HamariHawa

Report · 2 Nov 2026 · Revised 3 Nov 2026

Lahore's smog: the year-round baseline, and what changes in October

Pakistan's standard for fine particles (PM2.5) is 35 µg/m³ averaged over 24 hours, and 15 over a year. Lahore's monthly average was above 35 in every month from October 2025 to October 2026, not only in the smog season. Even from April to September 2026, its cleanest months, ground monitors averaged 40 to 59 µg/m³ a month. That is the baseline: traffic, industry, brick kilns, power plants, household fuel and rubbish burning, all year round. In October and November two things are added on top. The layer of air that smoke can mix into becomes about half as deep, and fires break out on farmland around the city after the rice harvest. October 2025's average was about two and a half times the April–September 2026 average, and November 2025's about three and a half times. The satellite record suggests the burning near Lahore is not over yet: in each of the last seven seasons, between 37% and 63% of the crop-fire detections within 100 km of the city on the Pakistani side came after 1 November.

40–59µg/m³ monthly average PM2.5 in Lahore, April–September 2026, its cleanest months
860 → 260 mdepth of air the smoke mixes into, May against November, averaged over all hours
781crop-fire detections within 40 km of Lahore, 15 Oct–1 Nov 2026
37–63%of each season's crop-fire detections within 100 km, Pakistani side, came after 1 November, 2019–2025

Check this report. Every number here is in data.json, with its source, period, place, units and how it was calculated. Fire counts are NASA satellite heat detections, not confirmed fires. This is a monitoring report: it sets out what instruments show, and says where they can't separate one source from another.

What's in the air: a quick guide

This report focuses on fine-particle pollution, PM2.5. Other gases can harm health, help form particles or ozone, or warm the climate. This guide explains those roles and what satellites can observe.

The main pollutants in this report. The sources listed are typical examples, not estimates of how much each contributes in Lahore.
PollutantWhat it isTypical sourcesRole in smog and air pollutionSeen from space?
PM2.5
fine particles
Particles 2.5 µm across or smaller, small enough to reach deep into the lungs [49]Burning of fuel and waste (engines, kilns, industry, crop and rubbish fires, household fuel); some form in the air from gases [38, 49]Fine particles that contribute to haze; the main particle measure examined in this reportNot directly. Satellites measure aerosols through the whole column of air, not the concentration at street level; surface PM2.5 can be estimated by combining satellite observations with models and ground measurements [54]. This report uses ground monitors
PM10
fine and coarse particles
Particles up to about 10 µm, including PM2.5 [49]Dust from construction sites, unpaved roads and fields, plus all the sources of PM2.5 [49]Includes PM2.5 and larger inhalable particles. Dust's contribution to winter pollution is not quantified here (see Dust, below)Satellites observe dust and other aerosols [54]; this report uses a model dust field from CAMS [43]
NO₂
nitrogen dioxide
A gas from fuel burnt at high temperatureVehicles, power plants, industrial furnaces [56]A harmful gas that contributes to the formation of particles and ozone [49, 50, 56]Sentinel-5P retrieves its tropospheric column daily [20, 55]; used in this report
SO₂
sulphur dioxide
A gas from burning fuel that contains sulphur [51]Fossil fuels burnt in power plants and industry; high-sulphur fuel in vehicles [51]A harmful gas that can form sulphate particles [51, 58]Sentinel-5P retrieves its total column [55]; not used here
Ammonia
NH₃
A gasFertiliser and manure [38]Reacts with acidic products of nitrogen and sulphur oxides to form fine particles (ammonium nitrate and ammonium sulphate) [58]The IASI satellite instrument retrieves atmospheric ammonia columns [57]; not used here
CO
carbon monoxide
A colourless, odourless gas from incomplete burning; it reduces the oxygen the blood can carry [52]Vehicles and other machinery that burn fuel [52]; firesA harmful gas and an ozone precursor; it is not particulate matter [52, 56]Sentinel-5P retrieves its total column [55]; not used here
Methane
CH₄
A gas, the main component of natural gasWaste in dumps, including Lakhodair [17]; also livestock and gas leaks [53]A greenhouse gas and an ozone precursor; it is not particulate matter [53, 56]Sentinel-5P measures it over large areas [55]; Carbon Mapper detected plumes over Lakhodair [17]
Ground-level ozone
O₃
Not emitted directly: it forms when nitrogen oxides and volatile organic compounds (VOCs) react in sunlight [50, 56]Formed from gases emitted by traffic, industry and other sources [50]A component of photochemical smog; not assessed in this report [50]Satellites retrieve ozone columns and profiles, which are not direct measurements of ground-level concentration [55]; not used here

A note on "seen from space": satellites measure a gas or particles through the whole column of air above a place, not the concentration at street level, and how much they can retrieve, and how well, varies with cloud, the ground surface and the retrieval method [54, 55].

The baseline: above 35 in every month

Bar chart of Lahore's monthly PM2.5 from October 2025 to October 2026, above 35 in every month, with a second chart showing the mixing height falling from about 860 m in May to about 190 m in December

IQAir's city history for Lahore [1], which averages its network of ground monitors, recorded a monthly average of 40 µg/m³ in June 2026, the lowest month, and 45 to 59 in the other months from April to September. From October 2025 to February 2026 the monthly average was 124 to 183, and in October 2026 it was 100. Pakistan's PM2.5 standards are 35 µg/m³ over 24 hours and 15 over a year; the World Health Organization's annual guideline is 5 [14, 15]. The 35 line on the chart is the daily standard, shown for context: these are monthly averages, and they can't tell us how many days were above it. Even the warm-season months alone averaged about 50 µg/m³, more than three times the annual standard.

All monthly figures here come from the same IQAir product, read on 14 September and 3 November 2026; the months the two readings share are identical. IQAir's 2025 annual report table gives slightly different values for the same months (for example 171 rather than 178 for December 2025) [2]; we don't mix the two.

The sources behind the baseline run all year. They include vehicles, factories, brick kilns, power plants, generators, wood and dung burnt for cooking and heating, rubbish burnt in the open and dust. Some PM2.5 also forms in the air itself from gases such as sulphur dioxide, nitrogen oxides and ammonia, including ammonia from fertiliser and manure [38]. The sections below show what satellites and published data can say about some of them; the list isn't complete.

Traffic and industry: nitrogen dioxide

Nitrogen dioxide (NO₂) comes from fuel burnt at high temperature: vehicle engines, power plants and furnaces. The European Sentinel-5P satellite measures the amount of NO₂ in the column of air between the ground and the top of the lower atmosphere (the tropospheric NO₂ column, in molecules per cm²) every day, a record that starts in 2018 [20, 45]. We averaged the same 18 days, 15 October to 1 November, in each year from 2023, the first year for which we could read the images NASA's map service publishes over Lahore.

Tropospheric NO₂ column over Lahore, average of 15 October–1 November, 2023–2026
Four maps of nitrogen dioxide over Lahore, 2023 to 2026, each showing the highest levels over the city

Every year, Lahore's NO₂ column is about 1.6 to 2 times the level over farmland to its south-west. 2026 is about 10% below 2023, but we don't read that as a trend: the values are read back from the colour bands of published images, each band spans a range of values, clouds leave a different number of usable days each year (about 6 to 7 per pixel in 2023 and 2026), and the 2024 and 2025 images have a markedly different coverage pattern, with a value for every pixel on every day, whose cause we have not established here. This limits comparison between the years. The highest levels usually sit over the city core; in 2023 the peak was in the north-west, near Shahdara and the Ravi. On HamariHawa's live map, which ranks places by how far their NO₂ column stands above their surroundings, Lahore's south-centre was third in Pakistan for the 14 days to 1 November, after Rawalpindi and Karachi [29]. That is a relative ranking of a satellite measurement, not an official ranking of the air at street level.

What NO₂ can't do: it can't separate a motorcycle from a truck, a bus or a factory chimney, and it says nothing about fuel quality.

Vehicles: what the official numbers say

Punjab's Excise department counted 7.48 million vehicles registered in Lahore on 30 September 2024 [7], including 5.23 million motorcycles (70%), 1.68 million cars, 229,000 rickshaws, 131,000 pickups and vans, 69,000 tractors, 57,000 buses and 44,000 trucks. That is 30% of all vehicles registered in Punjab, for a city with about 10% of its people, and the Urban Unit notes that many of the province's vehicles are registered in Lahore. Registrations don't measure the vehicles actually operating in the city, and we found no published count of those. Traffic police have estimated that about 15,000 heavy vehicles enter the city each day (2022) [8].

Exhaust PM2.5 from road vehicles in all of Punjab, 2020, by type (Urban Unit, Emission Inventory of Punjab 1990–2020)
Vehicle typeThousand tonnes a yearShare of exhaust
Heavy-duty vehicles and buses17.655%
Motorcycles5.818%
Cars4.314%
Light-duty vehicles4.313%

Motorcycles are most of the fleet but not most of the exhaust particles. In the European emission factors these inventories use [24], an older diesel bus or truck has a PM2.5 exhaust factor per kilometre tens of times that of a four-stroke motorcycle; that compares technology categories, not vehicles measured in Lahore. The Urban Unit's Punjab inventory for 2020 [6] puts heavy-duty vehicles and buses at 55% of road vehicles' exhaust PM2.5 across the province and motorcycles at 18%; for NOx, heavy vehicles are about two-thirds and motorcycles about 1%. In the Urban Unit's Lahore inventory [5], motorcycles account for most of transport's emissions by weight, and carbon monoxide makes up about 80% of transport's total emissions by weight; that is where the idea that motorcycles are the biggest source comes from. In the Punjab inventory, tyre and brake wear and road abrasion add about 15% on top of exhaust PM2.5 (about 13% of road traffic's total), and dust kicked up from roads is not counted at all.

Fuel matters too. Pakistan has required imported petrol to meet the Euro V standard since 2020 and imported diesel since 2021 [9]. Fuel from local refineries is a different matter: as of 2025, only one made Euro V diesel, two made Euro III and the rest made diesel with far more sulphur [10]. The refinery upgrade policy, as amended in August 2026, gives each refinery five years from signing its upgrade agreement, subject to specified extension provisions (a cure period of up to a year, force majeure and a possible further year); its tariff incentives run for seven years [11]. The federal cabinet's Economic Coordination Committee approved the upgrade agreement, with its five-year completion period, on 14 September 2026 [12]. New vehicles are required to meet only the older Euro II standard [13]. Satellites can't see any of this: what reaches them is the combined NO₂ of every engine and chimney.

Brick kilns, power plants and industry

Research datasets that map kilns from satellite images [25] place 890 brick kilns within 50 km of central Lahore on the Pakistani side, 196 of them within 25 km. We merged the two datasets and removed points within about 60 m of each other; across the whole 50 km circle about 70 pairs remain less than 100 m apart, so a few kilns may be counted twice. These are mapped kilns, not a census of kilns operating today. Kilns burn fuel within their firing zones and discharge exhaust through chimneys [48]. Fire satellites rarely detect them: only 20 detections within 40 km were at kilns from 15 October to 1 November. Whether a kiln is running or shut can't be read from these images.

OpenStreetMap shows seven oil- and gas-fired power plants within 50 km, five of them to the west and north-west around Sheikhupura, the largest being the 1,180 MW Bhikki plant [27], and two near Raiwind to the south [26]. Atlas and Saba had their power purchase agreements terminated effective 30 September 2024 [42, 47]. Plant locations and contract status don't establish which plants run today or what they emit. OpenStreetMap also shows 181 mapped industrial areas within 50 km on the Pakistani side (as of October 2026).

Rubbish dumps: methane, and no large fires seen

Satellite view of north-east Lahore showing the capped Mehmood Booti mound and the Lakhodair landfill

Mehmood Booti, a 14-million-tonne dump on Lahore's Ring Road, closed in 2016. In March 2025 BBC Urdu reported [16] that it had been covered with soil and fitted with pipes to vent its methane; the gas would be sold to nearby industry if there was enough of it, or burnt off if not. According to the same report, most of the city's rubbish, about 5,000 tonnes a day, now goes to Lakhodair, a few kilometres to the north-east.

No clear large surface fire was identified on either dump in the usable images we examined. NASA's fire satellites caught a handful of detections on the Mehmood Booti mound in the 2020–2022 seasons and none since, and the clear Sentinel-2 images of Lakhodair we checked since October 2024 show no large fire on the dump's surface. The satellites do pick up heat on several days a season within about a kilometre of both dumps; in the images we checked, these were on neighbouring sites. Small, short-lived, smoke-obscured or underground fires can be missed. Lakhodair also gives off methane: Carbon Mapper [17, 44] detected methane plumes from it on 11 of the 12 dates it observed the site (19 plumes in all, the first in June 2023 and the latest in March 2026), and estimates its emissions at about 1,560 kg of methane an hour, with a reported uncertainty of about 440 kg an hour (no confidence level is given). That figure is Carbon Mapper's persistence-adjusted source estimate, combining emission rates by observation date; it is not a continuous measurement of the whole landfill. Methane warms the climate and helps form ozone [53], but it isn't itself PM2.5, and finding methane doesn't rule out other emissions. Rubbish burnt in small open fires across the city is a different matter: those fires can be missed by satellites.

Dust

The Copernicus CAMS global model, as served by Open-Meteo, has a field labelled dust. It is the model's largest dust size bin, with particle radii of 0.9–20 µm, not total dust [23, 43]. Its daily mean at central Lahore reached 50 µg/m³ or more on nine days in October 2026: 1–3, 7–11 and 30 October. On those days the model's PM2.5 averaged about 65 µg/m³, against about 68 on the other days. In this CAMS sample, days with a higher modelled dust signal did not have higher average modelled PM2.5. This comparison does not isolate dust's contribution to measured pollution or rule out local road and construction dust, which a model with a grid of about 45 km can't resolve, and the model reads Lahore well below ground monitors. (PM10 includes particles with aerodynamic diameters up to about 10 µm, including PM2.5.)

What changes in October

1. The air gets shallower

Diagram comparing May and June, when everyday smoke mixes through about 850 m of air stirred by the sun-heated ground, with November, when a lid of warmer air holds the same smoke, plus crop smoke, in a layer about 260 m deep

Wind at Lahore is light all year, about 1.8 to 2.4 m/s as a monthly average, and October's is no calmer than September's. What changes is the height of the layer the smoke can mix into. Averaged over all hours of the day, it reaches around 830–860 m in May and June, when the sun-heated ground stirs the air; by October it is about 400 m, by November 260 m and by December 190 m. Afternoons are deeper and cold, clear nights much shallower. In an idealized picture, with the same emissions and everything else held constant, air half or a third as deep makes the smoke roughly two or three times as concentrated, before a single extra fire is lit.

This year's October was not unusual. The average mixing height over 1–26 October was about 360 m, against 320 to 490 m in 2019–2025; 2023 and 2025 were lower. IQAir's city history gives Lahore's October 2026 average as 100 µg/m³, against 124 in October 2025 [1]. The Pakistan Air Quality Initiative (PAQI) [3], comparing the 26 monitoring sites that reported in both years, found the October average down 27%, from 136.3 to 100.1 µg/m³; that figure is PAQI's, not a comparison of the chart's bars. PAQI's model estimated that this October's weather alone lowered PM2.5 by about 19%, compared with what this October's emissions would have produced under last October's weather.

2. Crop burning on the city's edge

Every day from 1 October to 1 November 2026: NASA's satellite photo of Punjab around Lahore, with that day's hotspot detections of all kinds (all satellites)

The fire counts here are crop-associated hotspot detections, "detections" for short: heat spots that NASA's satellites picked up on farmland, away from known kilns, dumps, industry and towns. That a detection is crop burning is our inference from where it is. Detections are not the same as fires: a big field fire can cover several spots, a field burning over two days counts twice, and fires that start and end between satellite passes, or burn under cloud, are missed. They are not burned area or emissions.

In the time-lapse frame, the satellites detected 135 hotspots of all kinds on 1 October and 1,538 on 1 November; 121 and 1,490 of them were crop-associated. From 15 October to 1 November, there were 781 crop-associated detections within 40 km of central Lahore, about as far as Muridke or Raiwind. The biggest share was across the Ravi in Sheikhupura district (322), then Tarn Taran (149) and Amritsar (106) on the Indian side, Nankana Sahib (77), Lahore district itself (62) and Kasur (50). Within 100 km, on both sides of the border, there were 7,130.

Crop-associated fire detections within 40 km of central Lahore, 15 October–1 November 2026, all satellites
DistrictDetections
Sheikhupura322
Tarn Taran (India)149
Amritsar (India)106
Nankana Sahib77
Lahore62
Kasur50
Gujranwala15

Compared year to year on one satellite (NOAA-20), crop-associated detections within 100 km of Lahore on the Pakistani side numbered 1,654 over these 18 days, a quarter fewer than in 2025 (2,193) and 14% fewer than in 2024 (1,925), but more than in 2019 (812) or 2023 (622). Over the same days only 72 detections within 40 km were not on farmland (in towns, at kilns, industrial hot spots and near dumps); satellites see open field fires well and enclosed burning poorly or not at all.

On the Indian side, Punjab's crop-fire detections have fallen steeply since 2021; the busiest day so far this season had 122 in the whole state on one satellite. Part of that fall may not be less burning: NASA has reported that more farmers there now burn later in the afternoon, after the satellites pass over [39, 40], so fewer detections don't necessarily mean less burning or less smoke. In six of the last seven seasons, Indian Punjab's detections peaked in the first week of November. Diwali, on 8 November this year, moves between mid-October and mid-November, while the burning follows the harvest; the fire peaks don't line up with the festival's date, and harvest timing alone can't test whether fireworks add measurably to Lahore's air.

The burning is probably not over

Crop-associated fire detections within 100 km of central Lahore, Pakistani side, NOAA-20 satellite; the last column is the share of 1 October–30 November detections that came after 1 November
Season1–14 Oct15 Oct–1 Nov2–15 Nov16–30 NovAfter 1 Nov
20192628125711,28763%
20201,0329701,30980351%
20219343,1392,74977946%
20224831,6221,20371348%
20238976221,07376555%
20248311,9251,26150839%
20253722,1931,33716537%
20265081,654……

In every season since 2019, more than a third of the NOAA-20 crop-associated detections within 100 km of Lahore on the Pakistani side came after 1 November, and the first two weeks of November alone had between 570 and 2,750. In recent years detections have ended earlier, with few left after mid-November in 2025. With November's air shallower still than October's, the next two weeks are when crop smoke and the year-round baseline are most likely to stack up.

Was there smog before?

Yes. Thick winter haze over Lahore is decades old. A study that sampled Lahore's air every sixth day from January 2007 to January 2008 found a year-long average PM2.5 of about 194 µg/m³ [31]. That was one study with its own sites and methods, so it can't be set against today's monitor networks to show a trend. Researchers were already describing denser, longer-lasting winter fog in 2008 [33]. Reliable public monitors were scarce until about 2016–2019, so earlier episodes left fewer numbers behind.

Left: share of December–January days with daily-mean visibility under 1 km at Amritsar airport rising from 0.8% in the 1970s to 41% in 2020–24. Right: share of October–November MODIS active-fire detections recorded in November, rising from 5% to over 70% in Indian Punjab and falling from about 90% to about 50–60% in Pakistani Punjab, 2002–2025

Airport weather records show when visibility got worse [30]. At Lahore's airport, days whose average visibility was under 1 km were almost absent in the October–January winters from 1992–93 to 1997–98 (none or one a winter) and became common from 1998–99, at 1 to 23 a winter and often more than five. Amritsar's airport, about 47 km away and with the longest record, shows a similar step and a further rise since: such days made up under 1% of December and January in the 1970s, 10% in the 1990s, 20% in the 2010s and 41% in 2020–24. These are unadjusted daily averages, and the airport took 5 to 7 readings a day in the 1970s and 1980s against 22 to 24 since the 2010s, which can shift a daily average on its own. A study of 82 weather stations across Pakistan and India found fog becoming more frequent in two steps, around 1990 and 1998, alongside a gradual rise in relative humidity [32]. Visibility records can't separate fog from smoke.

What changed over those decades:

What satellites can't tell you

Satellites can't give each source's share of Lahore's PM2.5. Two recent emission inventories estimate how much each source emits. The Urban Unit's 2023 emission inventory for Lahore [5] put transport at about 35% of PM2.5 emissions (397 of 1,120 tonnes a year), with crop-residue burning (371 tonnes) and waste burning (320 tonnes) next, while noting that its vehicle figures may be an overestimate. Its often-quoted 83% for transport is a share of all pollutants added together by weight, most of it carbon monoxide. PAQI's national landscape report [37], reported by Dawn in December 2025 [4], estimated Lahore's primary PM2.5 emissions as 35% from transport, 28% from heavy industry and 17% from brick kilns; Punjab's Environment Protection Agency disputed its data at the time. These are shares of what is emitted, not of what people breathe: some PM2.5 forms in the air from gases, wind carries smoke in and out, and rain and settling remove it. Inventories, ground measurements and source-apportionment studies answer different questions, and their percentages can't be converted into one another.

What the record does show is narrower but solid: Lahore's monthly average was above 35 µg/m³ in every month we examined; its NO₂ column is high every year, with no clear trend in the images from 2023 to 2026; and in October the air gets shallower and crop-fire detections rise within reach of the city. Vehicle types, fuel quality, dust, construction and street burning of rubbish need measurements on the ground.

How we did this

Sources

  1. IQAir. Air quality in Lahore: historic monthly averages (city history). Read 14 September and 3 November 2026. www.iqair.com
  2. IQAir. World's most polluted cities, 2025 city table with monthly values (2025 World Air Quality Report data). Read 14 September 2026. www.iqair.com
  3. ProPakistani. “Lahore's smog drops 27% this October, but weather did most of the work” (PAQI analysis). 2 November 2026. propakistani.pk
  4. Imran Gabol. “Transport sector emerges as key cause of air pollution” (on PAQI's national landscape report). Dawn, 10 December 2025. www.dawn.com
  5. The Urban Unit, Government of Punjab. Sectoral Emissions Inventory of Lahore. 2023. urbanunit.gov.pk
  6. The Urban Unit, Government of Punjab. Emission Inventory of Punjab 1990–2020 (Annexes B and D). 2023. urbanunit.gov.pk
  7. Excise, Taxation & Narcotics Control Department, Punjab. Total vehicles registered in Punjab as of 30 September 2024. excise.punjab.gov.pk
  8. Asif Chaudhry. “40pc of 7m registered vehicles in Lahore contributing to smog”. Dawn, 3 December 2022. www.dawn.com
  9. Khaleeq Kiani. “Govt orders all petrol, diesel imports be Euro-V compliant”. Dawn, 9 July 2020. www.dawn.com
  10. “Policy delays threaten $5b refinery investments”. The Express Tribune, 7 May 2025. tribune.com.pk
  11. Petroleum Division, Government of Pakistan. Pakistan Oil Refining Policy for Upgradation of Existing (Brownfield) Refineries, 2023, amended February 2024 and August 2026: §6.1.2.2, §6.1.3.5, §6.1.4.1. petroleum.gov.pk
  12. Press Information Department. PR No. 146, Economic Coordination Committee meeting, 14 September 2026. pid.gov.pk
  13. Associated Press of Pakistan. National Assembly climate committee briefing on vehicle emission standards. 22 October 2025. www.app.com.pk
  14. Pakistan Environmental Protection Agency. National Environmental Quality Standards for Ambient Air (SRO 1062(I)/2010). leap.unep.org
  15. World Health Organization. WHO global air quality guidelines. 22 September 2021. www.who.int
  16. Umar Draz Nangiana. Report on the Mehmood Booti landfill (in Urdu). BBC Urdu, 24 March 2025. www.bbc.com
  17. Carbon Mapper. Source record for the Lakhodair landfill, Lahore (emission estimate, persistence, plume list), via the Carbon Mapper catalog API. Read 2 November 2026. api.carbonmapper.org
  18. NASA FIRMS. VIIRS and MODIS active fire data. firms.modaps.eosdis.nasa.gov
  19. NASA Worldview and GIBS. VIIRS NOAA-20 corrected reflectance; TROPOMI tropospheric NO₂ layer. worldview.earthdata.nasa.gov
  20. Copernicus Sentinel-5P TROPOMI (ESA / European Commission). sentinels.copernicus.eu
  21. Copernicus Sentinel-2 L2A, via Element 84 Earth Search. earth-search.aws.element84.com
  22. ECMWF / Copernicus Climate Change Service. ERA5 hourly reanalysis, via the Open-Meteo historical weather API. open-meteo.com
  23. Copernicus Atmosphere Monitoring Service (CAMS) global forecasts, via the Open-Meteo air quality API. open-meteo.com
  24. European Environment Agency. EMEP/EEA air pollutant emission inventory guidebook 2023, road transport chapter. www.eea.europa.eu
  25. SentinelKilnDB (IIT Gandhinagar, CC BY-NC 4.0) and the Brick Kiln Dataset for Pakistan's IGP Region (CC BY 4.0).
  26. OpenStreetMap contributors. Power plants and industrial land use (ODbL). www.openstreetmap.org
  27. NEPRA. Tariff determination for Quaid-e-Azam Thermal Power (Bhikki), January 2023. nepra.org.pk
  28. Pakistan Bureau of Statistics. Population and Housing Census 2023 and 1998, district results. www.pbs.gov.pk
  29. HamariHawa. Nitrogen dioxide hotspots, 14 days to 1 November 2026. hamarihawa.org
  30. NOAA National Centers for Environmental Information. Global Surface Summary of the Day, stations 41641099999 (Lahore airport) and 42071099999 (Amritsar airport). www.ncei.noaa.gov
  31. Stone, E., Schauer, J., Quraishi, T. A. & Mahmood, A. Chemical characterization and source apportionment of fine and coarse particulate matter in Lahore, Pakistan. Atmospheric Environment 44(8), 2010. doi.org
  32. Syed, F. S., Körnich, H. & Tjernström, M. On the fog variability over south Asia. Climate Dynamics 39(12), 2012. doi.org
  33. Report on a Suparco study of winter fog. Dawn, 30 December 2008. www.dawn.com
  34. Agarwala, M., Bhattacharjee, S. & Dasgupta, A. Unintended consequences of Indian groundwater preservation law on crop residue burning. Ashoka University working paper. www.ashoka.edu.in
  35. Jethva, H. et al. Connecting crop productivity, residue fires, and air quality over northern India. Scientific Reports 9, 2019. www.nature.com
  36. JICA. The Project for Lahore Urban Transport Master Plan, final report. 2012. openjicareport.jica.go.jp
  37. Abid Omar, Pakistan Air Quality Initiative. Unveiling Pakistan's Air Pollution: A National Landscape Report on Health Risks, Sources and Solutions. 27 November 2025. zenodo.org
  38. World Bank. Punjab Clean Air Program (P508222): Environmental and Social Systems Assessment. 3 March 2025. documents1.worldbank.org
  39. NASA Earth Observatory. “New timing for stubble burning in India”. 15 December 2025. science.nasa.gov
  40. Jethva, H. Timing the flames: geostationary satellite detection of diurnally shifting stubble burning in northwestern India. Remote Sensing 18(10), 1506, 2026. www.mdpi.com
  41. “Toxic emissions: alternative factory fuels causing respiratory disease”. The Express Tribune, 24 April 2013. tribune.com.pk
  42. “Govt ends contracts with five IPPs to save Rs411bn”. Dawn, 11 October 2024. www.dawn.com
  43. Open-Meteo. CAMS variable mapping (CamsDomain.swift: dust = aermr06), and ECMWF parameter database, parameter 210006, dust aerosol (0.9–20 µm) mixing ratio. github.com; codes.ecmwf.int; ECMWF forum, “PM10 and PM2.5 global products” (dust bins given as radii). forum.ecmwf.int
  44. Carbon Mapper. Product guide (plume and source emission estimates, persistence). carbonmapper.org
  45. TROPOMI science team (KNMI, for ESA/Copernicus). Nitrogen dioxide data product. www.tropomi.eu
  46. “Industrialists switch to alternative power” (Express Tribune, 28 April 2012) and “Tyre burning: three steel factories reopened” (Express Tribune, June 2011). tribune.com.pk; tribune.com.pk
  47. Central Power Purchasing Agency (Guarantee) Ltd. Annual Report 2025, table of terminated power purchase agreements, printed page 42. www.cppa.gov.pk
  48. National Energy Efficiency and Conservation Authority (NEECA). “Industrial Innovation – ZigZag Technology”. 29 December 2020. neeca.gov.pk
  49. US Environmental Protection Agency. “Particulate Matter (PM) Basics”. www.epa.gov
  50. US Environmental Protection Agency. “Ground-level Ozone Basics”. www.epa.gov
  51. US Environmental Protection Agency. “Sulfur Dioxide Basics”. www.epa.gov
  52. US Environmental Protection Agency. “Basic Information about Carbon Monoxide (CO) Outdoor Air Pollution”. www.epa.gov
  53. United Nations Environment Programme and Climate and Clean Air Coalition. Global Methane Assessment: Benefits and Costs of Mitigating Methane Emissions. 2021. www.ccacoalition.org
  54. NASA Goddard Space Flight Center. “Deep Blue: Science”. earth.gsfc.nasa.gov
  55. Copernicus Data Space Ecosystem. “Sentinel-5P Level 2” (product documentation). documentation.dataspace.copernicus.eu
  56. European Environment Agency. “Sources and emissions of air pollutants in Europe”, in Air quality in Europe 2022. 24 November 2022. www.eea.europa.eu
  57. L. Clarisse et al. “The IASI NH₃ version 4 product: averaging kernels and improved consistency”. Atmospheric Measurement Techniques 16, 5009–5028, 2023. amt.copernicus.org
  58. US Environmental Protection Agency. PM2.5 Precursor Demonstration Guidance, EPA-454/R-19-004. May 2019. www.epa.gov