Advertisement
Browse Subject Areas
?

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here.

  • Loading metrics

Spatio-Temporal Variation and Futuristic Emission Scenario of Ambient Nitrogen Dioxide over an Urban Area of Eastern India Using GIS and Coupled AERMOD–WRF Model

Spatio-Temporal Variation and Futuristic Emission Scenario of Ambient Nitrogen Dioxide over an Urban Area of Eastern India Using GIS and Coupled AERMOD–WRF Model

  • Sharadia Dey, 
  • Srimanta Gupta, 
  • Precious Sibanda, 
  • Arun Chakraborty
PLOS
x

Abstract

The present study focuses on the spatio-temporal variation of nitrogen dioxide (NO2) during June 2013 to May 2015 and its futuristic emission scenario over an urban area (Durgapur) of eastern India. The concentration of ambient NO2 shows seasonal as well as site specific characteristics. The site with high vehicular density (Muchipara) shows highest NO2 concentration followed by industrial site (DVC- DTPS Colony) and the residential site (B Zone), respectively. The seasonal variation of ambient NO2 over the study area is portrayed by means of Geographical Information System based Digital Elevation Model. Out of the total urban area under consideration (114.982 km2), the concentration of NO2 exceeded the National Ambient Air Quality Standard (NAAQS) permissible limit over an area of 5.000 km2, 0.786 km2 and 0.653 km2 in post monsoon, winter and pre monsoon, respectively. Wind rose diagrams, correlation and regression analyses show that meteorology plays a crucial role in dilution and dispersion of NO2 near the earth’s surface. Principal component analysis identifies vehicular source as the major source of NO2 in all the seasons over the urban region. Coupled AMS/EPA Regulatory Model (AERMOD)–Weather Research and Forecasting (WRF) model is used for predicting the concentration of NO2. Comparison of the observed and simulated data shows that the model overestimates the concentration of NO2 in all the seasons (except winter). The results show that coupled AERMOD–WRF model can overcome the unavailability of hourly surface as well as upper air meteorological data required for predicting the pollutant concentration, but improvement of emission inventory along with better understanding of the sinks and sources of ambient NO2 is essential for capturing the more realistic scenario.

Introduction

The surface emission sources and patterns of major air pollutants have been substantially changing over the tropical region. Rapid urbanization has led to an increasing number of large population agglomerations. Gradual degradation of air quality is one of the negative outcomes of modernization on human beings and environment. Escalating air pollution in urban areas is a matter of concern worldwide. The increasing levels of gaseous air pollutants pose a serious risk to human health and environment due to their detrimental effects. Nitrogen dioxide (NO2) is one of the criteria pollutants identified by Clean Air Act of 1970. It is an important trace gas which has a potential direct role in global climate change and plays a central role in tropospheric chemistry. It acts as a precursor for a number of harmful secondary air pollutants such as tropospheric ozone (O3) and plays a crucial role in the formation of acid rain. In the troposphere, nitric oxide (NO) is mainly emitted which in turn is rapidly converted to NO2. During daytime, a steady state is established NO and NO2 leading to the formation of tropospheric O3. The residence time of NO2 in the atmosphere is found to be approximately 0.5–2 days. Current scientific evidence links short-term NO2 exposures (ranging from 30 minutes to 24 hours) with adverse respiratory effects including airway inflammation in healthy people, increased respiratory symptoms in peoples suffering from asthma and increased epilepsy attack [1]. Oxides of nitrogen i.e. NOx (including NO2) and volatile organic compounds react in the presence of heat and sunlight to form O3 which in turn causes reduction in lung function, aggravation of pre-existing respiratory disease (such as asthma), increased daily hospital admissions and emergency department visits for respiratory causes and excess mortality.

The increasing levels of NO2 and NOx especially in the urban areas have gained attention worldwide. Along with other gaseous pollutants, NO2 and NOx were monitored and analyzed in in Pakistan [2], Al-Ain city, UAE [3], Metropolitan area of Monterrey, Mexico [4] etc. Investigation of the concentration of NO2 and NOx were carried out in different spatial and temporal scale in different corners of India like Lucknow, Haryana, Kolkata, Delhi, Burdwan and Gopalpur [510]. Zhao et al. [11] explored the association of higher concentration of ambient NO2 with high ozone days (HODs) over Shanghai, China. The interaction of multiple sources and various processes in different spatial and temporal scales make the urban air quality modeling more complicated. Borge et al. [12] performed a comprehensive source apportionment study in the Madrid metropolitan area by using a multi-scale, multi-pollutant air quality modeling system (WRF- SMOKE-CMAQ). He et al. [1314] predicted particulate matters at urban area by using coupled artificial neural network—chaotic particle swarm optimization algorithm as well as by hybrid model combining multi layer perceptron model and principal component analysis. Several researches have been performed on driver’s bound rationality, fuel consumption and emissions [1516]. Dispersion of a pollutant is a complex function of meteorological factors, planetary boundary layer characteristics and interactions with other species present in the ambient air. Therefore, quality data of these parameters are required as inputs to the dispersion models used for modeling the urban air quality status. Rao et al. [17] and Sharma et al. [18] evaluated the performance and predictive capacity of some commonly used dispersion models and concluded that these models are consistence with the dynamic nature of the atmosphere and are suitable for exploring the dispersion of pollutants. Such dispersion models are increasingly used for forecasting urban air quality status [1920] and the necessary meteorological inputs are generated using suitable prognostic models like e.g. MM5 [21], Eta [19], WRF [20] etc. Weather Research and Forecasting (WRF) can successfully generate the meteorological inputs required for AERMOD [20].

The study area (Durgapur) has witnessed rapid industrialization and urbanization in the last few decades and it is known to be one of the most polluted urban areas of the country. Deteriorating air quality scenario of this urban area has posed a serious risk to human health and environment due to their detrimental effects. To the best of authors’ knowledge, spatio–temporal variation of air pollutants along with its future projection has not yet taken place over this region. Such a study is essential for formulation and effective implementation of air pollution abatement measures. The focus of this study is on the spatial and temporal variation of the NO2 over this tropical urban area (Durgapur) and to obtain futuristic emission scenario over this region. The spatial and temporal variation of NO2 is obtained by using Geographical Information System (GIS) based Digital Elevation Model (GeomaticaV.10.1). A Gaussian air pollutant dispersion model AMS/EPA Regulatory Model or AERMOD [22] is used for understanding the dispersion of NO2 over the chosen area. AERMOD needs hourly surface and upper air meteorological observations for simulating the pollutant dispersion which is not available over this urban area. A high resolution prognostic model, Weather Research and Forecasting (WRF), is used for generating the required meteorological data. Coupled WRF–AERMOD simulates the present as well as future emission scenario of NO2 over the urban area. Finally, a comparative study of the simulated and observed values of NO2 at different sites over the urban is performed.

Data and Method

Description of the study area

Durgapur (chosen urban area) is situated in the Burdwan district of West Bengal, India. It is located on the bank of River Damodar. This area is covered with Red and Yellow Ultisols soil and the topography of this area is undulating, with an average elevation of 65 m MSL. This area experiences a transitional climate between the tropical wet and dry climate and the more humid subtropical climate. Three different sites are selected in the urban area for sampling (denoted by red dots in Fig 1).The sites are located in private ownership areas, so no specific permissions are required for the sampling activities in the chosen sites. Moreover, the field studies do not involve any endangered or protected species. The details of these sites are as follows:

  1. Site I (Muchipara) is situated at 23° 30 ′13.79 ″ N and 87° 21′16.82 ″ E. The sampling site is adjacent to National High Way (NH-2) or the Grand Trunk Road. This site represents an area with high vehicular density.
  2. Site II (B Zone) is located at 23°33′ 54.21ʺ N and 87°19′16.22ʺ E. This site is situated in a residential area which is approximately 4.1 km from national highway and 7.5 km industrial area.
  3. Site III (DVC-DTPS Colony) is situated in an industrial area at 23° 31′ 35.11ʺ N and 87° 15′ 27.94ʺ E. River Damodar flows to its south. The Durgapur Steel Plant (DSP) and the Durgapur Thermal Power Station (DTPS) of the Damodar Valley Corporation (DVC) are close to this site.

Data collection

High Volume Sampler (Envirotech APM 460BL) is used for 24 hour sampling and the concentration of ambient NO2 is determined by Modified Jacobs and Hochheiser Method [23]. Relative humidity and temperature are measured by a portable hygrometer (Model-HTC-1), wind speed is measured by a digital anemometer (Model-Lutron-AM-4201) and wind direction is recorded by a wind vane. The meteorological parameters are recorded at a regular interval of 1 hour.

The data of different criteria pollutants are collected from the Durgapur Station of West Bengal Pollution Control Board (www.wbpcb.gov.in).

Statistical analyses

Correlation analysis.

Pearson correlation coefficients between NO2 and different meteorological parameters are calculated by using the formula (1) Where, X and Y are two variables, with means and respectively and with standard deviations Sx and Sy respectively.

Regression analysis.

Multiple linear regression (MLR) attempts to model the relationship between two or more explanatory variables (independent variables) and a response variable (dependent variable) by fitting a linear equation to observed data. MLR technique has the capability of exploring the contribution of selected variables to chosen air pollutant concentration. The general equation of MLR is expressed as [24] (2) where,

bi is the regression coefficient,

xi is the independent variable, and

ξ is the stochastic error associated with the regressions.

Principal component analysis (PCA).

Among multivariate techniques, Principal components analysis (PCA) is designed to classify variables based on their correlations with each other. The goal of PCA is to consolidate a large number of observed variables into a smaller number of factors (components) that can be more readily interpreted as these underlying processes. It is often used as an exploratory tool to identify the major sources of air pollutant emissions [2526].

In general, principal components (PCs) are expressed by the following equation (3) where,

PCi is principal component i and

Ani is the loading (correlation coefficient) of the original variable Vn

All the statistical analyses are performed by using XLSTAT 2010.

Model description

GIS based digital elevation model.

Digital Elevation Model (DEM) has been used as an effective tool for exploring the spatial and temporal variation of air pollutants in a GIS environment. DEM is generated on the basis of sampling points stored as point layer along with the NO2 by using VEDIMINT algorithm in the Geomatica V.10.1. The output of DEM is represented as a zonation map of the NO2 which gives an idea of the spatial distribution of the NO2 in four different seasons (i.e. monsoon, post monsoon, winter and pre monsoon) over the study area.

Numerical modeling.

Dispersion model (AERMOD): It is a steady-state Gaussian plume model useful for the computation of pollutant dispersion applicable for multiple sources (point, area and volume) of emissions in rural and urban areas [22, 2728]. This model accounts for the vertical in- homogeneity of the PBL in its dispersion calculation by averaging the parameters of actual PBL into effective parameters of an equivalent homogeneous planetary boundary layer (PBL). The PBL parameters such as friction velocity, Monin—Obukhov length, convective velocity scale, temperature scale, mixing height, surface heat flux are computed by AERMET (meteorological preprocessor of AERMOD) by using local surface characteristics in the form of surface roughness and Bowen ratio in combination with standard meteorological observations (wind speed, wind direction, temperature and cloud cover). These obtained parameters are then passed through an interface present in AERMOD for calculating vertical profiles of wind speed, lateral and vertical turbulent fluctuations and potential temperature gradient. AERMOD is found to be useful for simulation of short–range (less than 50 km) pollutant dispersion especially for urban areas.

Hourly surface as well as upper air meteorological observations are required along with the emission inventory to integrate AERMOD model. But unfortunately, such kind of meteorological data which is essential for the computation of the required boundary layer parameters that serve as input to AERMOD is not available over the study area. The surface parameters and PBL parameters which serve as input in AERMOD are obtained from the output of WRF model.

Gridded emission inventory.

Preparation of emission inventory is an indispensible scientific tool for prevention of air pollution and air quality management. Quantitative understanding of the emission helps in better identification of the actual emission sources and estimation of the future emission scenario. The calculation of emissions from vehicles over the study area is performed on the basis of emission factors for different types of vehicles according to Automotive Research Association of India [29], number of vehicles of specific type, the distance travelled by a particular vehicle and their distribution based on the type of the fuel used. The estimation of vehicular emission is based on earlier works [3032], (4) where Ei is the emission of compound (i), Vehj is the number of vehicles per type (j), Dj is the distance travelled in a year per different vehicle type (j) and Ei;j;km is the emission factor of compound (i) of vehicle type (j) per driven kilometer.

The obtained gridded emission inventory serves as one of the inputs of AERMOD.

Mesoscale atmospheric model.

The Advanced Research Weather Research and Forecasting (WRF—ARW) model using Eulerian Mass Dynamical core developed by National Center for Atmospheric Research (NCAR) is a flexible, state-of-the-art atmospheric simulation system which can be used for research work in different spatial scales ranging from meters to kilometers. A detailed description of the model physics, equations and dynamics is available in Skamarock et al. [33]. In this model, eight PBL schemes are available for parameterization of the sub-grid scale turbulent vertical fluxes of heat, moisture and momentum within the PBL. Out of eight available PBL schemes of WRF model, YSU i.e. Yonsei University [34] scheme is found to give good performance over urban area and is useful for air pollution dispersion studies [35]. So this PBL scheme is used for the future projection of air pollutant in the present study. This model is run using 1° X 1° resolution with 6 hourly National Centre for Environmental Prediction (NCEP) Final Analysis (FNL) data for the initial and boundary conditions.

The mesoscale atmospheric model (WRF) and dispersion model (AERMOD) are coupled for obtaining the futuristic emission scenario over the chosen urban area.

Result and Discussion

Seasonal variation of NO2 over the study area

GIS based DEM model is used for obtaining the spatial distribution of NO2 over the urban area in four different seasons. During the monsoon season, the average concentrations of NO2 are found to be 63.795 μg/m3, 11.100 μg/m3 and 35.693 μg/m3 in Site I, Site II and Site III respectively. The concentration of NO2 lies between 10–30 μg/m3 over 59.434 km2, between 30–60 μg/m3 over 53.624 km2 and between 60–80 μg/m3 over 1.924 km2 area of the urban region as shown in Fig 2(A). Shallow PBL height, low temperature and pressure leads to accumulation of air pollutants near the earth’s surface. These combines effects meteorology and PBL result in the increase of pollutant loads near the ground in both post monsoon and winter seasons. The total area of the urban region under consideration is 114.982 km2. The DEM of spatial distribution of NO2 over the urban area for post monsoon season [Fig 2(B)] shows that the concentration of NO2 lies between 30–60 μg/m3 over 87.036 km2 and between 60–80 μg/m3 over 22.946 km2 area of the urban region under consideration. The concentration of NO2 is above the permissible limit of NAAQS (i.e. above 80 μg/m3) over an area 5.000 km2. In winter season, a large portion of the urban area (86.573 km2) is covered by the concentration range of 10–30 μg/m3. 23.38 km2 and 4.243 km2 are covered by the concentration range of 30–60 μg/m3 and 60–80 μg/m3 respectively [Fig 2(C)]. The NAAQS permissible limit for 24 h NO2 concentration is exceeded over an area of 0.786 km2. The average concentrations (ranges) of NO2 concentration are found to be 182.836 μg/m3 (84.490–315.651 μg/m3), 60.605 μg/m3 (35.173–98.604 μg/m3) and 135.123 μg/m3 (87.786–229.204 μg/m3) at Site I, Site II and Site III respectively. The concentration of NO2 is highest over Site I followed by Site III and Site II. The pattern of spatial distribution of NO2 gradually changes with the advent of pre monsoon season [Fig 2(D)]. The average concentrations (ranges) of NO2 are found to be 93.591 μg/m3 (67.955–145.575 μg/m3), 29.149 μg/m3 (10.092–51.899 μg/m3) and 61.915 μg/m3 (34.147–129.334 μg/m3) at Site I, Site II and Site III respectively. Increase in PBL height leads to gradual dispersion of pollutants which in turns dilutes the pollutant concentration near the earth’s surface. In this season, 74.815 km2 lies in the range of 10–30 μg/m3, 36.122 km2 lies in the 30–60 μg/m3 category and 3.392 km2 lies in 60–80 μg/m3 concentration range. The concentration of NO2 is above the NAAQS permissible limit over 0.653 km2 only during pre monsoon season.

thumbnail
Fig 2.

Spatial distribution of NO2 over the urban area during (a) monsoon season, (b) post monsoon, (c) winter and (d) pre monsoon seasons.

https://doi.org/10.1371/journal.pone.0170928.g002

Comparison of concentration of NO2 with other urban areas

The NO2 level at a residential area of Kolkata, India was to be 32.500±14.200 μg/m3 whereas in the industrial area, the concentration of NO2 was 49.900±9.800 μg/m3 [7]. The average concentration of NO2 was found to be 10.70 ± 3.25 ppb (~20.116 ± 6.11 μg/m3) with a range of 0.78–38.79 ppb (~1.466–72.925 μg/m3) in an urban area of Delhi during winter period [8]. The NO2 levels in Haryana, India lies in the range of 10.600 μg/m3–83.600 μg/m3 in sensitive area and between 17.700–117.100 μg/m3 in industrial region [6]. Verma et al. [5] reported the concentration of NO2 to be 38.240 μg/m3 in Lucknow. The concentration of ambient NO2 were found to be 97.645 ± 79.034 μg/m3, 95.126 ± 52.355 μg/m3 and 126.557 ± 83.245 μg/m3 in pre monsoon, post monsoon and winter seasons respectively in Burdwan, India [9]. The concentration of nitrogen dioxide was found in range of 0.02–0.08 ppm (~ 37.600–150.400 μg/m3) in Pakistan by Ali and Athar [2].

Source identification

The data of various criteria pollutants are collected and analyzed for identifying the major sources of pollutants over this urban area.

PCA is performed over the data set of two years (June 2013 –May 2015) on seasonal basis for identifying the sources of NO2 in four different seasons. The loading of variables on the component are computed for the physical interpretation of the component. An analysis of the PC loadings on the chosen variables allows the identification of the PCs as pollution sources affecting the data and this constitutes the basis of classification. The factors with eigen value more than 1 are chosen for the study as the normalized variables each carry one unit of variance. The numbers of factors (PCs) are selected such that the cumulative percentage variance explained by all the chosen factors is more than 75%. The total variance explained by various variables is 81.897 (in monsoon), 83.017 (in post monsoon), 84.536 (in winter) and 77.389 (in pre monsoon) and are given in Table 1, Table 2, Table 3 and Table 4 respectively. It is observed that vehicular emission is the major source of ambient NO2 over the urban region in all the seasons. Therefore, vehicular emission can be used for estimating the future emission scenario over the study area.

thumbnail
Table 1. Varimax rotated PC matrix for the criteria pollutants during monsoon season over the urban area.

https://doi.org/10.1371/journal.pone.0170928.t001

thumbnail
Table 2. Varimax rotated PC matrix for the criteria pollutants during post monsoon season over the urban area.

https://doi.org/10.1371/journal.pone.0170928.t002

thumbnail
Table 3. Varimax rotated PC matrix for the criteria pollutants during winter season over the urban area.

https://doi.org/10.1371/journal.pone.0170928.t003

thumbnail
Table 4. Varimax rotated PC matrix for the criteria pollutants during pre monsoon season over the urban area.

https://doi.org/10.1371/journal.pone.0170928.t004

Role of meteorology

Local meteorology plays a crucial role in the determination of concentration of NO2 over the urban region. High temperature and wind speed results in dispersion and dilution of the air pollutants. It is evident from Fig 3 that wind predominantly flows from the north–west direction in post monsoon and winter seasons. This North-west wind might be responsible for higher concentration of NO2 over this urban area during the winter and post monsoon seasons.

thumbnail
Fig 3.

Wind rose diagrams showing the variation of wind speed and wind direction over the urban area in (a) monsoon, (b) post monsoon, (c) winter and (d) pre monsoon seasons.

https://doi.org/10.1371/journal.pone.0170928.g003

Correlation analysis.

The relationships among the NO2 concentration and the meteorological parameters (i.e. temperature, relative humidity and wind speed) have been explored by correlation analysis (Table 5). It is observed that NO2 concentration and temperature hold positive correlation in monsoon (r = 0.803 and p < 0.0001) and pre monsoon (r = 0.194 and p = 0.365) seasons whereas in opposite situation prevails in post monsoon (r = -0.176 and p = 0.411) and winter (r = -0.664 and p = 0.000). Relative humidity bears significant negative relationship with NO2 level in all the seasons except monsoon. Wind speed is inversely proportional to NO2 level in monsoon (r = -0.238 and p = 0.263), post monsoon (r = -0.606 and p = 0.002), winter (r = -0.532 and p = 0.007) and pre monsoon (r = -0.675 and p = 0.000) seasons. High wind speed helps in dispersion of pollutants thereby diluting the pollutant load near the earth’s surface.

thumbnail
Table 5. Correlation coefficients (r) of NO2 concentration with respect to meteorological parameters in different seasons over the urban area.

https://doi.org/10.1371/journal.pone.0170928.t005

Regression analysis.

Linear regression analysis is implemented for understanding the influence of individual meteorological parameter on the concentration of NO2 by expressing the concentration of NO2 as function of temperature, relative humidity and wind speed separately (Table 6). This table suggests that the concentration of NO2 is mostly influenced by relative humidity (R2 = 0.384) followed by temperature and wind speed.

thumbnail
Table 6. Outcomes of regression analysis and equations of NO2 and meteorological factors (T, RH and WS).

https://doi.org/10.1371/journal.pone.0170928.t006

The regression analysis of NO2 w.r.t temperature, relative humidity and wind speed (Table 6) suggest that all the chosen meteorological parameters hold inverse relationship with the concentration of ambient NO2. Higher temperature elevates the planetary boundary layer height which in turn dilutes the pollutant level near the earth’s surface.

Multiple regression analysis (step—wise) has been performed on the NO2 data set. In the analysis, concentration of NO2 (CNO2) is assumed as the dependent variable whereas temperature, relative humidity and wind speed are considered as the independent variables. The proposed equation suggests that the chosen meteorological parameters (temperature, relative humidity and wind speed) are responsible for 40.5% variation of the NO2 concentration. The constructed equation is as follows (5)

Futuristic emission scenario

Vehicular emission is found to be the major source of NO2 in the chosen urban atmosphere. So the NOx emitted from the vehicular exhaust is used for portraying the futuristic emission scenario. NOX generally refers to the mixture of NO and NO2 and NO rapidly get converted into NO2 in the atmosphere. On this basis, the model generated data of NOX have been compared with the primary and secondary data of NO2 in this work.

The NOx emission due to vehicles in the chosen urban area is estimated using registered vehicular data of RTO (Regional Transport Office), pollutant emission factors recommended by ARAI (Automotive Research Association of India), Pune and roads length for the present scenario taking 2014 as base year. On the basis of the report, the average annual growth rate of vehicles in this area is found to be 12.05%. Considering 2014 as the base year, the number of different types of vehicles has been calculated for the year 2024 and 2034. Table 7 shows the number of different vehicles in 2014, 2024 and 2034 as well as the emission factors of NOX according to ARAI [25] for different types vehicles.

thumbnail
Table 7. Various types of vehicles with emission factors of NOX according to ARAI (2007) and the number of vehicles in 2014, 2024 and 2034 over the urban area.

https://doi.org/10.1371/journal.pone.0170928.t007

Using obtained emission scenario (Table 7) and WRF generated surface and upper air data, the ground level concentrations of NOX over the study area in four different seasons (winter, pre monsoon, monsoon and post monsoon) are computed by AERMOD (Table 8). It is observed that the concentrations of NOX in 2014 obtained as per ARAI norms have exceeded the 24 h NAAQS permissible limit (i.e.80 μg/m3) in winter (116.005 μg/m3), pre monsoon (115.50 μg/m3) and post monsoon (110.96 μg/m3) seasons. The predicted values for 2024 and 2034 are quite high. It might be due to the fact average annual growth rate of the vehicle has been considered but the number of vehicles de-registered every year has not been taken into consideration due to unavailability of data. Moreover, use of modern technology and introduction of better control measures in automobiles are expected to improve efficiency of engines and reduce the emission rate of NOX thereby decreasing the concentration of ambient NOX in the coming year.

thumbnail
Table 8. NOX concentration 2014, 2024 and 2034 by using different emission scenarios over the urban area.

https://doi.org/10.1371/journal.pone.0170928.t008

Comparison of model output with field observation

The model generated concentration of NOX (μg/m3) for the year 2014 is compared with the archived data of NO2 concentration obtained from WBPCB (Durgapur unit) and the primary data recorded at three different sites during the field work in different seasons of 2014 (Fig 4).

thumbnail
Fig 4. Comparative study of the concentration of NOX/NO2 (μg/m3) in 2014 over the urban area.

https://doi.org/10.1371/journal.pone.0170928.g004

The comparative study reflects the concentrations of NOX as per ARAI norms are higher than that of data recorded by WBPCB in all the seasons of 2014 but fair agreement exists between model generated NOX data (as per ARAI norms) and primary data obtained from three different sites (especially the site with high vehicular density, Site I). It is found that the model generally overestimates the concentration of NO2 over the urban area in all the seasons (except winter season).

Conclusion

The present study highlights the spatial distribution of NO2 in different seasons over the urban area. The spatial and temporal variation of NO2 level over the urban area is a manifestation of combined effects of emission sources, meteorology and planetary boundary layer characteristics. Prediction of NOX concentration using coupled WRF–AERMOD model shows encouraging results. This works imparts an idea of future emission scenario of NO2 over the urban area. Although source identification of NO2 over the urban area shows that vehicular emission is the major source of NO2, but the industrial emission of this region also influences the ambient concentration of NO2. It is observed that model estimated concentrations of NO2 exceed the observed average concentrations in all the seasons (except winter). An understanding the season–wise sources and sinks of NO2 over this study area might improve the obtained result. Therefore, preparation of comprehensive emission inventory by considering all available polluting sources (especially including the industrial emission) of this urban area as well as incorporation of data of yearly de-registered vehicles will portray a more realistic scenario.

Acknowledgments

Authors are thankful to the University of KwaZulu-Natal for the necessary support.

Author Contributions

  1. Conceptualization: SD PS SG AC.
  2. Data curation: SD.
  3. Formal analysis: SD PS SG AC.
  4. Funding acquisition: PS.
  5. Investigation: SD.
  6. Methodology: SD PS SG AC.
  7. Resources: PS SG AC.
  8. Software: SD SG.
  9. Supervision: PS SG AC.
  10. Validation: SD SG PS AC.
  11. Writing – original draft: SD.
  12. Writing – review & editing: SD PS SG AC.

References

  1. 1. Xu C, Fan Y-N, Kan H-D, Chen R-J, Liu JH, Li Y-F, et al. The novel relationship between urban air pollution and epilepsy: A time series study. PLoS ONE 2016; 11(8): e0161992. pmid:27571507
  2. 2. Ali M, Athar M. Air pollution due to traffic, air quality monitoring along three sections of national highway N-5, Pakistan. Environmental Monitoring and Assessment 2008; 136: 219–226. pmid:17385053
  3. 3. Salem AA, Soliman AA, Haty IAE. Determination of nitrogen dioxide, sulfur dioxide, ozone, and ammonia in ambient air using the passive sampling method associated with ion chromatographic and potentiometric analyses. Air Quality Atmosphere and Health 2009; 2: 133–145.
  4. 4. Cerón-Bretón JG, Cerón-Bretón RM, Ramírez-Lara E, Rojas-Domínguez L, Vadillo-Saénz MS, Guzmán-Mar JL. Measurements of atmospheric pollutants (Aromatic Hydrocarbons, O3, NOx, NO, NO2, CO, and SO2) in ambient air of a site located at the northeast of Mexico during summer 2011. World Scientific and Engineering Academy and Society Transactions 2013; 12: 55–66.
  5. 5. Verma A., Singh SN, Shukla MK. Air quality of the Trans-Gomti area of Lucknow city, India. Bulletin of Environmental Contamination and Toxicology 2003; 70: 166–173. pmid:12478440
  6. 6. Kaushik CP, Ravindra K, Yadav K, Mehta S, Haritash AK. Assessment of ambient air quality in urban centres of Haryana (India) in relation to different anthropogenic activities and health risks. Environmental Monitoring and Assessment 2006; 122: 27–40. pmid:16897524
  7. 7. Gupta AK, Karar K, Ayoob S, John K. Spatio-temporal characteristics of gaseous and particulate pollutants in an urban region of Kolkata, India. Atmospheric Research 2008; 87: 103–115.
  8. 8. Sharma SK., Datta A, Saud T, Mandal TK, Ahammed YN, Arya BC, et al. Study on concentration of ambient NH3 and interactions with some other ambient trace gases. Environmental Monitoring and Assessment 2010; 16: 225–235.
  9. 9. Chattopadhyay S. Spatial and temporal variations of ambient air quality in Burdwan town, West Bengal, India. Ph. D. Thesis, The University of Burdwan, West Bengal, India. 2012. Available from: shodhganga.inflibnet.ac.in/bitstream/10603/21860/2/thesis.pdf
  10. 10. Dey S, Pati C, Gupta S. Measurement and analysis of surface ozone and its precursors at three different sites in an urban region in eastern India. Environica; 2014:112–120.
  11. 11. Zhao H, Wang S, Wang W, Liu R, Zhou B. Investigation of ground-level ozone and high-pollution episodes in a megacity of Eastern China. PLoS ONE 2015; 10(6): e0131878. pmid:26121146
  12. 12. Borge R, Paz D, Lumbreras J, Pérez J, Vedrenne M. Analysis of contributions to NO2 ambient air quality levels in Madrid city (Spain) through modeling. Implications for the development of policies and air quality monitoring. Journal of Geosciences and Environment Protection 2014; 2: 6–11.
  13. 13. He H, Lu W-Z, Xue Y. Prediction of particulate matter at street level using artificial neural networks coupling with chaotic particle swarm optimization algorithm. Building and Environment 2014; 78: 111–117.
  14. 14. He H, Lu W-Z, Xue Y. Prediction of particulate matters at urban intersection by using multilayer perceptron model based on principal components. Stochastic Environmental Research and Risk Assessment 2015; 29: 2107–2114.
  15. 15. Tang T-Q, Huang H-J, Shang H-Y. Influences of the driver’s bounded rationality on micro driving behavior, fuel consumption and emissions. Transportation Research Part D 2015; 41: 423–432.
  16. 16. Tang T-Q, Huang H-J, Shang H-Y. An extended macro traffic flow model accounting for the driver’s bounded rationality and numerical tests. Physica A 2017; 468: 322–333.
  17. 17. Rao ST, Sistla G, Keenan MT, Wilson JS. An evaluation of some commonly used highway dispersion models. Atmospheric Environment 1980; 20: 1095–1103.
  18. 18. Sharma N, Chaudhry KK, Chalapati Rao CV. Vehicular pollution prediction modeling: a review of highway dispersion models. Transport Reviews 2004; 24: 409–435.
  19. 19. Otte TL, Pouliot G, Pleim JE, Young JO, Schere KL, Wong DC, et al. Linking the Eta model with the Community Multiscale Air Quality (CMAQ) modeling system to build a national air quality forecasting system. Weather Forecasting 2005; 20: 367–384.
  20. 20. Kesarkar AP, Dalvi M, Kaginalkar A, Ojha A. Coupling of the weather research and forecasting model with AERMOD for pollutant dispersion modeling: A case study for PM10 dispersion over Pune, India. Atmospheric Environment 2007; 41: 1976–1988.
  21. 21. Sistla GW, Hao JY, Ku G, Kallos K, Zhang HM, Rao ST. An operational evaluation of two regional–scale ozone air quality modeling systems over the Eastern United States. Bulletin of the American Meteorological Society 2001; 82: 945–964.
  22. 22. Cimorelli AJ, Perry SG, Venkatram A, Weil JC, Paine RJ, Wilson RB, Lee RF, Peters WD, Brode RW, Paumier JO. AERMOD: Description of model formulation. US Environmental Protection Agency. 2004. EPA Report No. 454/R-03-002d: 85.
  23. 23. Jacobs MB, Hochcheiser S. Continuous sampling and ultra microdetermination of nitrogen dioxide in air. Analytical Chemistry 1958; 30: 426–428.
  24. 24. Gvozdic V, Kovac–Andric E, Brana J. Influence of meteorological factors NO2, SO2, CO and PM10 on the concentration of O3 in the urban atmosphere of Eastern Croatia. Environmental Modeling and Assessment 2011; 16: 491–501.
  25. 25. Bruno P, Caselli M, Gennaro G, Traini A. Source apportionment of gaseous atmospheric pollutants by means of an absolute principal component scores (APCS) receptor model. Fresenius Journal of Analytical Chemistry 2001; 371: 1119–1123. pmid:11798109
  26. 26. Guo H, Wang T, Louie PKK. Source apportionment of ambient non-methane hydrocarbons in Hong Kong: Application of a principal component analysis/absolute principal component scores (PCA/APCS) receptor model. Environmental Pollution 2004; 129: 489–498. pmid:15016469
  27. 27. Cimorelli AJ, Perry SG, Venkatram A, Weill JC, Paine RJ, Wilson RB, Lee RF, Peters WD, Brode RW. AERMOD: A dispersion model for industrial source applications. Part I: General model formulation and boundary layer characterization. Journal of Applied Meteorology 2005; 44: 682–693.
  28. 28. Perry SG, Cimorelli AJ, Paine RJ, Brode RW, Weil JC, Venkatram A, Wilson RB, Lee RF, Peters WD. AERMOD: a dispersion model for industrial source applications. Part II: model performance against 17 field study databases. Journal of Applied Meteorology 2005; 44: 694–708.
  29. 29. ARAI (The Automotive Research Association of India), Pune. Draft report on emission factor development for Indian vehicles as a part of ambient air quality monitoring and emission source apportionment studies. Air Quality Monitoring Project—Indian Clean Air Programme (ICAP) 2007; Project Rep No.: AFL/2006–07/IOCL/Emission Factor Project/Final Rep. Available from: http://www.cpcb.nic.in/Emission_Factors_Vehicles.pdf, 1–89.
  30. 30. Kandlikar M, Ramachandran G. The causes and consequences of particulate air pollution in urban India: a synthesis of the science. Annu Rev Energy Environ 2000; 25: 629–684.
  31. 31. Gurjar BR, Aardenne JA, Lelieveld J, Mohan M Emission estimates and trends (1990–2000) for mega city Delhi and implications. ‎Atmospheric Environment 2004; 38:5663–5681.
  32. 32. Mohan M, Bhati S, Rao A. Application of air dispersion modeling for exposure assessment from particulate matter pollution in mega city Delhi. Asia-Pacific Journal of Chemical Engineering 2011; 6: 85–94.
  33. 33. Skamarock WC, Klemp JB, Dudhia J, Gill DO, Barker DM, Dudha MG, et al. A description of the advanced research WRF Ver.30 2008. In: NCAR technical note. NCAR/TN-475STR. Meso-scale and Micro-scale meteorology divison, National Centre for Atmospheric Research, Boulder Colorado, USA, 113.
  34. 34. Hong SY, Noh Y, Dudhia J. A new vertical diffusion package with explicit treatment of entrainment processes. Monthly Weather Review 2006; 134: 2318–2341.
  35. 35. Boadh R, Satyanarayana ANV, Rama Krishna TVBPS. Sensitivity of PBL schemes of WRF-ARW model in simulating boundary layer flow parameters for its application to air pollution dispersion modeling over a tropical station. Atmósfera 2016; 29: 61–81.