Only 12 per cent of India’s census cities and towns have air quality monitoring stations: CSE

New analysis by Centre for Science and Environment (CSE) exposes sad state of air quality monitoring in the country

Highlights

Number of manual monitoring stations have doubled since 2010; 20-fold jump in number of real time monitoring stations — yet 47 per cent population lives outside the air quality monitoring network; 62 per cent do not have access to daily alerts on local air quality index
National monitoring grid is only 6-8 per cent of the recomemnded requirement
Less than 200 cities monitor all six key criteria pollutants
Most manual monitoring stations do not meet the minimum requirement of 104 days of monitoring a year; real time stations perform better — over 70 per cent meet the minimum data requirement for assessing air quality trends
Only Chandigarh, Delhi and Goa have full population coverage of 100 per cent
Distribute regulatory monitors equitably in unmonitored habitations. Adopt hybrid system of air quality monitoring that combines regulatory monitors with standardised air sensors network and satellite monitoring with appropriate protocols for maximum and cost effective coverage of the entire population to assess risk and inform action

New Delhi: Of India’s 4,041 census cities and towns, a mere 12 per cent have air quality monitoring systems. What’s more, only 200 of these cities monitor all six key criteria pollutants. This is when compliance with the National Ambient Air Quality Standards (NAAQS) and clean air targets under the National Clean Air Programme (NCAP) requires robust air quality monitoring. A new analysis by Centre for Science and Environment (CSE) released here today has laid bare this abysmal state of the country’s air quality monitoring network.

The analysis points out that this means nearly 47 per cent of the country’s population remains outside the maximum radius of the air quality monitoring grid (manual and real time combined), while 62 per cent is outside that of the real time monitoring network.

“Limited air quality monitoring makes it challenging to identify non-attainment status of a vast number of towns/cities and regions and also impedes effective evaluation of clean air action and improvement in air quality needed for evaluation of performance of clean air action, especially under the 15^th Finance Commission grant. More harmful PM2.5 and ozone are not considered for compliance under NCAP due to limited monitoring and data. It is necessary to ensure more equitable distribution of monitors and adoption of hybrid monitoring with a standardised and certified air sensor network and satellite-based monitoring with appropriate protocols for maximum and cost effective coverage of population to support action,” says Anumita Roychowdhury, executive director, research and advocacy, CSE.

“The current monitoring network also faces the challenge of inadequate data generation, lack of data completeness and poor quality control of monitoring. This makes air quality trend assessment difficult to establish compliance with clean air targets. The current urban monitoring grid is highly concentrated in a few big cities and there are vast areas in other regions with no monitoring. This needs to be rationalised to cover a wider population and habitats to support implementation of clean air action plans, provide information to public about the daily risks and design emergency response and longer term action,” adds Avikal Somvanshi, senior programme manager, urban lab, CSE.

The methodology

CSE’s assessment has analysed the adequacy of the air quality monitoring network and air quality data, spatial spread, population coverage and data completeness in the country. This has considered both manual monitoring under the National Air Quality Monitoring Programme (NAMP) and real time monitoring under the Continuous Ambient Air Quality Monitoring System (CAAQMS). The key focus is on adequacy and completeness of the PM2.5 data.

The study covers 883 manual stations and 409 real time stations. It has accessed and analysed publicly available data from the websites and publications of Central Pollution Control Board (CPCB) as of December 31, 2022. The extent of population coverage by the monitoring grid and the population estimates are based on the 100 m x 100 m spatial distribution of population in 2020, developed by the WorldPop research programme of the School of Geography and Environmental Sciences at the University of Southampton.

Air quality reported by a station is understood to be accurate representation of ambient air in its 2-km radius, while it can also be fairly representative of a 2-10 km radius around the station. Without major topographical or human-made features the air quality reported at a station can also be a good proxy for ambient air 10-50 km around the station. This method is drawn from the 2019 UNICEF report on assessment of coverage of children by monitoring stations in Africa that also accounts for the population overlaps.

Key findings

Number of manual monitoring stations has doubled since 2010: There were 411 manual stations operating in 2010. According to the CPCB website, currently, there are 883 operating manual stations in 379 cities/towns in 28 states and seven Union territories (UTs) of the country. The CPCB has discontinued its practice of publishing station-wise monitoring data after the NAMP 2020 report; in 2020, the NAMP report had monitoring information from about 711 stations (despite 818 stations listed on record for that year).

Number of real time monitoring stations has grown 20-fold since 2010: There are 409 real time CAAQMS stations spread across 209 cities/towns in 27 states and four UTs. Of these, 77 stations were added in 2022. As of February 22, 2023, 23 new stations and 18 new cities have been added to the network, making it 423 stations in 221 cities. It is not always possible to assess actual operating stations. For example, at least four stations have not reported any monitoring data in last few years. These include Airoli in Navi Mumbai, Bandra in Mumbai, PWD Grounds in Vijaywada and Nishant Ganj in Lucknow.

Only 476 of 4,041 cities/towns have air quality monitoring stations (manual or real time): A majority of these (267 cities) have manual stations; 98 cities have only real time stations; and 111 cities have both manual and real time stations.

Air quality monotoring grid falls short of the recommended guidelines: The monitoring capacity as of January 1, 2023 barely adds up to 6-8 per cent of the minimum recommended as per the guidelines of the Indian Standard 5182 Methods for Measurement of Air Pollution-Part 14. This recommends a minimum number of monitoring stations for cities and towns to be set up as per their population size and type and level of pollutants. The number of stations increases with rise in population, and this is based on factors defined in the standard itself.

As per the 2011 census, India has 4,041 satutory towns. The UN population database for 2020 says there are 63 census towns have more than a million inhabitants. As per the guidelines, these million-plus cities require 959 PM monitors, 643 SO₂ monitors, 630 NO₂ monitors and 320 monitors for CO and surface ozone each.

Cities with their population in the range of 100,000 and a million need 2,417 PM, 1,863 SO₂ and 2,417 NO₂ monitors and 525 monitors for CO and surface ozone, each. The requirement for small towns with population less than 100,000 is staggering and adds up to 14,172 PM, 10,629 SO₂, and 10,629 NO₂ monitors, besides 3,543 monitors for CO and surface ozone, each.

What is available is miniscule in comparison. India has only 1,187 PM10 monitors, 717 PM2.5 monitors, 1,161 SO₂ monitors, 1,185 NO₂ monitors, 406 CO monitors and 396 surface ozone monitors. A lot more will have to be added if the rural population is also considered.

47 per cent of the population lives outside the 50-km radius coverage of combined manual and real time air quality monitoring network: About 47 per cent of Indian population or about 655 million people lives outside 50 km radius of the air quality monitoring stations (NAMP and CAAQMS combined). Only 4 per cent of the population or about 50 million people lives within the immediate coverage zone of 2 km radius of the monitoring stations; 15 per cent lives in the 2-10 km radius; and about 34 per cent lives in 10-50 km radius.

Only Chandigarh, Delhi and Goa have full population coverage under the combined monitoring network: In Delhi, 26 per cent of the population lives within a 2-km radius of its 40 real time and 10 NAMP monitoring stations; 100 per cent of the population is within the 50-km radius. Chandigarh has the second best coverage with 40 per cent of its population residing within the 2-km radius of its eight manual and real time monitoring stations and 100 per cent within the 50-km radius. Haryana is able to cover 95 per cent of its population within the 50-km radius of its monitoring grid. No other state or UT has over 90 per cent population coverage (within the 50-km radius).

62 per cent of people live outside the 50-km radius coverage of real time air quality monitoring network: About 860 million people live outside the 50-km radius of real time air quality monitoring stations (CAAQMS only) – which means they do not receive the daily AQI-based health alerts. Only 2 per cent of the population lives in the immediate coverage zone (2-km radius) of real time monitoring stations; 11 per cent is within 2-10 km radius, and about 25 per cent within 10-50 km radius.

Less than 200 cities monitor all six criteria pollutants: The group of pollutants covered for NAAQS includes particulate matter less than 10 and 2.5 micron size (PM10 and PM2.5), nitrogen dioxide (NO₂), sulphur dioxide (SO₂), carbon monoxide (CO), ozone (O₃), ammonia (NH₃) and lead (Pb). These also qualify for Air Quality Index (AQI) calculations. While PM10, NO₂ and SO₂ are most widely monitored, the monitoring of PM2.5 is comparatively not as extensive and that of ozone is very limited. Monitoring of ammonia and lead and air toxins like benzene etc is very sparse.

Manual stations are mandated to monitor SO₂, NO₂, PM10 and PM2.5. Real time stations usually monitor six pollutants — SO₂, NO₂, PM10, PM2.5, CO and ozone. Many stations do not monitor all the pollutants. Over 100 manual monitoring stations have been added since 2020, but their data is not in the public domain. As per available data from CAAQMS stations (updated as of January 1, 2023), 199 cities are capable of monitoring all six criteria pollutants. Eight cities monitor only five pollutants (barring PM2.5); 102 cities monitor four pollutants – they do not have CO and O₃ monitors; 161 monitor three pollutants — they do not monitor CO, O₃ and PM2.5. There are six cities that monitor only one or two pollutants.

Only 43 districts have their whole population monitored for all six criteria pollutants: Considering the combined coverage of NAMP and CAAQMS networks, only 43 districts have 100 per cent of their pollution within the 50-km radius of each of the six criteria pollutant monitors. These districts are mostly located in Delhi-NCR. There are 237 districts where PM10 monitoring covers more than 75 per cent of the population within the 50-km radius of the combined monitoring networks. The number falls to 173 districts for PM2.5 monitoring. Monitoring of SO₂ and NO₂ covers over 75 per cent of the population within the 50-km radius of the combined monitoring networks in 235 districts and 236 districts, respectively. Numbers are much lower for CO and surface ozone monitoring, as it is only done under the CAAQMS network. For CO and ozone, 75 per cent or more population coverage is limited to 134 and 129 districts, respectively.

Less than half of the manual stations have PM2.5 monitors and only 51 stations meet the 104 days of minimum monitoring: The NAAQS requires a minimum of 104 days of monitoring at manual stations with two 24-hour monitoring sessions every week of the year. In 2020, the CPCB discontinued publishing information on number of monitoring days for stations; therefore, this analysis is based on data from station-level monitoring days published in the NAMP 2019 report. Only 315 out of 711 stations that reported data in the report had PM2.5 monitors. Out of them, only 51 met the minimum monitoring requirement of 104 days or 100 per cent data completeness; 103 stations reported data completeness of 75-99 per cent. Over half of the PM2.5 stations did not even clock 52 days of monitoring in 2019.

Half the manual stations meeting minimum requirement of 104 days of monitoring in a year are located in just two states: These two states are Odisha (14 stations) and West Bengal (12 stations). About 11 states and one UT that have manual PM2.5 monitors do not meet the minimum requirement. Other stations meeting the minimum requirement are Goa (six stations), Himachal Pradesh (five stations), Chandigarh (four stations), Karnataka (three stations), Bihar (two stations), Delhi (two stations), and one station each in Uttar Pradesh, Telangana and Meghalaya. The maximum number of manual PM2.5 monitors are located in Andhra Pradesh (47 stations) — none of them meet the minimum requirement.

Over 70 per cent of real time stations meet minimum data completeness requirement: The NAAQS has not defined minimum data completeness requirement for real time stations. The European Union requires minimum 90 per cent of hourly values to compute a valid annual average. In the US, annual data completeness is set at greater than or equal to 75 per cent of scheduled monitoring. The CAAQMS requires real time stations to report data every 15 minutes. Says Somvanshi: “If the US criteria is applied, 290 stations out of 409 that reported PM2.5 data in 2022 had data completeness of 75 per cent or more; 69 stations reported data completeness of 50-74 per cent, while 37 reported 1-49 per cent data completeness. Seven stations did not have any data.”

Minimum data completeness requirement for 24-hour average PM2.5 was met at all stations of only nine states and UTs: The CPCB requires a minimum of 16 hours of data to determine a 24-hour average. All stations of six states and three UTs met the 75 per cent valid 24-hour values in 2022. These were Assam, Himachal Pradesh, Nagaland, Punjab, Tripura, West Bengal, Chandigarh, Jammu & Kashmir and Puducherry. The maximum number of stations that did not meet the values were located in Maharashtra (20 stations), Karnataka (16 stations) and Gujarat (12 stations).

The way forward

Air quality monitoring is important for assessing the status of growing risk and the impacts of clean air action on air quality; it is also instrumental for informing people and vulnerable communities about the daily pollution levels; enabling emergency and long-term action; and supporting health impact studies. The CSE analysis recommends the following way forward to bolster the monitoring systems:

More equitable distribution of reference-grade regulatory air quality monitoring stations for wider population coverage and for reaching areas that are without a monitoring network.
Expand monitoring network for pollutants that pose higher public health risk — PM2.5 and ozone.
Implement framework for hybrid air quality monitoring system that combines a network of regulatory monitors with air sensors and satellite-based monitoring — but based on proper standardisation, certification, calibration requirements and detailed protocols.
All communities, especially the vulnerable ones, need to be covered.