during winter. During winter highest ratio is in the segment between 0.6-0.7, however fewer value 8, 12 and 16% are found for pre-monsoon, monsoon and post-monsoon, respectively. High PM2.5/PM10 ratio during winter indicates that significant portion of pollution particles fall under the size distribution of PM2.5. Table 1: Average PM concentrations, their seasonal ratio and annual exceedance from Bangladesh National Ambient Air Quality Standard (BNAAQS) Year PM2.5 Average Concentration PM10 Average Concentration PM2.5 Win/ Mon. ratio PM10 Win/ Mon. ratio Annual PM2.5 exceedance from BNAAQS (%) Annual PM10 exceedance from BNAAQS (%) 2013 Winter 187.41 ± 57.37 277.24 ± 84.43 6.09 4.22 47.12 41.87 Pre-monsoon 78.96 ± 52.17 155.22 ± 96.12 Monsoon 30.77 ± 14.68 65.68 ± 28.22 Post-monsoon 78.69 ± 40.88 136.94 ± 80.05 Annual 89.61 ± 73.62 152.31 ± 108.65 2014 Winter 168.98 ± 47.93 260.55 ± 68.95 5.56 4.54 53.15 43.53 Pre-monsoon 89.28 ± 52.55 175.58 ± 89.83 Monsoon 30.42 ± 11.55 57.34 ± 20.35 Post-monsoon 117.76 ± 69.83 181.72 ± 117.00 Annual 95.11 ± 70.50 159.63 ± 108.53 2015 Winter 173.48 ± 46.14 258.45 ± 76.65 5.04 4.11 47.67 40.5 Pre-monsoon 72.89 ± 43.28 139.16 ± 81.85 Monsoon 34.39 ± 16.11 62.92 ± 30.21 Post-monsoon 93.32 ± 46.09 169.04 ± 85.18 Annual 88.32 ± 65.34 148.13 ± 100.82 2016 Winter 188.28 ± 34.84 301.96 ± 76.72 6.20 4.95 44.11 43.01 Pre-monsoon 69.70 ± 43.59 151.52 ± 77.20 Monsoon 30.38 ± 14.91 61.06 ± 28.68 Post-monsoon 79.14 ± 48.10 148.32 ± 78.37 Annual 87.88 ± 70.43 158.55 ± 112.23 Figure 2: Frequency distribution of PM2.5/PM10 ratio Figure 3: Annual variation of PM2.5/PM10 for the years of 2013, 2014, 2015 and 2016 3.3 Meteorological Parameters Influencing PM Levels As the dispersion condition of atmosphere is primarily responsible for the accumulation of PM particle in air, the primary focus is on the role of temperature, relative humidity and precipitation for the variation of PM levels. The results of the correlation analysis between meteorological parameters and different sized PM Journal of Engineering Science 12(1), 2021, 1-8 5 particles are shown in Table 2. A significance level of 5% (p=0.05) has been chosen to be the threshold for determining the significance of correlation analysis. Table 2 shows that, dominant meteorological parameters those drive PM around Dhaka city are temperature (adj. r2= 0.681 for PM2.5; adj. r2=0.566 for PM10) and relative humidity (adj. r2= 0.244 for PM2.5; adj. r2= 0.413 for PM10). Solar radiation (adj. r2= 0.232 for PM2.5; adj. r2=0.131 for PM10) and rainfall (adj. r2= 0.141 for PM2.5; adj. r2=0.157 for PM10) are observed to exert weak influence over PM variation. Temperature is found to have strong negative relation with particulate matters (rx-corr= -0.825 at lag 0 for PM2.5; rx-corr= -0.752 at lag 0for PM10). This represents the fact that, with increase of temperature, PM decreases and vice versa. When temperature becomes lower i.e. during winter season, formation of stagnant air condition occurs and simultaneous biomass burning activities increases in the kilns, which gives rise to the particulate matter concentration. The correlation coefficient between air pollutants and relative humidity is found to be significant (rx-corr= -0.494 at lag 0 for PM2.5; rx-corr= -0.643 at lag 0 for PM10). High humidity indicates higher precipitation events with in cloud scavenging, reduction in the formation of OC (Organic Carbon) and EC (Elementary Carbon), higher moisture absorption and subsequent settling down of particles, all of which eventually result in low concentration of particulate matters. Relative humidity has been found to have slightly higher correlation with coarser particle. This is because the wash out effect of humidity and precipitation is more profound for the case of coarser particles. Negative correlation is observed between solar radiation and PM (rx-corr= -0.482 at lag 0 for PM2.5; rx-corr= -0.362 at lag 0 for PM10). This relation might indicate the phenomena that, increase of incident solar radiation leads to surface warming which cause rise of boundary layer height (BLH). When BLH increases, PM gets more space for dispersion. Higher dispersion results in decrease of PM concentration in the ambient air. Table 2: Cross correlation and single linear regression analysis results between PM concentrations and meteorological variables Predict-ion Variable (CAMS 3) Using Monthly Data (n=48) Cross Correlation Single Linear Regression (SLR) r(lag) r² Adjusted r² Error Sum of Square PM2.5 Temperature -0.825(0)** 0.681** 0.674** 57763 Rainfall -0.376(0)* 0.141 0.123 155600 Humidity -0.494(0)* 0.244* 0.228* 136990 Solar Radiation -0.482(0)* 0.232 0.215 139210 PM10 Temperature -0.752(0)** 0.566** 0.556** 178320 Rainfall -0.396(0)* 0.157 0.138 346130 Humidity -0.643(0)** 0.413* 0.401* 240850 Solar Radiation -0.362(0)* 0.131 0.112 356600 Statistical significance indicators are as follows: **, p < 0.001; *, 0.01>p > 0.001, otherwise 0.05>p>0.01 Table 3: Annual cross correlation coefficients between PM and other meteorological parameters (Rainfall & Temperature) at CAMS-3 and CAMS-8 Meteorological Parameters CAMS No. PM2.5 PM10 n a Correlation Coefficients n a Correlation Coefficients Temperature CAMS-3(Dhaka) 48 -0.825** 48 -0.752** CAMS-8 (Sylhet) 47 -0.855** 48 -0.862** Rainfall CAMS-3 (Dhaka) 48 -0.340* 48 -0.396* CAMS-8 (Sylhet) 47 -0.731** 47 -0.732** a Cross correlation coefficients at zero lag. Statistical significance indicators are as follows: **, p < 0.001; *, 0.01>p > 0.001, otherwise 0.05>p>0.01 Correlation analysis between PM concentration and rainfall is also carried out where weak negative correlation is found between them (rx-corr= -0.376 at lag 0 for PM2.5; rx-corr= -0.396 at lag 0for PM10). However, a strong negative correlation has been reported between average monthly PM and rainfall in Sylhet (CAMS-8) (Table 3). (r= -0.731 at lag 0 for PM2.5; r= -0.732 at lag 0 for PM10) (Table 3). In Dhaka, amount of rainfall is low compared to Sylhet where consistent rainfall is a distinct feature (in Northeastern part of Bangladesh). In 6 Nafisa Islam et al. Meteorological Influences on Urban Air Quality ……….. Bangladesh, rainfall is caused by the influence of the Southwest monsoon (Hossain et al., 2013). Total precipitation in Dhaka in 2016 is 462.02 mm with highest rainfall recorded in July (157.74 mm) and total precipitation in Sylhet in 2016 is 1606.8 mm with the highest rainfall recorded in the month of April (567.4 mm). Therefore, lower correlation in Dhaka may occur due to the long interval of consistent rainfall occurring in this area. This leads to the conclusion that, rainfall amount and duration both contributes combinedly in PM fluctuation. Like RH, higher correlation is observed between PM10 and rainfall which indicates that scavenging effect is more effective on PM10 than on PM2.5. 3.4 Global Comparison of Correlation Coefficients A comparative analysis has been conducted between this study and other literature values, based on the calculated correlation coefficients for PM2.5 and PM10 with meteorological parameters and is presented in Table 4. From Table 4, it is evident that, our results show similarity with the analyses conducted in India, Turkey and Egypt, whereas contradictory relation has been observed for studies conducted in Greece, Germany and USA. Negative correlation for temperature with PM has been observed for Bangladesh, India, Turkey and Egypt whereas, positive correlation has been obtained for Greece, Spain, Germany and USA. This variation mainly occurs due to the difference in weather condition and chemical composition of particulate matters all over the world. Weather patterns are similar for Bangladesh, India and Turkey, since all of them fall under the subtropical region. Biomass burning activities during winter season contribute to the higher concentration of PM. During low temperature period, particulate matter concentration becomes high and thus inverse relationship is formulated between PM and temperature. However, considering USA, Greece and Germany, high temperature is favorable for atmospheric chemical reaction. Hence, secondary particle formation is favored by temperature increase which produces positive correlation between PM and temperature. Table 4: Pearson’s correlation coefficients between PM10 and meteorological parameters in different regions Reference Country Tempera -ture Relative Humidity Wind Speed Precipitation PM2.5 (Galindo et al., 2011) Spain(yearly) -0.016 0.048 -0.496 -- (Tai et al., 2010) USA(yearly) 0.4-0.7 (-0.1)-(-0.15) (South) (-0.05)-1.0 -- 0.05-0.14 (North) (Bhaskar, and Mehta, 2010) India(yearly) -0.64 -- -0.53 0.01(Northern) - 0.74(Southern) (Akyuz, and Cabuk, 2009) Turkey(winter) -0.324 -0.108 -0.350 -- (Chaloulakou et al., 2003) Greece(winter) 0.46 -- -0.54 -- -- Bangladesh -0.681 -0.244 -- -0.141 PM10 (Elminir, 2005) Egypt(yearly) -0.48 0.252 -- -- (Galindo et al., 2011) Spain(yearly) 0.601 0.189 -0.334 -- (Hien et al., 2002) Germany(yearly) 0.17 -0.15 -0.49 -0.38 (Akyuz, and Cabuk, 2009) Turkey (winter) -0.155 -0.237 -0.409 -- (Bhaskar, and Mehta, 2010) India(yearly) -0.34 -0.44 -0.17 -0.53 (Chaloulakou et al., 2003) Greece(yearly) 0.39 -- -0.43 -- -- Bangladesh -0.566 -0.413 -- -0.157 Considering Relative humidity (RH), high humidity condition leads to higher moisture absorption and subsequent settling down of particles in the subtropical region. Therefore, negative correlation occurs for RH with PM in Bangladesh, India and Turkey, whereas around the European countries i.e. in USA, Greece and Germany, ultrafine particulate formation is found to be positively affected in the presence of high humidity, thus resulting in positive correlations between particulate matter and relative humidity. 3.5 Relationship between Different PM Size Fractions The relationship between fine particles PM2.5 and coarse particle PM10 is studied using cross correlation coefficients. The data are divided into four seasons for this analysis. The results are presented in the Table 5. Winter and monsoon correlation coefficients have been found to be lower than the other seasons. These Journal of Engineering Science 12(1), 2021, 1-8 7 differences between the coefficients are due to meteorological conditions those drive the PM concentration to change. It is linked to the seasonal changes of the weather conditions. During winter, there are major natural sources of PM, such as biomass burning activities, fossil fuel burning form vehicles and burning of agricultural soil and clays in the brick kilns which operate mainly during winter. These activities produce significant number of fine particles i.e. PM2.5. Therefore, the contribution of fine particles becomes much higher in winter. Besides, during monsoon season, significant reduction in PM10 occurs by the wet deposition mechanism of continuous precipitation. Thus, increase in PM2.5 in winter and decrease in PM10 in monsoon leads to the reduction of correlation coefficients between PM2.5 and PM10 during these seasons. During other seasons, the contribution of abovementioned sources, which mainly enriches PM2.5, becomes low. As a result, the obtained coefficients are higher for other seasons compared to winter. Annual PM2.5/PM10 variability is examined from 2013 to 2016 and it is found that the ratio remains low during the period of March to September i.e. pre-monsoon-monsoon season (Figure 3). During this period, wind speed remains at its maximum which induces the re-suspension effect of PM10. Therefore, the concentration of PM10 becomes higher in the atmosphere compared to PM2.5. Due to these reasons, the PM2.5/PM10 ratio remains low during March-September. Besides, a distinct rise of PM2.5/PM10 ratio can be observed in August. During August, rainfall remains at its maximum, which induces scavenging mechanism that works more effectively on PM10. As a result, average concentration of PM10 drops more compared to PM2.5, resulting in a distinct rise in PM2.5/PM10 ratio. Therefore, this figure is indicative of 3 mechanisms: i) dilution-effect of wind speed (WS) on PM2.5 during December-February, ii) re-suspension effect of WS on PM10 during March-September, and iii) more pronounced scavenging-effect of rainfall on PM10 during August. Table 5: Cross correlation analysis results between PM2.5 and PM10 concentrations PM10 Winter Pre- monsoon Monsoon Post- monsoon All Data PM2.5 Winter 0.792(0)** Pre-monsoon 0.922(0)** Monsoon 0.812(0)** Post-monsoon 0.954(0)** All Data 0.944(0)** Statistical significance indicators are as follows: **, p < 0.001; *, 0.01>p > 0.001, otherwise 0.05> p>0.01 4. CONCLUSION The influence of meteorological parameters on seasonal variation of particulate matter (PM) is examined using a 4-year (2013-2016) monitoring data of air quality parameters. Using monthly-data of the Continuous Air Monitoring Station (CAMS) of Darus Salam, Dhaka, cross-correlation and Pearson’s correlation analysis are performed between PM and meteorological parameters. Significant seasonal variation is observed and it has been found that winter PM concentrations are 4.5-5.5 times higher than monsoon PM concentrations. Major meteorological parameters that control PM in the air of Dhaka city are – temperature and relative humidity. Inverse relationships of PM with temperature, rainfall and relative humidity have been found. Increased biomass burning during low temperature period, washout effect of rainfall, dry deposition effect of higher humidity may be held responsible for these negative correlations. Besides, comparison of correlations between rainfall and PM for CAMS-3 (Dhaka) and CAMS-8 (Sylhet) indicates that, rainfall duration along with rainfall amount play major role to dominate PM. Dry deposition and scavenging effect are more effective to drive coarser particles (PM10) compared to finer particles (PM2.5). ACKNOWLEDGEMENTS The authors would like to thank the Department of Environment (DoE) for providing the background air quality and meteorological data. REFERENCES Akyuz, M., and Cabuk H., 2009. Meteorological variations of PM2.5/PM10 concentrations and particle- associated polycyclic aromatic hydrocarbons in the atmospheric environment of Zonguldak,Turkey, J. Hazard. Mater., 170, 13-21. Begum, B. A., Hopke P. K., and Markwitz A., 2013. Air pollution by fine particulate matter in Bangladesh, Atmos. Pollut. Res., 4(1), 75-86. Begum, B. A., Saroar G., Nasiruddin M., Randal S., Sivertsen B., and Hopke P., 2014. Particulate matter and Black Carbon monitoring at urban environment in Bangladesh, Nucl. Sci. Appl., 23(1 & 2). 8 Nafisa Islam et al. Meteorological Influences on Urban Air Quality ……….. Bhaskar, B.V., and Mehta V. M., 2010. Atmospheric Particulate Pollutants and their Relationship with Meteorology in Ahmedabad, Aerosol Air Qual. Res., 10, 301-315. Chaloulakou, A., Kassomenos P., Spyrellis N., Demokritou P., and Koutrakis P., 2003. Measurements of PM10 and PM2.5 particle concentrations in Athens, Greece, Atmos. Environ., 37(5), 649-660. Dayan, U., and Levy I., 2005. The Influence of Meteorological Conditions and Atmospheric Circulation Types on PM10 and Visibility in Tel Aviv, J. Appl. Meteorol. 44(5), 606-619. Dockery, D., Schwartz J., and Spengler J., 1993. Air pollution and daily mortality: Associations with particulates and acid aerosols, Environ. Res., 59, 362-373. Elminir, H. K., 2005. Dependence of urban air pollutants on meteorology, Sci. Total Environ. 350(1), 225-237. Galindo, N., Varea M., Gil-Moltó J., Yubero E., and Nicolás J., 2011. The Influence of meteorology on particulate matter concentrations at an urban Mediterranean location, Water Air Soil Poll. 215(1), 365- 372. Gietl, J. K., and Klemm O., 2009. Analysis of traffic and meteorology on airborne particulate matter in Munster, Northwest Germany, J. Air Waste Manage. 59(7), 809-818. HEI, 2017. State of Global Air 2019: Special Report, Health Effect Institute, Institute of Health Metrics and Evaluation (IHME), Boston, MA. Hien, P. D., Bac V. T., Tham H. C., Nhan D. D., and Vinh L. D., 2002. Influence of meteorological conditions on PM2.5 and PM2.5−10 concentrations during the monsoon season in Hanoi, Vietnam, Atmos. Environ. 36(21), 3473-3484. Hossain, M. A., Hossain M. S., Bhuyan M. A., Karmakar R., and Hossain M. M., 2013. Annual Flood Report 2013. Flood Forecasting & Warning Centre, Processing & Flood Forecasting Circle, Bangladesh Water Development Board, Ministry of Water Resources, Dhaka. Kulshrestha, A., Satsangi G., Masih J., and Taneja A., 2009. Metal concentration of PM2.5 and PM10 particles and seasonal variations in urban and rural environment of Agra, India, Sci. Total Environ., 407(24), 6196-6204. Tai, A., Mickley L., and Jacob D., 2010. Correlations between fine particulate matter (PM2.5) and meteorological variables in the United States: Implications for the sensitivity of PM2.5 to climate change, Atmos. Environ., 44(32), 3976-3984. Weather Underground, https://www.wunderground.com. Last accessesd on 17 Dec. 2017. © 2021 the Authors. Journal of Engineering Science published by Faculty of Civil Engineering, Khulna University of Engineering & Technology. This is an open access article under the terms of the Creative Commons Attribution- NonCommercial-NoDerivatives License, which permits use and distribution in any medium, provided the original work is properly cited, the use is non-commercial and no Modifications or adaptations are made.