Influence of coal based thermal power pl

Environ Monit Assess DOI 10.1007/s10661-011-2087-6 Influence of coal-based thermal power plants on the spatial–temporal variability of tropospheric NO2 column over India Anup K Prasad · Ramesh P Singh · Menas Kafatos Received: 12 September 2010 / Accepted: 14 April 2011 © Springer Science+Business Media B.V 2011 Abstract The oxides of nitrogen—NOx (NO and NO2 )—are an important constituent of the troposphere The availability of relatively higher spatial (0.25◦ grid) and temporal (daily) resolution data from ozone monitoring instrument (OMI) onboard Aura helps us to better differentiate between the point sources such as thermal power plants from large cities and rural areas compared to previous sensors The annual and seasonal (summer and winter) distributions shows very high mean tropospheric NO2 in specific pockets over India especially over the Indo-Gangetic plains (up to 14.2 × 1015 molecules/cm2 ) These pockets correspond with the known locations of major thermal power plants The tropospheric NO2 over India show a large seasonal variability that is also observed in the ground NO2 data The multiple regression analysis show that the influence of a unit of power plant (in gigawatts) over tropospheric NO2 (×1015 molecules/cm2 ) is around ten times compared to a unit of population (in millions) over India The OMI data show that the NO2 increases by 0.794 ± 0.12 (×1015 molecules/cm2 ; annual) per GW compared to a previous estimate of 0.014 (×1015 molecules/cm2 ) over India The increase of tropospheric NO2 per gigawatt is found to be 1.088 ± 0.18, 0.898 ± 0.14, and 0.395 ± 0.13 (×1015 molecules/cm2 ) during winter, summer, and monsoon seasons, respectively The strong seasonal variation is attributed to the enhancement or suppression of NO2 due to various controlling factors which is discussed here The recent increasing trend (2005–2007) over rural thermal power plants pockets like Agori and Korba is due to recent large capacity additions in these regions Keywords NO2 · OMI · Thermal power plants · India Introduction A K Prasad (B) · R P Singh · M Kafatos School of Earth and Environmental Sciences, Schmid College of Science, Chapman University, Orange, CA 92866, USA e-mail: aprasad@chapman.edu A K Prasad · R P Singh · M Kafatos Center of Excellence in Earth Observing, Chapman University, Orange, CA 92866, USA The coal-fired power generation capacity in India has increased tremendously since 1980s A majority of the coals fired thermal power plants have been set up in and around the northern plains of India in the last two to three decades The National Thermal Power Corporation of India (NTPC) was set up in 1975 which is currently the largest coal-fired power utility company Environ Monit Assess in India (Fig 1a) The particulate and gaseous pollution from these point sources are believed to be major culprits behind increasing anthropogenic pollution level over India (Prasad 2007; Ghude et al 2008) In India, regions around Delhi, Mumbai, and Kolkata show increasing trends due to anthropogenic emissions (van der A et al 2008; Ghude et al 2008) Further, the northern part of the Indian sub-continent, namely the IndoGangetic (IG) plains, is subjected to dense haze, fog, and smog in the winter season (December– January) affecting millions of people (Hameed et al 2000; Ramanathan and Ramana 2005; Prasad et al 2006a, b; Prasad and Singh 2007; Choudhury Fig a The growth of power generation capacity in India (NTPC, 1986–2007; source: www.ntpc.co.in), b the annual mean tropospheric NO2 (×1015 molecules/cm2 ) for period 2005–2007 over the Indian sub-continent shows major hot spots (higher NO2 ) Boxes (blue or black) show location of selected industrialized rural, industrialized urban, and urban regions The blue box shows 100 km grid and black box shows 250 km grid around these locations c Pockets of major TPPs classified into coastal and noncoastal are shown as red and orange zones Cities with >1 and 0.2–1 million human population outside these pockets of TPPs are marked as violet and blue square dots (area of 0.25 × 0.25◦ ), respectively d Population density (persons per square kilometer) over India The TPPs (>100 MW) locations are shown as white circles with radius directly proportional to the production capacity (year 2007) Environ Monit Assess et al 2007) This winter phenomenon is considered to be a relatively recent one (last two to three decades) Major sources The major sources of NOx are fossil fuel combustion (coal and petroleum products), biomass and bio-fuel burning, and natural activities such as wild fires, and emission by soil and lightning (Lamsal et al 2008, 2010) Soil and biomass burning emissions account for 22% and 14% of global surface NOx emissions (Jaegle et al 2005) Various models and ground data show an increase in NOx and ground ozone levels during 1990–1995 that are likely attributed to the biomass burning, and fossil fuel consumption in the transport sector and power plants (Saraf and Beig 2004; Brasseur et al 2006; Beig and Brasseur 2006) Kunhikrishnan et al (2004, 2006) used Global Ozone Monitoring Experiment (GOME) data and chemistry transport modeling approach to quantify uncertainty over NOx sources of emission (urban, rural, and natural sources), emission strength, chemistry, and influence on ozone The tropospheric mass of NO2 is found to be ∼35% of the stratospheric column over India with strong seasonal variations (Kunhikrishnan et al 2006) The local sources of emission (vehicular, biomass, power generation, and Industries) were found to be responsible for 60–70% of NOx in the lower troposphere while remote sources (emissions from Africa, Southeast Asia, and China) contributed ∼15–20% of NOx in the middle and upper troposphere (Kunhikrishnan et al 2006) Inventory and trend of NOx emissions India is considered to be the fourth highest (5.2 Terra grams—Tg, year 2000) country in terms of NOx (nitrogen oxides) emissions after USA (15.3 Tg), China (11.3 Tg), and Brazil (5.5 Tg) (http://geodata.grid.unep.ch) The NOx emissions over India have been projected to grow by a factor of by 2020 with the largest increase in transport sector constituting 50% of emissions compared to 1990 (van Aardenne et al 1999) Global inventories of NOx emissions show an increase over India and China during 1990–2000 (van Aardenne et al 1999; Martin et al 2003; Richter et al 2005) Based on GOME (1996–2002) and Scanning Imaging Absorption Spectrometer for Atmospheric Cartography (SCIAMACHY; 2003– 2006) measurements, an increasing trend of NO2 (1.65 ± 0.52 % per annum) was observed over India with varying magnitude of the trend over different regions (Ghude et al 2008) The groundbased measurements of the gaseous pollutants (NO2 , SO2 ) carried out by the Central Pollution Control Board (CPCB) show high (moderate to critical) level over major cities of the IG plains (CPCB 2000) In the last 30 years, the total installed capacity and the number of thermal power plants (TPPs) in India have increased ∼10 times (National Thermal Power Corporation—NTPC, http://www.ntpc.co.in) The NO2 emission from TPPs in India was estimated at 1.129 Tg based on emission coefficients for the year 2003–2004 (Chakraborty et al 2008) The sub-regional NOx emission (inventory) estimates by Garg et al (2001) were 2.63 and 3.46 Tg during year 1990 and 1995, respectively The sector-wise contribution was 32%, 30%, and 19% for transport, electric power generation, and biomass consumption, respectively Unlike transport sector which is more evenly distributed and linked with human population, Garg et al (2001) found few districts in India as hot spots for NOx emission Lin et al (2009) estimated the annual NOx emission over China to be 6.8 TgN using ozone monitoring instrument (OMI) and GOME-2 data Impact on the ecosystem Other than gaseous pollution, Jamil et al (2009) found that fly ash and heavy metal pollution in the vicinity of thermal power plants creates problems in the form of land use, health hazards, and hazards to entire ecosystems Recently, Singh and Agrawal (2010) discussed heavy atmospheric depositions nearby thermal power plants and coal mines Higher heavy metal accumulation on the surface of leaf and leaf tissue was found in the study emphasizing impact of combustion of coal in the nearby industries and power utility companies Several studies have documented higher atmospheric deposition of heavy metals affecting air Environ Monit Assess and water quality and effect of gaseous pollutants on the local plants and agriculture in the vicinity of thermal power plants near Varanasi and other parts of the world (Sharma et al 2008; Singh and Agrawal 2010; Fagbeja et al 2008; Jamil et al 2009) in tropospheric NO2 and their relationship with the recent increase in the capacity of TPPs The results are compared with previous works on the tropospheric emissions of NO2 over India, China, Europe, and USA Issues and objectives of present study Data used The knowledge of seasonal spatio-temporal distribution of NO2 is important in helping us to understand the role played by these TPPs in the tropospheric chemistry over the IG plains In earlier studies, increasing NO2 in the IG plains was attributed to increasing atmospheric pollution, mostly from the biomass burning, vehicular emissions, and industrialization Recent studies show influence of the industrial emissions from coalbased TPPs and smelters towards pollution over the Indo-Gangetic plains (Prasad et al 2006b; Prasad 2007; Ghude et al 2008) The relatively coarse spatial and low temporal resolution (relatively poor temporal sampling) of data from GOME and SCIAMACHY limit the distinction between non-point urban and rural areas compared to point sources such as TPPs Besides, the error factor is relatively higher in GOME and SCIAMACHY (∼50%) Due to better spatial resolution (0.25◦ grid) and relatively low error (15–30%) in the ozone monitoring instrument derived tropospheric NO2 , the point sources of pollution, such as TPPs, show a larger peak with a gradient from center to the periphery (Celarier et al 2008; Wenig et al 2008) Further, the relative influence of the population (a proxy for vehicular NOx emissions in cities and biomass burning in rural areas) and TPPs over the satellitederived tropospheric NO2 is not well quantified over India We have carried out a multiple linear regression to quantify relative contribution of TPPs and population to the observed NO2 over India and compared our results with the previous estimates The seasonality of the tropospheric NO2 over India is discussed taking into account the effects of other sources of emission and controlling factors that enhances or suppresses NO2 A short-term trend analyses (2005–2007) have been carried out to delineate regions showing the largest increase OMI The OMI is a part of NASA’s Earth Observing System (EOS) Aura satellite that has been in orbit since July 15, 2004 (Levelt et al 2006) The characteristics of the instrument (spectrometers) are described in Bucsela et al (2006) The OMI provides an enhanced spatial (Nadir Field of view 13 × 24 km2 ) and temporal resolution (daily global coverage) compared to its predecessors such as SCIAMACHY (Nadir Field of view 30 × 60 km2 , days) and GOME (320 × 40 km2 , days) (Boersma et al 2008; Wenig et al 2008; Sitnov 2009) The EOS Aura crosses the equator at 13:30 local time We have used level (L3) version (V003) daily OMI Aura NO2 product OMNO2E, obtained from the National Aeronautics and Space Administration (NASA) Goddard Space Flight Center (GSFC) Goddard Earth Sciences (GES) Distributed Active Archive Center (http://daac gsfc.nasa.gov) The OMNO2E is a grid product covering the entire globe at a spatial resolution of 0.25 × 0.25◦ The gridded dataset provides daily total tropospheric NO2 concentrations over the globe since September 2004 We have included cloud-free scenes with a cloud fraction of less than 30% The satellite-based NO2 retrieval algorithm makes use of the differential optical absorption spectroscopy technique (Platt 1994; Platt and Stutz 2006) to determine the slant column densities (Bucsela et al 2006, 2008; Wenig et al 2008; Celarier et al 2008; Boersma et al 2004) The retrieval algorithm for standard product (SP by NASA GSFC) is described by Bucsela et al (2006, 2008), Wenig et al (2008), and Celarier et al (2008) The near-real-time product from Netherlands Royal Meteorological Institute (KNMI) also provides NO2 products (Boersma et al 2007) These two products employ Environ Monit Assess same technique to yield NO2 slant column densities, but differ in the subsequent steps toward retrieval of stratospheric and tropospheric components (Bucsela et al 2006) The tropospheric slant column is converted to the initial tropospheric vertical column by applying the tropospheric air mass factor (AMF) (Bucsela et al 2006) In the computation of tropospheric NO2 column density, the major uncertainties are associated with the slant column density, the AMF, and with the separation of the stratosphere and troposphere (Lamsal et al 2008) The AMF calculation becomes a more important contributor to the total error in tropospheric NO2 estimation over enhanced NO2 columns due to clouds, aerosols, and surface albedo (Martin et al 2002; Boersma et al 2004; Wenig et al 2008; Vidot et al 2010) The uncertainty due to cirrus cloud properties causes a major source of error in the estimation of tropospheric NO2 columns (Vidot et al 2010) Over moderately to heavily polluted regions, the cloud fraction and albedo may lead to relative errors of 20–30% and 20–50%, respectively (Boersma et al 2004) The OMI NO2 retrieval scheme uses O2 –O2 algorithm (Acarreta et al 2004) for cloud information (Bucsela et al 2008; Lamsal et al 2008; van der A et al 2008) The effects of aerosols are not taken into account in the retrieval algorithm (Bucsela et al 2006; Wenig et al 2008) However, for urban regions, the effects due to aerosols are expected to be relatively small (∼5%) (Martin et al 2003) The validation experiment using the measurements from ground-based Brewer at NASA GSFC and OMI show good correlation (R = 0.9) The time series data from Brewer and OMI show consistent under-estimation by OMI The values from Brewer were 25% higher than the OMI data (Wenig et al 2008) The comparison of the aerosol data (aerosol optical thickness at 440 nm) from Aerosol Robotic Network (AERONET) with the observed differences between two vertical column density dataset (OMI and Brewer) showed no obvious correlation However, the errors in the estimation of AMF due to error in the estimation of surface albedo and NO2 profile may explain the differences between OMI and Brewer vertical column densities (Wenig et al 2008) It is also observed that a finer resolution OMI NO2 data show better agreement with ground station data (Wenig et al 2008) Population The population data, density (persons/km2 ) and count (total), has been obtained from Gridded Population of the World database (v3) and values represent those for year 2005 unless otherwise stated (CIESIN 2007) The database, available at the resolution of 2.5 arc-minutes, contains values that have been adjusted to match United Nations totals The current production capacity of TPPs (year 2007) has been obtained from CARMA (Carbon Monitoring for Action, http://carma.org) Tropospheric NO2 hot spots The annual mean of tropospheric NO2 column density (2005–2007) over India is shown in Fig 1b It is apparent that the northern plains show relatively higher density compared to the southern India The most prominent feature is the limitation of the high-density NO2 regions to certain pockets, to be called as the NO2 hot spots from here on (Fig 1b) An examination of the location of all major TPPs (>100 MW) in India show their presence in certain pockets (Fig 1c) that seems to overlap with almost all of the hot spots present in Fig 1b The TPP pockets, which are a union of regions of 0.5◦ radius around the known TPP locations, are grouped into two classes: coastal and non-coastal, as the inland TPPs (non-coastal) is known to show relatively higher aerosol loading compared to coastal TPPs (Prasad et al 2006b) All the TPPs that are located up to 300 km from the coastline are classified as coastal (Prasad et al 2006b) In contrast to the NO2 hot spots that are scattered all over India, the population density is relatively high over the northern plains of India with highest concentrations around mega cities such as Delhi (Table 1) Qualitatively, there is no direct correspondence between all the NO2 hot spots and major cities (Fig 1b–d) A time-series of mean monthly NO2 column density show that the highest NO2 values are observed over the pockets of TPPs (Fig 2) Throughout the year, the mean NO2 density 3.7 1.9 2.8 3.4 1.8 2.3 1.7 1.5 1.3 10.5 3.6 3.2 6.5 2.8 4.0 2.0 2.2 1.7 7.5 4.7 3.7 5.4 3.2 4.7 2.5 1.9 1.5 17.8 9.8 4.6 9.1 5.6 6.2 2.8 2.8 2.2 6.3 4.3 4.1 5.1 4.1 3.5 3.8 2.8 1.9 Max 14.2 7.4 3.7 8.1 4.4 5.2 2.8 2.8 2.1 Avg 5.8 3.7 3.3 4.5 3.0 3.5 2.6 2.1 1.6 U I+U The mean (and maximum) annual and seasonal tropospheric NO2 over 100 km grid area Avg 6.5 3.8 3.2 4.5 3.0 3.5 2.6 2.3 1.6 Seasonal variability 5.9 3.8 3.5 4.6 3.1 3.5 2.6 1.9 1.5 Avg Avg 5.2 3.6 3.3 4.5 3.0 3.6 2.6 2.0 1.5 9,109 3,910 1,860 3,408 5,500 3,324 530 – – 11 0 1,619 3,610 860 2,108 – 1,320 – – – Total cap No Total cap No 0 1.94 1.87 3.14 24.48 7.19 9.04 13.26 8.01 5.27 Agori Korba Bhatinda Delhi Dhanbad Ahmedabad Patna Hyderabad Bangalore I+R Total (millions) 250 km2 area 100 km2 area is highest over TPPs (non-coastal) followed by TPPs (coastal) The peaks in tropospheric NO2 are observed during summer (April to June) and winter (December–January) seasons, with a sharp decline during the monsoon (July to October)season The washout effect of northeast and southwest monsoon on the tropospheric NO2 density is prominent over India (Fig 2a, b) The major cities (>1 million population), which are located outside of these TPP pockets (Fig 1c), show higher values compared to that of all India estimates However, The ground-based monthly mean NO2 measurements (CPCB, 2001–2005) over mega cities such as Delhi, Mumbai (Fig 2b), and Pune (Beig et al 2007) also show similar seasonal variability 14.0 9.5 4.9 8.7 5.7 4.8 4.2 3.7 2.4 Monsoon 2005–2007 Max Avg Seasonal (tropo NO2 ) (×1015 molecules/cm2 ) Summer Winter 2005–2007 2004–2008 Max Avg Max Avg Annual (tropo NO2 ) (×1015 molecules/cm2 ) 2005 2006 2007 2005–2007 Thermal power plants (in MW) Population count 2005 Name Type of area Table Total population (year 2005, over 100 × 100 km2 area) and total installed capacity (MW) of thermal power plants around industrialized rural (I+R), industrialized urban (I+U), and urban (U) regions of the Indian sub-continent Environ Monit Assess The three-dimensional view (Fig 3) of mean tropospheric NO2 for the period 2004–2008 shows that the northern plains show relatively higher NO2 in each of these seasons—summer, winter, and monsoon The locations of the TPPs (>100 MW) are shown as black bars (Fig 3) The length of a black bar is directly proportional to the capacity of TPP The factors behind the strong seasonal variability of tropospheric NO2 and ground NO2 as observed over India is discussed in this section (Figs 2a, b and 3a–c) Winter season The highest (peak) tropospheric NO2 is found during the winter and the lowest during the monsoon season (Figs and 3a) The large NO2 values, visible as peaks in Fig 3a, during the winter season (mean of December and January, 2004– 2008) correspond to the location of major TPPs The lifetime of NO2 is higher during the winter season due to reduction in photolysis efficiency The winter season shows relatively low water vapor content, negligible rainfall, lower OH production rate, smaller day length, and low sunlight that favors build up of NO2 The lower boundary layer and calm wind during the winter season further enhance NO2 concentrations due to low dispersion The convection is almost absent or Environ Monit Assess Fig a The time-series of tropospheric NO2 (September 2004–February 2008) over pockets of TPPs (coastal and non-coastal), cities >1 and 0.2–1 million population situated outside pockets of TPPs, and all India (excluding and including TPPs pockets) The TPPs and non-TPPs areas are shown in Fig 1c b The seasonal variation of ground NO2 over Delhi and Mumbai cities (CPCB data) weak, that limit vertical transport of NO2 above the boundary layer activities are known to cause formation of NOx in the middle to upper troposphere (Tie et al 2001; Martin et al 2007) During the summer, deep convection increases the sensitivity of the upper troposphere to local emissions (40–50%) and 10– 20% by lightning (Kunhikrishnan et al 2004) The burning of biomass or agricultural waste during April–May also contributes to NOx emissions Summer season The lifetime of NO2 decreases during summer due to an increase in the photolysis frequency due to higher temperature and more sunlight The boundary layer goes up during the summer and the vertical transport of NO2 across weaker inversion layer causes more vertical dispersion Higher horizontal wind speed also causes more dispersion around source regions causing decrease in the value of peak over major point sources Thus, the summer season shows more even distribution of NO2 with lower peaks compared to the winter season (Figs and 3b) The NOx emission from soil depends on the soil temperature and is more common in arid regions (northern mid-latitudes) during summer time with maxima peak observed during afternoon (Jaegle et al 2005; Ludwig et al 2001; van der A et al 2008) The vast alluvial lands and high temperature lead to near surface emission of NOx from soil over the northern plains of India (Ludwig et al 2001) The lightning Monsoon season The heavy rainfall during the onset of a monsoon season (July) causes washout effect leading to the wet deposition of NO2 (van der A et al 2008) Thus, the monsoon season conspicuously shows low density of tropospheric NO2 (Figs and 3c) The increase in cloud cover during the monsoon season also causes substantial decrease in the cloud free pixels The middle to upper troposphere cloud layers also inhibits view of the NOx in the boundary layer that increase error in the retrieval of NO2 The selection of cloud-free scenes (cloud fraction of less than 30%) minimizes error due to cloud layers The OMI instrument has better sampling count per grid cell compared to Environ Monit Assess Environ Monit Assess Fig The three-dimensional view of seasonal mean tropospheric NO2 over the Indian sub-continent a Winter season (December + January, 2004–2008), b summer season (April + May + June, 2005–2007), and c monsoon season (July + August + September + October, 2005– 2007) The color scale (red to brown) and peaks show high tropospheric NO2 hot spots and regions The industrialized regions show higher NO2 that is proportional to the installed capacity (bar charts) The black bars are locations of thermal power plants (>100 MW capacity) The pink bars are major cities and blue bars are the location of study regions (Fig 1, Table 1) GOME or SCIAMACHY The effect of rainfall on the sudden decrease of NO2 is also evident in the surface (ground) measurements (Fig 2b) The minimum variation of NOx during the monsoon is of the order 1–3 ppbv compared to maximum variations (60–70 ppbv) during the winter season over Pune (Beig et al 2007) The seasonal variations (summer, winter, and monsoon) of surface NOx over Pune (Beig et al 2007) are found to be similar to those observed from OMI Aura The large seasonal variability of tropospheric NO2 column is also observed from satellite and CAMx model (Comprehensive Air Quality Model) observations over Europe (Zyrichidou et al 2009) The seasonal cycle of tropospheric NO2 around Moscow show maximum and minimum during September and December, respectively This implies that the dominant source of NO2 around Moscow metropolitan area is anthropogenic particularly fossil fuel combustion (Sitnov 2009) Such large seasonal variations of NO2 are also observed in the ground NO2 data measured by CPCB (Fig 2b) (Beig et al 2007) The unique topography and meteorology of the IG plains also influence the observed seasonal variability of tropospheric NO2 Further, the seasonal variability of tropospheric NO2 from sensors onboard satellites could also be partially attributed to over-estimation or under-estimation associated with seasonal changes in surface albedo, boundary layer height, cloud fraction (below-cloud estimation), and NO2 profile (Wenig et al 2008) Such errors lead to systematic errors in the calculation of air mass factor that causes under- or overestimation of the derived product (tropospheric NO2 ) (Wenig et al 2008; Bucsela et al 2008) Regression analysis A quantitative analysis is performed to look for relative influence of the total population (in millions) and all the non-coastal TPPs in a region (Fig 4a) This group of non-coastal TPPs not only varies with the installed capacity but is also located in regions of varying population density (Fig 4a) We have chosen × 1◦ grid (Fig 4a) to club all major TPP capacity in one particular grid For nearby located TPP installations, choosing a grid takes into account NO2 emissions from all installations in the grid This is necessary as choosing individual TPPs which are very close to each other will affect actual NO2 estimates from these that will negatively affect the regression analysis For instance, two power plants 2,000 MW and 200 MW, if located close to each other will show similar NO2 as the larger TPP will affect estimates for smaller TPP Therefore, for statistical regression analysis, we have performed gridding of data into × 1◦ by combining all TPPs in a grid and their combined NO2 emission There are 36 such grids (Fig 4a) over which seasonal NO2 , total capacity, and population was calculated individually over a 3-year period and the regression analysis was carried out (Table 2) The sets of response (NO2 ) and predictors (TPP and POP) were made by extracting monthly data for all the × 1◦ grid cells shown in Fig 1a incorporating all TPPs in the region The sets were divided into groups annual and seasonal (monsoon, summer, and winter) to perform multiple regression analysis The results of a multiple linear regression are given below The standard error of coefficient (SE coeff.), relative influence (ratio of coefficient), and R2 values are summarized in Table Linear model : NO2 = C + (a1 × TPP) + (a2 × POP) Annual : NO2 = 1.77 + (0.794 × TPP) + (0.0814 × POP) (1) Monsoon : NO2 = 1.14 + (0.395 × TPP) + (0.0856 × POP) (2) Environ Monit Assess Environ Monit Assess Fig (a) The × 1◦ regions, 300 km away from the coast, where TPP density (measured as capacity, year 2007) is high is selected with corresponding population count (in millions) for multiple regression (Eq 1–4) The relationship between TPP capacity and population where size of a circle (radius) is proportional to the tropospheric NO2 during (b) annual, (c) monsoon, (d) summer, and (e) winter seasons Summer : NO2 = 2.36 + (0.898 × TPP) + (0.0825 × POP) (3) Winter : NO2 = 1.95 + (1.090 × TPP) + (0.0898 × POP) (4) Where C = constant, TPP = thermal power plant capacity in gigawatts, POP = population in million, units: tropospheric NO2 (×1015 molecules/cm2 ) The linear regression over all months (annual) shows that the influence of a unit of TPP capacity (in gigawatts, GW) over the NO2 density is approximately ten times compared to a unit of population (in millions) The influence of TPP (per GW) over NO2 density is at its peak during the winter season (Table 2) During the summer season, the relative influence of TPPs (per GW) is more than ten times (10.88) that of popula- tion (per million) During the monsoon season, the influence of TPPs (per GW) decreases However, the relative influence of TPPs (per GW) is still 4.61 times that of population (per million) The changes in the relative influence (ratio of coefficients) and constant of the linear multipleregression with the season is anticipated due to influence of various other sources and meteorological factors that affect the NO2 density during different seasons The higher constant coefficient (2.36 ± 0.18) during the summer season compared to the winter season (1.95 ± 0.23) show the presence of other factors such as lightning, biomass burning, and soil emission of NOx affecting the spatial distribution of the tropospheric NO2 The distribution of TPPs capacity, population, and seasonal estimates of NO2 are shown in Fig 4b–d that show a pattern of distribution such as higher NO2 is mostly associated with the higher capacity The magnitude of NO2 , however, strongly varies with seasons (summer, winter, and monsoon; Fig 4b–d) A linear regression analysis between TPPs capacity (GW) and tropospheric NO2 (×1015 molecules/cm2 ) show a different slope with seasons with all points within the 95% predicted interval (Fig 5a–d) The increasing size of dots in Fig shows a larger population (in millions) The linear regression (Fig 5) between tropospheric Table The results of multiple linear regression (at 95% confidence interval) showing the relative influence of capacity of TPPs (in GW) and population count (in millions) over the tropospheric NO2 (×1015 molecules/cm2 ) as measured by OMI Aura Season Predictor Short name Coeff SE coeff P value R2 Ratio of coeff (TPP/POP) Annual (2005–2007) Constant TPP capacity (GW) Population (millions) Constant TPP capacity (GW) Population (millions) Constant TPP capacity (GW) Population (millions) Constant TPP capacity (GW) Population (millions) C TPP POP C TPP POP C TPP POP C TPP POP 1.7742 0.7937 0.08138 1.1383 0.3948 0.08561 2.3625 0.8977 0.08247 1.9537 1.0876 0.08979 0.1545 0.1212 0.0183 0.1636 0.1283 0.01937 0.1772 0.139 0.02098 0.2334 0.183 0.02763 0.000 0.000 0.000 0.000 0.004 0.000 0.000 0.000 0.000 0.000 0.000 0.003 63.40% 9.75 44.50% 4.61 61.40% 10.88 56.20% 12.13 Monsoon (2005–2007) Summer (2005–2007) Winter (2004–2008) There are 36 such grids (figure 4a), for each case: annual, monsoon, summer, and winter over which the regression analysis was performed Environ Monit Assess Fig The linear correlation analysis between TPP capacity and tropospheric NO2 during a annual, b monsoon, c summer, and d winter seasons The dotted lines show predicted intervals at 95% confidence interval The radius of dots is directly proportional to the population count (range 0.02–25, in millions) NO2 (×1015 molecules/cm2 ) and TPPs capacity (in GW), show similar changes in the influence of TPPs (per GW) over NO2 density during summer, winter, and monsoon seasons The relative influence of TPPs varies significantly compared to population during monsoon season (Eq 2) The possible explanation is that TPPs are a dominant source compared to population and the TPP emissions are much more concentrated (point source) compared to population; therefore, effects are more visible on TPPs emission The rain washout effect is dominant over TPP emissions as they are concentrated in time and space In a recent study based on GOME and SCIAMACHY (1996–2006), Ghude et al (2008) carried out a linear regression between the installed capacities of only ten TPPs with the tropospheric NO2 over India They reported the increase of NO2 by 0.014 (×1015 molecules/cm2 ) per GW Our regression analysis, based on OMI Aura derived tropospheric NO2 and all TPPs located in non-coastal regions (Fig 4a), show that NO2 increases by 0.794 ± 0.12 (×1015 molecules/cm2 ; annual) per GW The increase of tropospheric NO2 per gigawatt is found to be 1.088 ± 0.18, 0.898 ± 0.14, and 0.395 ± 0.13 (×1015 molecules/cm2 ) during winter, summer, and monsoon seasons, respectively (Table 2) The difference in estimates could be due to sampling of only ten TPPs out of ∼90 TPPs (>100 MW) in India and relatively poor resolution of the GOME and SCIAMACHY compared to OMI Aura The multiple linear regression analysis, in the current and previous studies, not quantitatively include the influence Environ Monit Assess of other coal-fired industries such as smelters, and other sources of NO2 emission such as biomass, soil, and lightening The installed capacity of TPPs is used for regression analysis due to non-availability of operational capacity of TPPs over India This may lead to some uncertainty in the estimates of relative influence of TPPs and population (Table 2) Besides, surface albedo, clouds, and count (number of samples per pixel) also introduce error in the estimation of tropospheric NO2 that vary with seasons Short-term trend analysis The short-term trend analysis has been performed by applying the linear fit model to each grid cell (0.25◦ ) The increase or decrease of NO2 (×1015 molecules/cm2 ) per year as depicted in Fig only implies a relative rate of change during 3-year period starting from 2005 to 2007 The increase in NO2 per year is positive (shown by red color) over parts of the IG plains and southern India (Fig 6a) The largest and significant (at 95% confidence interval) positive rate (1.327 × 1015 molecules/cm2 per annum) is observed over a 0.25◦ grid in the Agori Region This positive trend in the tropospheric NO2 can be attributed to recent large capacity additions in the region and its surroundings during period 2006–2007 (Vindhyachal, 500 MW; Birsinghpur, 500 MW; Sipat, 500 MW, Central Electrical Authority, India: http://www.cea.nic.in) Similarly, Korba Region shows an increase of tropospheric NO2 at 0.276742 (×1015 molecules/cm2 ) per annum Korba, south of Agori Region, also shows recent capacity additions (Korba East phase 1, 250 MW; phase 2, 250 MW) Since, a large change in population count is not anticipated during this period, the increasing trend over known TPP pockets like Agori and Korba is directly linked with the recent large capacity additions in these regions The northern region—around Delhi—also shows a positive trend The negative trend is observed over eastern regions that show decrease in the NO2 during Fig a The short-term linear trend (×1015 molecules/cm2 per year) of tropospheric NO2 over India (2005–2007) b The intercept (×1015 molecules/cm2 ) component of the linear regression analysis over the study region Environ Monit Assess 2005–2007 High NO2 regions during most of the year show high intercept (Fig 6b) in pockets that qualitatively correspond with the combined strength (capacity) of the point source (TPPs) in the region (Table 1) A relatively long-term data (10 years) from OMI would give more reliable estimates of trend of NO2 over India in future Discussion The distribution of tropospheric NO2 over megacities such as Delhi not only shows the coalbased TPPs but also the influence of fossil fuel burning The vehicular emission is one of the major sources of NO2 in Delhi with highest vehicular population count The tropospheric NO2 density over Delhi region cannot be explained only by vehicular emission The rural regions (200–300 persons/km2 ) with high density of TPP (>3,000 MW installed capacity) show higher NO2 compared to Delhi with ∼25 million population (2,500 persons/km2 ) and 5.6 million vehicles (Table 1) The earlier estimates or projections showing transport sector or vehicular pollution as a major source of NO2 pollution under-estimated the effects of TPPs (van Aardenne et al 1999; Garg et al 2001) The combined effect of vehicular emission and TPP emissions over mega cities such as Delhi (NO2 = 4.5 × 1015 molecules/cm2 ) cannot catch up with an industrialized rural region such as Agori with higher TPP density (NO2 = 5.8 × 1015 molecules/cm2 ) The relative influence of TPP capacity and population is clearly evident in the multiple regression analysis (Eqs 1–4, Table 2) Recently, Chakraborty et al (2008) have suggested ways to reduce emissions from TPPs in India The mitigation efforts for transport sector emissions that are well distributed across the country will be higher (Garg et al 2001) In USA, high levels of decline (negative trend) in NOx pollution have been achieved with a control on TPP emissions (Frost et al 2006; Kim et al 2006; van der A et al 2008) The satellite estimates show declining regional NOx levels during 1999– 2005 (Kim et al 2006) Larger decrease in NOx has been observed over Ohio River Valley where TPPs dominate NOx emissions [50% of total NOx emissions] A modest decrease in ozone has been observed in some regions in response to NOx emission reduction especially regions downwind of the source (Frost et al 2006) An overall decrease of 9–16% in ozone has been observed by Environmental Protection Agency using surface monitors during 1990–2004 Richter et al (2005) also observed substantial reduction in the satellite measured tropospheric NO2 during 1996–2004 over areas of Europe and USA The trend analysis, based on NO2 dataset from GOME and SCIAMACHY, shows the largest positive trend over east China (van der A et al 2006) Satellite measurements during 1996–2006 from GOME and SCIAMACHY show a strong increase (29% per year) in tropospheric NO2 in Asia, particularly China, and a pronounced decline (up to 7% per year) in Europe The reduction in tropospheric NO2 column observed from the satellite instruments (GOME-2 and OMI) during Beijing Olympic Games emphasizes influence of fossil fuel combustion on the air quality (Mijling et al 2009) Based on the worldwide experiments and strong localization of NO2 into certain pockets, we suggest that the tropospheric NO2 density over India can be controlled by strategically targeting reduction in the emission from these specific point sources Conclusions The analysis of tropospheric NO2 observations from OMI Aura over the Indian sub-continent highlights the presence and strong influence of a network of large point sources of emissions (TPPs) These pockets of coal-based TPPs and industries are found to be associated with the higher levels of tropospheric NO2 (hot spots) The space-based NO2 observations over India as well as ground observations of NO2 show strong seasonal variability which can be attributed to enhancement or suppression of NO2 by various controlling factors, transport, and input from other seasonal sources of emissions such as soil, lightening etc The multiple linear regression analysis using TPPs capacity and population count over dense TPP regions (1 × 1◦ blocks) show the influence of TPPs per GW over tropospheric Environ Monit Assess NO2 (×1015 molecules/cm2 ) The short-term trend analysis (3 years, 2005–2007) shows the largest increase in tropospheric NO2 over the rural (Agori and Korba) region that is attributed to the recent increase of power generation capacity in this area Acknowledgements We are thankful to GES DAAC for providing OMI Aura NO2 data We are thankful to Center for International Earth Science Information Network (CIESIN) and Carbon Monitoring for Action (CARMA) for providing necessary population and power plant data used in this study We also would like to acknowledge the support from Science Systems and Applications (SSAI # 2804-09-024) References Acarreta, J R., de Haan, J F., & Stammes, P (2004) Cloud pressure retrieval using the O2 –O2 absorption band at 477 nm Journal of Geophysical Research, 109, D05204 doi:10.1029/2003JD003915 Beig, G., & Brasseur, G P (2006) Influence of anthropogenic emissions on tropospheric ozone and its precursors over the Indian tropical region during a monsoon Geophysical Research Letters, 33, L07808 doi:10.1029/2005GL024949 Beig, G., Gunthe, S., & Jadhav, D (2007) Simultaneous measurements of ozone and its precursors on a diurnal scale at a semi urban site in India Journal of Atmospheric Chemistry, 57(3), 239–253 (15) Boersma, K F., et al (2007) Near-real time retrieval of tropospheric NO2 from OMI Atmospheric Chemistry Physics, 7, 2103B2118 Boersma, K F., Eskes, H J., & Brinksma, E J (2004) Error analysis for tropospheric NO2 retrieval from space Journal of Geophysical Research, 109, D04311 doi:10.1029/2003JD003962 Boersma, K F., Jacob, D J., Eskes, H J., Pinder, R W., Wang, J., & van der A, R J (2008) Intercomparison of SCIAMACHY and OMI tropospheric NO2 columns: Observing the diurnal evolution of chemistry and emissions from space Journal of Geophysical Research, 113, D16S26 doi:10.1029/2007JD008816 Brasseur, G P., et al (2006) Impact of climate change on the future chemical composition of the global troposphere Journal of Climate, 19, 3932–3951 Bucsela, E J., Celarier, E A., Wenig, M O., Gleason, J F., Veefkind, J P., Boersma, K F., et al (2006) Algorithm for NO2 vertical column retrieval from the ozone monitoring instrument IEEE Transactions on Geoscience and Remote Sensing, 44(5), 1245–1258 Special Issue on the EOS Aura Mission Bucsela, E J., Perring, A E., Cohen, R C., Boersma, K F., Celarier, E A., Gleason, J F., et al (2008) Comparison of tropospheric NO2 from in-situ aircraft measurements with near-real time and standard product data from OMI Journal of Geophysical Research, 113, D16S31 doi:10.1029/2007JD008838 Celarier, E A., et al (2008) Validation of ozone monitoring instrument nitrogen dioxide columns Journal of Geophysical Research, 113, D15S15 doi:10.1029/ 2007JD008908 Chakraborty, N., Mukherjee, I., Santra, A K., Chowdhury, S., Chakraborty, S., Bhattacharya, S., et al (2008) Measurement of CO2 , CO, SO2 , and NO emissions from coal-based thermal power plants in India Atmospheric Environment, 42, 1073B1082 Choudhury, S., Rajpal, H., Saraf, A K., et al (2007) Mapping and forecasting of North Indian winter fog: An application of spatial technologies International Journal of Remote Sensing, 28(16), 3649– 3663 CIESIN (2007) Center for International Earth Science Information Network Columbia University; United Nations Food and Agriculture Programme (FAO); and Centro Internacional de Agricultura Tropical (CIAT) 2005 Gridded Population of the World: Future Estimates (GPWFE) Palisades, NY: Socioeconomic Data and Applications Center (SEDAC), Columbia University Available at http://sedac.ciesin columbia.edu/gpw CPCB (2000) Central Pollution Control Board, 2000 Air quality status and trends in India National Ambient Air Quality Monitoring Series, NAAQMS/ 14/1999B2000 Fagbeja, M A., Chatterton, T J., Longhurst, J W S., et al (2008) Air pollution and management in the Niger Delta—Emerging issues, Conference Information: 16th International Conference on Modelling, Monitoring and Management of Air Pollution, Date: SEP, 2008 WessexInstTechnolSkiathos GREECE Air Pollution, XVI, 116, 207–216 Frost, G J., et al (2006) Effects of changing power plant NOx emissions on ozone in the eastern United States: Proof of concept Journal of Geophysical Research, 111, D12306 doi:10.1029/2005JD006354 Garg, A., Shukla, P R., Bhattacharya, S., & Dadhwal, V K (2001) Subregion (district) and sector level SO2 and NO X emissions for India: Assessment of inventories and mitigation flexibility Atmospheric Environment, 35, 703B 713 doi:10.1016/S1352-2310(00)00316-2 Ghude, S D., Fadnavis, S., Beig, G., Polade, S D., & van der A, R J (2008) Detection of surface emission hot spots, trends, and seasonal cycle from satelliteretrieved NO2 over India Journal of Geophysical Research, 113, D20305 doi:10.1029/2007JD009615 Hameed, S., et al (2000) On the widespread winter fog in northeastern Pakistan and India Geophysical Research Letters, 27(13), 1891B1894 Jaegle, L., Steinberger, L., Martin, R V., & Chance, K (2005) Global partitioning of NOx sources using satellite observations: Relative roles of fossil fuel combustion, biomass burning and soil emissions Faraday Discussions, 130, 407B–423B doi:10.1039/b502128f Jamil, S., Abhilash, P C., Singh, A., Singh, N., & Behl, H M (2009) Fly ash trapping and metal accumulating capacity of plants: Implication for green belt around Environ Monit Assess thermal power plants Landscape And Urban Planning, 92(2),136–147 Kim, S.-W., Heckel, A., McKeen, S A., Frost, G J., Hsie, E.-Y., Trainer, M K., et al (2006) Satellite observed U.S power plant NOx emission reductions and their impact on air quality Geophysical Research Letters, 33, L22812 doi:10.1029/2006GL027749 Kunhikrishnan, T., Lawrence, M G., von Kuhlmann, R., Richter, A., Ladstätter-Weißenmayer, A., & Burrows, J P (2004) Analysis of tropospheric NOx over Asia using the model of atmospheric transport and chemistry (MATCH-MPIC) and GOME-satellite observations Atmospheric Environment, 38, 581B596 Kunhikrishnan, T., Lawrence, M G., von Kuhlmann, R., Wenig, M O., Asman, W A H., Richter, A., et al (2006) Regional NOx emission strength for the Indian subcontinent and the impact of emissions from India and neighboring countries on regional O3 chemistry Journal of Geophysical Research, 111, D15301 doi:10.1029/2005JD006036 Lamsal, L N., Martin, R V., van Donkelaar, A., Celarier, E A., Bucsela, E J., Boersma, K F., et al (2010) Indirect validation of tropospheric nitrogen dioxide retrieved from the OMI satellite instrument: Insight into the seasonal variation of nitrogen oxides at northern midlatitudes Journal of Geophysical Research-Atm, 115, D05302 Lamsal, L N., Martin, R V., van Donkelaar, A., Steinbacher, M., Celarier, E A., Bucsela, E., et al (2008) Ground-level nitrogen dioxide concentrations inferred from the satellite-borne ozone monitoring instrument Journal of Geophysical Research, 113, D16308 doi:10.1029/2007JD009235 Levelt, P F., et al (2006) Science objectives of the ozone monitoring instrument IEEE Transactions on Geoscience and Remote Sensing, 44(5), 1199B1208 Lin, J.-T., McElroy, M B., & Boersma, K F (2009) Constraint of anthropogenic NOx emissions in China from different sectors: A new methodology using multiple satellite retrievals Atmospheric Chemistry Physics Discussion, 9, 19205–19241 Ludwig, J., Meixner, F X., Vogel, B., & Förster, J (2001) Soil–air exchange of nitric oxide: An overview of processes, environmental factors, and modelling studies Biogeochemistry, 52, 225B–257B Martin, R V., et al (2002) An improved retrieval of tropospheric nitrogen dioxide from GOME Journal of Geophysical Research, 107(D20), 4437 doi:10 1029/2001JD001027 Martin, R V., Jacob, D J., Chance, K., Kurosu, T P., Palmer, P I., & Evans, M J (2003) Global inventory of nitrogen oxide emissions constrained by space-based observations of NO2 columns Journal of Geophysical Research, 108(D17), 4537 doi:10.1029/2003JD003453 Martin, R V., Sauvage, B., Folkins, I., Sioris, C E., Boone, C., Bernath, P., et al (2007) Space-based constraints on the production of nitric oxide by lightning Journal of Geophysical Research, 112, D09309 doi:10.1029/2006JD007831 Mijling, B., van der A, R J., Boersma, K F., Van Roozendael, M., De Smedt, I., & Kelder, H M (2009) Reductions of NO2 detected from space during the 2008 Beijing Olympic Games Geophysical Research Letters, 36, L13801 doi:10.1029/2009GL038943 Platt, U (1994) Differential optical absorption spectroscopy (DOAS) In M Siegrist (Ed.), Air Monitoring by Spectroscopic Techniques New York: Wiley Platt, U., & Stutz, J (2006) Dif ferential Optical Absorption Spectroscopy (DOAS), Principle and Applications Heidelberg: Springer Prasad, A K (2007) Multi-sensor appraisal of aerosols, vegetation and monsoon dynamics over Indian subcontinent Ph.D Thesis, Indian Institute of Technology Kanpur, 1–238 (Unpublished) Prasad, A K., & Singh, R P (2007) Comparison of MISRMODIS aerosol optical depth over the Indo-Gangetic basin during the winter and summer seasons (2000– 2005) Remote Sensing of Environment, 107(1–2), 109– 119 Prasad, A K., Singh, R P., & Singh, A (2006a) Seasonal climatology of aerosol optical depth over the Indian sub-continent: Trend and departures in recent years International Journal of Remote Sensing, 27(12), 2323– 2329 Prasad, A K., Singh, R P., & Kafatos, M (2006b) Influence of coal based thermal power plants on aerosol optical properties in the Indo-Gangetic basin Geophysical Research Letters, 33, L05805 doi:10.1029/ 2005GL023801 Ramanathan, V., & Ramana, M V (2005) Persistent, widespread, and strongly absorbing haze over the Himalayan foothills and the Indo-Ganges plains Pure and Applied Geophysics, 162, 1609–1626 Richter, A., Burrows, J P., & Nüß, H., Granier C., Niemeier, U (2005) Increase in tropospheric nitrogen dioxide over China observed from space Nature, 437, 129–132 doi:10.1038/nature04092 Saraf, N., & Beig, G (2004) Long-term trends in tropospheric ozone over the Indian tropical region Geophysical Research Letters, 31, L05101 doi:10.1029/ 2003GL018516 Sharma, R K., Agrawal, M., & Marshall, F M (2008) Atmospheric deposition of heavy metals (Cu, Zn, Cd and Pb) in Varanasi City Environmental Monitoring and Assessment, 142(1–3), 269–278 Singh, R K., & Agrawal, M (2010) Atmospheric depositions around a heavily industrialized area in a seasonally dry tropical environment of India Environmental Pollution, 138(1), 142–152 Sitnov, S (2009) Analysis of spatial–temporal variability of tropospheric NO2 column over Moscow megapolis using OMI spectrometer (Aura satellite) data Doklady Earth Sciences, 429(2), 1511–1517 Tie, X., Zhang, R., Brasseur, G., Emmons, L., & Lei, W (2001) Effects of lightning on reactive nitrogen and nitrogen reservoir species in the troposphere Journal of Geophysical Research, 106(D3), 3167B–3178B van Aardenne, J A., Carmichael, G R., Levy, H II, Streets, D., & Hordijk, L (1999) Anthropogenic NOx Environ Monit Assess emissions in Asia in the period 1990 B 2002 Atmospheric Environment, 33, 633B–646B van der A, R J., Eskes, H J., Boersma, K F., van Noije, T P C., Van Roozendael, M., De Smedt, I., et al (2008) Trends, seasonal variability and dominant NOx source derived from a ten year record of NO2 measured from space Journal of Geophysical Research, 113, D04302 doi:10.1029/2007JD009021 van der A, R J., Peters, D H M U., Eskes, H J., Boersma, K F., Van Roozendael, M., De Smedt, I., et al (2006) Detection of the trend and seasonal variation in tropospheric NO2 over China Journal of Geophysical Research, 111, D12317 doi:10.1029/2005JD006594 Vidot, J., Jourdan, O., Kokhanosvky, A A., Szczap, F., Giraud, V., & Rozanov, V V (2010) Retrieval of tro- pospheric NO2 columns from satellite measurements in presence of cirrus: A theoretical sensitivity study using SCIATRAN and prospect application for the A-Train Journal of Quantitative Spectroscopy and Radiative Transfer, 111(4), 586–601 Wenig, M O., Cede, A M., Bucsela, E J., Celarier, E A., Boersma, K F., Veefkind, J P., et al (2008) Validation of OMI tropospheric NO2 column densities using direct-Sun mode Brewer measurements at NASA Goddard Space Flight Center Journal of Geophysical Research, 113, D16S45 doi:10.1029/2007JD008988 Zyrichidou, I., Koukouli, M E., Balis, D S., et al (2009) Satellite observations and model simulations of tropospheric NO2 columns over south-eastern Europe Atmospheric Chemistry and Physics, 9(16), 6119–6134  ... cities of the IG plains (CPCB 2000) In the last 30 years, the total installed capacity and the number of thermal power plants (TPPs) in India have increased ∼10 times (National Thermal Power Corporation—NTPC,... depositions nearby thermal power plants and coal mines Higher heavy metal accumulation on the surface of leaf and leaf tissue was found in the study emphasizing impact of combustion of coal in the nearby... 2005–2007 Thermal power plants (in MW) Population count 2005 Name Type of area Table Total population (year 2005, over 100 × 100 km2 area) and total installed capacity (MW) of thermal power plants