Delhi is located at (28.37˚N; 77.12˚E, 216 m above mean sea level-MSL) and covers ~1500 km² of the industrial area including the suburban towns of Ghaziabad, Noida Faridabad, and Gurgaon. According to 2011 census, Delhi is the second highest populous (~11 million population) city in India, as well as it is one of the most polluted urban environments over the globe. Delhi climate is humid subtropical climate and hot semi-arid with high temperatures of 40˚C - 45˚C in pre-monsoon (warm) season while lowest of 13˚C - 15˚C during in winter period. The monsoon arrives at the end of June, along with an increase in humidity.

Pune is located at (18.5˚N; 73.8˚E) an altitude of about 550 m above the MSLs, it is situated in the western zone of the Deccan plateau and covers an area of ~700 km². It is one of the most rapidly developing cities in India in terms of industries as well as vehicular population and also urbanization [28] . Pune has a tropical wet and dry climate with average temperatures ranging between 20 to 33˚C [29] . Number of vehicles over Delhi and Pune are increasing 2 to 3 times [29] [30] within a decade. By the year 2010, the vehicular count for Delhi is 57 lakh’s and over Pune, 1000 vehicles per hour cross the signal [29] . The monitoring stations are placed at strategically selected locations by including different sectors with vehicular transport, industrial pollution, biomass burning, residential etc. More detailed information about the instrumentation, quality control of the data from the SAFAR network at different observational locations over Delhi and Pune (http://safar.tropmet.res.in/) are reported in [24] .

Continuous measurements of PM_2.5 and PM₁₀ with meteorological parameters were carried out from the System of Air Quality Forecasting and Research (SAFAR) network of the year of January-December, 2015. It consists of eight Air Quality Monitoring Stations (AQMS) and Automatic Weather Stations (AWS) set up across Delhi and Pune. PM_2.5 and PM₁₀ mass concentrations were observed using the Beta Attenuation Monitor (BAM-1020; Met One Instruments, Inc., USA). The BAM-1020 measures the mass concentrations of airborne particulate matters in (µg/m³) using the principle of beta ray attenuation. A small Carbon-14 (14^C) element provides a constant source of high energy electrons known as beta rays. The beta particles are detected and counted by a sensitive scintillation counter. Detailed information of BAM-1020 has been reported in a study by [19] . AWS was continuously operated and measures meteorological parameters, viz. wind speed, wind direction, temperature, relative humidity, etc., during the study period and the data were recorded at every 5 min interval and then binned at the one-hour interval after a quality check for further analysis and comparisons [31] . Planetary boundary layer height (PBLH) is taken from ERA-Interim reanalysis data.

The potential source contribution function (PSCF) is used to calculate the possible source locations using back trajectories [32] . The air mass back trajectories for Delhi and Pune during the year 2015 were calculated using the Hybrid Single-Particle Lagrangian Integrated Trajectory-4 (HYSPLIT-4) and are shown in Figure 7 and Figure 8 respectively. The back trajectories started from each 00:00 (UTC) on every month of the year 2015 at 100 m, which represent air masses from low altitudes. We calculated five days (120-h) back trajectories at 24 hourly based on HYSPLIT using National Centers for Environmental Prediction (NCEP) reanalysis data with a gridded resolution of 1˚ × 1˚. The arrival heights of the trajectories were at the surface (100 m). To calculate the PSCF, the geographic region covered by the trajectories was divided into an array of 0.3˚ × 0.3˚ grid cells with surface height and the PM_2.5 concentration kept within a threshold limit 95%.

The analysis was begun from the annual variations of PM at all the observational locations over Delhi and Pune. The higher values of PM₁₀ were observed over Delhi as compared to Pune (it was the same case for PM_2.5). All the stations over Delhi were having the same amount of PM in all the seasons with the higher values are found during the pre and post monsoon seasons, and the lower values are observed during the monsoon months due to the settling of pollutants by rainout and washout mechanism. And the same is the case for Pune with the higher values were found during the pre-monsoon and the lower is observed during the summer monsoon season (Figure 1). The annual mean mass concentration of PM_2.5 and PM₁₀ over Delhi were in the ranges of 27 - 796 μg/m³ and 9 - 600 (μg/m³), respectively. However, in Pune PM_2.5 and PM₁₀ were found in the ranges of 3 - 428 (μg/m³) and 1 - 180 (μg/m³), respectively in the year 2015. The annual mean concentrations (±standard deviations) of PM_2.5 and PM₁₀ for all air quality stations for Delhi and Pune are shown in Table 1 and Table 2, respectively. Mean values of PM₁₀ are higher over C.V Raman location with the value of around 229.15 (μg/m³) while the lower is observed for IMD Aya Nagar in terms of the mean values. While the maximum and minimum values that shown in Table 1 is over IGI airport is reported as highest among all the stations as 796.67 μg/m³ whereas the lowest is 38.46 μg/m³. Among the other observational locations, IMD Ayanagar is second to IGI airport with maximum (minimum) of PM₁₀ as 761.97(27.02) μg/m³.

On the other hand, IMD Lodhi road is recorded the maximum (minimum) values of PM₁₀ with 644.25 (42.85) μg/m³. On the overall statistics of PM₁₀ has shown the higher variability in terms of their standard deviation with 89.31 μg/m³ while the lower Standard Deviation (SD) is observed over IITM Delhi with the values of 72.06 μg/m³. On an average, all the stations over Delhi have the SD of more 90 ug/m³ except single station IMD Aya Nagar. Also, it is the

same case for PM_2.5 with the higher values of maximum values are recorded over IGI airport and IMD ayanagar with the minimum values of 14.85 (13.65) μg/m³ with SD of 82.87 (46.30) μg/m³ for the above two stations. From the statistical calculation, IMD Lodhi road recorded the lower values of both PM₁₀ and PM_2.5.

Both PM_2.5 and PM₁₀ concentrations are lower over Pune as compared to Delhi. Within Pune, overall results have shown the higher maximum values over IITM, Pune with Alandi is followed to IITM with the values of 428.04 for IITM and 418.41 μg/m³ for Alandi. Despite their maximum values, Alandi has shown higher variability of 51.32 μg/m³ than the IITM with a lesser value of about 34.44 μg/m³. Further analysis infers the minimum values over Katraj (Bhosari) as 3.05 (4.34) μg/m³ with higher SD values 47.10 (58.98). While it is the same case for the PM_2.5 over these locations following the minimum values as 1.34 (1.05) μg/m³ and higher SD as 28.43 (25.68) μg/m³. Over Pune, Nigidi observational locations have recorded the lowest variability in both PM₁₀ and PM_2.5 among all the other nine stations. However, the overall variations of PM₁₀ and PM_2.5 over Pune is 2 and 3 times lesser than that of the Delhi respectively, and the annual mean mass concentration of PM₁₀ and PM_2.5 over Delhi is varied from 27.02 μg/m³ to 796.67 μg/m³ and 9.37 μg/m³ to 600.03 μg/m³ respectively as well as in Pune the annual mean mass concentration of PM₁₀ and PM_2.5 is varied from 3.05 μg/m³ to 428.04 μg/m³ and 1.05 μg/m³ to 180.81 μg/m³ as recorded in the year 2015. The observed values of PM₁₀ and PM_2.5 over Delhi have substantially exceeded the standards of their annual averages with reference to Indian National Ambient Air Quality Standards (NAAQS) i.e. 100 μg/m³ and 60 μg/m³ for 24 hours average. But over Pune observed values of PM₁₀ and PM_2.5 exceeds in winter and Pre-monsoon seasons only and rest of the period maintains the standard levels. According to previous studies over Delhi, the transport sector has contributed 45% to total PM_2.5 concentrations, 27% of PM_2.5 is contributed from industrial sector, 24% from industries and 4% from power sector [30] however, and major sources of overall air pollution in Delhi are: emissions from vehicles (67%), coal-based thermal power plants (13%), industrial units (12%) and domestic exhaust (8%) [4] . Figure 1 shows that the particulate matter levels are higher in the November and December followed by the May months over the Delhi, we can observe in the Pune the PM concentration is high during November and December followed by the April month.

The average diurnal variations in PM₁₀, PM_2.5 and their ratio concentration at Pune and Delhi are shown in Figure 2. Hourly averaged mean values as well as standard deviations for their variability are presented from the observational locations over both Delhi and Pune. Pronounced diurnal variations of PM are observed in both Metro cities, but with different daily patterns in the two cities. The variability over different locations over Delhi found to be higher in terms of their SD than the locations over Pune (Figure 2). The maximum values of SD

was found in the early hours of the morning and later hours in the night over Delhi with also peak in the morning hours around 10 AM, on the other hand, the maximum values of SD are found in the morning hours around 8 to 10 am over different locations over Pune. These higher deviations are mainly attributed due to the usage of a high number of vehicles which cause the higher emissions in the observed values. In the Metro area, PM_2.5 concentrations are observed low at early morning hours to 8:00 am, followed by increase around 2:00 pm, and after which rises up to midnight 12:00 am. The metro area, therefore, displays a bimodal pattern with a weak signal. Whereas the Metropolitan cities are showing the strong bimodal pattern which is showing an almost similar pattern. The metro area has a dense traffic and huge pollution will lead to the high concentration of particulate matter as well as low diurnal variation. The morning peak is due to enhanced anthropogenic activity during rush hour. The condition is more or less same for both PM₁₀ and PM_2.5 with the less SD values over Pune as compared to Delhi. Interestingly the ratio of PM₁₀ and PM_2.5 has shown the higher values especially during the hours between 10 to 2 PM whereas it is more is less similar pattern over Delhi with higher deviations.

Further, we have analyzed the diurnal variations of PM₁₀, PM_2.5 and their ratio patterns during different seasons are shown in Figure 3(a)-(f) over Delhi and Pune. The mean values of both the concentrations are calculated using different locations over both the metropolitan cities and presented for the discussion. PM concentrations show bimodal diurnal distribution during all the seasons. The diurnal cycle of PM₁₀ is maximum in the midnight time around 12:00 AM with a

value of around 285 μg/m³ during the winter season in Delhi while the minimum is observed around 16:00 PM with a value of 200 μg/m³, however, their high deviations are also observed around same time with 115 μg/m³ while the minimum is observed as ~65 μg/m³. In case of Pune, it is observed that the maximum is around morning 10:00 AM with a value of about 200 μg/m³ and the minimum is same as Delhi at 16:00 PM with a value of 90 μg/m³. It is interesting to note that the variability is found to be higher during the midnight with a value around 40 μg/m³ while the minimum variability is observed in the afternoon hours with a value as low as 13 μg/m³. The mean and SD values over Pune are below the half of the values Delhi during the winter season. During the pre-monsoon season in the PM₁₀, the concentrations shows maximum in the early hours of the morning around 5:00 PM which is about 230 μg/m³ while the minimum is observed around 18:00 PM with a value of 150 μg/m³ over Delhi. As it is mentioned, the higher variability is reported during the winter season, and pre-monsoon also follows the winter season, showing the same pattern of variability in terms of their SD, on the other hand, minimum variability is observed at 18:00 PM. But in the case of Pune, PM₁₀ has shown two peaks in its diurnal time scale, one of the maximum is observed around 10:00 AM in the morning with a value of about 125 μg/m³ and the second peak is observed with the same magnitude at midnight 12:00 PM. Interestingly the diurnal scale of variability is more or less similar of a value of about approximately 50 ug/mg except for few hours between 14:00 PM to 20:00 PM with minimum variability. But it is clearly seen there is a diurnal variability in terms of their SD over Delhi that is observed with mean values showing two peaks with lesser amplitude difference between 10:00 AM to 16:00 PM during the summer monsoon season. Minimum values are observed during the early hours in the morning.

Higher SDs is observed in the morning hours while the lesser SDs is observed during the night hours over Delhi region in Figure 3(a) and Figure 3(b). Maximum (minimum) is found to be observed as 145 (110) μg/m³ over Delhi which is the half of the magnitude compared to the winter season. The lower values during this season are attributed to the rain/washout mechanism which makes the values lesser in amplitude. One can clearly notice that there are two peaks in the PM₁₀ over Pune during the monsoon season with a maximum observed during the late night around 11:00 PM with a value around 77 μg/m³ and the second maximum is found in the early hours of the morning with a value of 65 μg /m³. While the minimum is observed from 10:00 AM to 20:00 PM before start of a rise of its values in the diurnal cycle. This is may be the reason that the rain/washout mechanism is same as Delhi. It is also noticed that the values are lesser in both amplitude and as well as the standard deviations which are almost half of its values over Pune as compared to Delhi. After monsoon season completes, again the values of PM₁₀ start rising and the same is seen in their diurnal cycle during the post-monsoon season over Delhi with the values more than 300 μg/m³ during the late night and start to fall its values and it continues to reach a minimum of its value reaches 120 μg/m³ in the later afternoon. The variability also follows the mean values of maximum and minimum in timing. In the case of Pune, during the pre-monsoon, it follows same as like in the post-monsoon with two peaks one in the morning and other in the late night while the minimum is observed in the afternoon hours. Higher values of PM₁₀ are observed during the post-monsoon season due to its settling and trapping of pollutants makes more health-related problems. The similar patterns are also observed for the PM_2.5 during the different seasons over both Delhi and Pune. It also follows as like PM₁₀ for the maximum and minimum values and their variability during different seasons, but with the weaker in the amplitudes. Similar patterns are evident in the ratio of both pollutants in the ratios of mean values and standard deviations.

Ambient conditions of air pollution distribution will mainly depend on local meteorological conditions as well as long-range transport; Strong seasonality will lead to modulating the air quality levels [17] [27] . The seasonal distributions of ambient pollutants such as PM_2.5 and PM₁₀ are controlled by the emission sources, topography, local meteorology, transport factors and boundary layer height in the tropical atmosphere [19] . It is already well established that the meteorology will identify the transport and diffusion of the air pollution cycle, Meteorological factors such as winds direction, wind speed, temperature, precipitation, atmospheric boundary layer, relative humidity play an important role to determine the air pollution levels [33] . Impact of all these meteorological parameters in PM concentrations is significantly important and variability depends on the season or specific meteorological and atmospheric regimes dominated on certain days [34] . On the other hand, it is impossible to determine the influence of any individual meteorological parameter on PM concentrations with high accuracy due to contradictory and multiple effects that they may have on accumulation and dispersion of pollutants.