WELCOME TO SMOGISTAN

Smog season is here to stay.

For four long months, from October to January, a heavy pall shrouds the hills and plains, from Peshawar in the north all the way down to Larkana in the south, choking the life and spirit out of millions of hapless souls. Social media is abuzz. News articles describe the air as “methane-laden.” A WhatsApp message from Lahore reads: “It’s like inhaling directly from a car exhaust.”

In terms of scope, we have utterly shattered all records. The Air Quality Index (AQI) level — a measure of the concentration of fine-grained particulate matter in the atmosphere — deems a value of 50 or less as “good” and 300 or more as “hazardous” to health. This season we have registered AQI levels in excess of 1,000 as a matter of routine. Hospital admittances have skyrocketed. Multan even jumped the 2,000 mark.

In the visuals, the affliction is apocalyptic. In November, drone footage of Lahore went viral, the once-renowned “City of Gardens” resembling a dark Hollywood dystopia, a setting right out of Blade Runner. International media reported that the smog over Punjab was viewable from space. The satellite images captured an unnatural alien white patch, like factory chemical discharge in water, a seething living presence. 

Smog has become a global phenomenon over the last few years, pervading the world silently, affecting billions, but it is most concentrated here now — a narrow strip stretching from Dhaka in the east to Islamabad and Peshawar in the west, with the epicentre concentrated in Delhi and Lahore.

“The world has turned the corner on tobacco,” warned the World Health Organisation’s (WHO) Director General, Dr Tedros Adhanom Ghebreyesus, in 2018. “Now it must do the same for the ‘new tobacco’ — the toxic air that billions breathe every day…No one, rich or poor, can escape air pollution. It is a silent public health emergency.”

At our end, there is a flurry of activity, which is now all too predictable. There are prompt bans on burning crop stubble, fines on vehicles, mass school closures, work-from-home policies, and curbs on socialising. Think-tanks and NGOs are cycling through a crowded calendar of events, sessions and seminars on the topic.

Everyone has a statement ready. Our politicians tussle on social media. When Bilawal Bhutto shared AQI readings for various cities on X with a tongue-in-cheek invitation to Pakistanis to move to Karachi, PTI responded with the famed Marie Antoinette quote, “Let them eat cake.” Some users dragged up images of Karachi wrecked after rains.

The mainstream discourse is a litany of complaints, excuses and explanations. The official version tends to vary. Currently, as per Punjab information minister Azma Bokhari, “[T]he wind direction brings air from India into Pakistan, yet India does not seem to be taking this problem as seriously as it should.” Multiple policies and guidelines have been issued over the years but serious structural change still seems distant. 

There is cacophonous noise and activity in these four months of smog, doom and gloom, recriminations aplenty, but things are very quick to die down when the skies clear. Wash, rinse and repeat. And every winter season the smog returns, more intense and more lethal. It has been almost a decade of this menace and we have yet to even get a sense of what confronts us.

What would a serious conversation on smog look like? In this piece, we try to ground the discourse in facts and data — the precious little that we have of it — to try to get a handle on this new normal.

KILLING US SOFTLY

The cold hard numbers on air pollution are eye-opening: the WHO reports that around 99 percent of the world’s population inhales polluted air, which exceeds their prescribed safety limits. Middle- and low-income countries are worst-hit.

A robust body of evidence now exists for the heavy toll air pollution exacts on our health, a wide spectrum with a multitude of respiratory problems, heart disease, stroke, asthma and cancer. Prenatal and early exposure has been linked to IQ loss in children. Air pollution is dangerous even at low concentrations. And it is unique compared to other killers — it can amplify the harmful effects of other pollutants and health risk factors.

The leading authority on this topic is the landmark ‘Global Burden of Diseases, Injuries, and Risk Factors Study’, which has been running for 30 plus years now, tracking 88 health risk factors in over 200 countries and territories. According to their latest edition, which presents results up to 2021, the global impact of disease stands at a staggering 2.88 billion disability-adjusted life-years (DALYs) — where each DALY equates to a year of healthy life lost to disability or premature death.

Of this colossal tally, an estimated eight percent — some 230 million DALYs — are due to particulate matter air pollution. This count outpaces well-known killers such as smoking, high blood pressure, and birthing complications. It has now become official: air pollution is the world’s “leading contributor to the global disease burden.” 

The University of Chicago Energy Policy Institute, the leading air pollution tracking unit in the world, estimates that air pollution now cuts lifespans globally by 2.3 years, surpassing cigarettes and tobacco, which clock in at 2.2 years. Air pollution kills more than HIV/AIDS, tuberculosis and malaria combined. It may even be worse than war.

We lack high quality data for Pakistan but, if the pollution problem were somehow magically fixed overnight, the average Karachiite would gain an estimated 2.6 years of life expectancy, Islamabad residents 4.6 years, and residents in Lahore 5.3 years. 

With numbers like these, our current mainstream discourse on smog seems outright tame. Why is there no grand awakening? Where is the mass mobilisation, the national emergency?

“Air pollution is the greatest external threat to human health on the planet,” says Professor Michael Greenstone, renowned economist and director of the Energy Policy Institute, “and that is not widely recognised, or not recognised with the force and vigour that one might expect.”

There are several reasons for this. Unlike traditional health threats, such as smoke from cigarettes, lead in paints, or mercury in fish, it can be hard to connect the dots with air pollution. Particulate matter air pollution is largely invisible. It is hard to quantify. It is ubiquitous, it disperses rapidly. It is a slow and patient killer. The effect is disproportionate and the poor are hit hardest. 

There is also the question of geography, of infrastructure and agency: air pollution can easily travel large distances, it does not care for borders, and it can be very difficult to police.

And there is no easy fix.

IT STARTS WITH DATA

The very first thing one notices in this conversation is that we seem to be flying blind. A decade of smog and we still do not have reliable air monitoring networks, dedicated research groups, or policy frameworks. Our mainstream discourse is reminiscent of a famous quip from US management theorist Edwards Deming: “Without data, you’re just another person with an opinion.”

The conversation is an ongoing struggle. What are the real numbers for health impact in Pakistan? What is the economic cost of air pollution?  How does air pollution vary across our cities and larger regions? What are the different solutions open to us? How do these solutions compare with each other? Without solid data, we cannot answer the simplest of questions with any degree of confidence.

To give readers a flavour of what a data-driven exercise looks like, we pick apart a local air pollution dataset and tease out the insights buried within. We use publicly available air pollution data from US embassies and consulates. The US government has deployed high quality sensors at embassy locations worldwide to inform overseas citizens about air pollution conditions, and to assist local authorities and residents. To the best of our knowledge, this is the only publicly available, long-term, detailed air quality dataset for major Pakistani cities.

The embassy sensors log hourly particulate matter concentrations, specifically PM2.5 (ie particles 2.5 microns or less in diameter). An algorithm is used to convert these readings into AQI values that can help inform health-related decisions. The index is calculated based on data over a 3-to-12-hour period, depending on the variability of particulate concentration.

We examined data for Karachi, Lahore, Islamabad and Peshawar — these are the only locations in Pakistan for which US sensor data is available. We specifically focused on data for the last three years (2022-2024). 

This experiment has some very obvious limitations: the embassies only log PM2.5 readings and do not track other contaminants, such as coarse dust particles, ozone, carbon monoxide, nitrogen dioxide etc. The data is also highly localised to the embassy region and should not be considered representative of an entire city. Pollution levels can differ dramatically over large spaces. Moreover, we worked directly with AQI values. This is relatively quick and easy, but for rigour and accuracy, computations should be done on the PM concentration values.

However, the biggest challenge was the data itself. There were over 95,000 data points and over 10,000 missing values. There were periods in the dataset where entire days — even weeks — were missing from the record, ostensibly due to sensor malfunction or replacement. Likewise, there were over 2,000 invalid readings (zero values or abnormally low AQI readings). For our purposes, we interpolated missing values using a filtering technique for gaps which had 6 hours or less missing. Even then, significant gaps persist. 

A team of three undergraduate students collaborated over some 80 hours to sanitise this data. This was a significant exercise. Most AQI reporting portals (such as aqicn.org) report results using the incomplete and defective data. 

On the plus side, this effort makes a good starting point for a data-driven discussion. Numbers can help structure our thinking about pollution and smog in a way that simple visuals cannot. The results confirm our intuition in several instances. They also broadly agree with findings reported in the research literature from India. We also encounter some interesting insights, particularly for day-night cycles and rainfall.

A TALE OF FOUR CITIES 

We start with visualisations of average daily AQI readings for the period spanning January 1, 2022 - November 30, 2024. We aim for a monthly breakdown. The different colours on the horizontal lines indicate the number of good, moderate, unhealthy and hazardous days, as per the standard AQI scale for every month. The light grey represents the amount of missing days in every month.

Some trends spring right off the charts: the smog season can be clearly discerned in the crimson-purple concentrations spanning October-February in all four cities. Air pollution builds up in the winter months, with colder temperatures resulting in atmospheric inversions that trap air near to the ground. Coal, wood and biomass fuel are used to heat homes and buildings, crop stubble is burnt in the fields. These combustion emissions, together with the lack of air circulation, allow pollutants to accumulate, giving rise to the phenomenon of winter smog. Lahore is the worst hit, followed by Peshawar. Karachi is better off, likely due to the coastal geography and the sea winds.

Smog season is also progressively getting worse and more intense with every season, in all cities except Karachi. There is also a common perception that the brunt of the smog season is borne in November and December. But January fares almost as bad as December in Lahore, Islamabad and Peshawar. 2024 is the first year that Islamabad experienced significant smog. The results for January 2024 in Islamabad and Peshawar are a revelation — a dramatically worse January than the two preceding years. It remains to be seen what this January brings.

Some trends are less evident but equally disturbing: smog may be a seasonal four month phenomenon but elevated AQI levels are a year-long problem. Air pollution — unhealthy levels of it — is perennial. The green patches — the ‘good’ days — are concentrated in the summer and monsoon months and seem to be getting scarce with every year. But even here — in cities such as Lahore and Peshawar — red and purple patches (denoting ‘unhealthy’ and ‘very unhealthy’ days) are showing up in increasing numbers over the years. 

There are no really healthy periods in our cities anymore — there is bad AQI and then there’s worse AQI.

FOLLOWING THE WIND

Wind patterns play a huge role in diffusing air pollution and dispersing it across great distances. Multiple studies have shown, for instance, that air pollution from Chinese factories contribute significantly to smog in US cities, traversing the Pacific Ocean in as little as six days.

We discern the effect of wind when we look at monthly AQI averages for all four cities over 13 months, November 1, 2023-November 30, 2024.

Lahore, Islamabad and Peshawar exhibit common patterns — strong evidence for wind currents at play — where AQI peaks over the December-February period, then dips in the summer (April) and monsoon months (August) and then climbs again September onwards. Levels in Lahore in particular ascend right into the stratosphere after October. 

Things get interesting when we overlay Lahore AQI levels with those of New Delhi. The dips and the peaks synchronise almost perfectly. New Delhi lies across the border from Lahore, at a distance of 265 miles, and is often implicated in our discourse on smog. The two cities share a common geography, they are sister cities on the grand Indo-Gangetic plain, and one of the most densely populated regions in the world, along with cities such as Multan, Kolkata and Dhaka. 

Scientists call this the ‘valley effect.’ The plain acts as a kind of low-lying bowl – it runs parallel to the Himalaya mountains. Pollutants get sucked in from multiple directions. Particulate matter from vehicles and industries, along with construction debris in the cities, mixes with dust blowing in from the Thar Desert and stubble and wood burning in winters. Cold air descends from the Himalayas, effectively trapping the polluted air near the ground, a deadly hazy cocktail, a layer of smog almost a kilometre in height. 

On the other hand, in Karachi, pollution is easily dispersed due to winds from the sea. Karachi’s AQI levels closely resemble those of Mumbai, another coastal city, with significant dips into healthy territory over the monsoon period.

THE DAY NIGHT CYCLES

We find a similar pattern when we look at the day-night AQI cycles over a month, say November, 2024. If we average out the hourly changes in AQI for the month, we see that Lahore and New Delhi are again in good synchronicity: AQI levels peak in the early morning hours, likely when the weather is coldest (around 4-5am) and then again during morning rush hour (9-10 am).

The big dip is in the late afternoon, around 4-6pm. A 2021 research study of Indian cities attributes this fall to “higher PBLH allowing for vertical mixing and dilution of the surface pollutants.” PBLH refers to the height of the planet boundary layer, the lowest layer of the Earth’s atmosphere, home to most types of clouds and weather events. Air pollution concentrates in this layer. The height of this layer tends to vary (from a hundred meters to several kilometres), depending on the time of day and location, and therefore impacts the dispersal of air pollution upwards into the atmosphere.

Karachi and Mumbai again correspond in terms of dramatically lower levels than the other cities, a flat line for the most part, with a more pronounced dip in the case of Karachi. 

RAIN TO THE RESCUE

When things get dreary, we look forward to a shower of rain to clear the skies. And the aftereffects of rain during smog is often very marked, with a clear change in visibility and the emergence of blue skies. 

We investigate a few instances to see how the change manifests in AQI values. Results are mixed: for instance, on November 23-24, 2024, when it rained in Islamabad, AQI dropped significantly, as depicted in Figure X. The day-night peaks persist clearly, but the average AQI (the white line) drops one entire category from red to orange (from ‘unhealthy’ to ‘unhealthy for sensitive groups’). But it takes only about four to six days for the average to ascend to the older pre-rainfall levels. Again, this makes intuitive sense. The pollution has not stopped.

Moreover, this mitigating effect of rain may be limited to smog season. We investigated the case of monsoon rainfall in Lahore on August 1, 2024 — these were torrential rains that broke a 44-year record for the city — and we did not find a significant change reflecting in AQI levels. 

DATA TO ACTION

These results above are preliminary and not suitable for policy recommendations — but hopefully they give a bit more substance to our intuitions about smog. We hope they also provide readers a snapshot of what kind of insights we may be able to derive if we had actual high quality, real-time data from multiple sources, ie a proper monitoring network. In advanced countries, real-time monitoring is the bedrock of national policy to combat air pollution and smog.

With a proper network of distributed sensors, we could understand and track actual sources of air pollution. We could undertake targeted interventions. As an example, China routinely halts cement production in regions when AQI levels cross a certain threshold, with significant success. China has also implemented numerous industrial relocation policies for pollution-intensive industries. Several European cities have used monitoring data to plan new roads and demarcate car-free zones.

Day-night data can be used to stagger office timings, reducing traffic congestion and transport emissions. Rainfall data can feed into cost-benefit analyses for very costly and controversial solutions like artificial rain. We can use AQI data to provision our hospitals accordingly. Data-backed ad campaigns can encourage behavioural change. Citizens around the world often use hourly readings to schedule outdoor activities, when the air quality is better.

Data can also give us deeper and more fundamental — and even illuminating — insights.

For instance, when the Covid-19 pandemic hit in late 2019, Stanford University Professor Marshall Burke investigated the impact of the lockdown on air pollution levels in Wuhan, the epicentre of the pandemic and the first city to shut down. He used US embassy readings over a two month period — the same type of data we use here — and found dramatic declines in PM2.5 concentrations.

He then drew on the health research literature and calculated, using conservative estimates, that the lockdown likely saved the lives of 1,400 children and 51,700 elderly people across China. To put these numbers in context — the lockdown likely saved 20 times more lives than those lost to the Covid-19 virus. 

This is a staggering result. We locked down the world and put our lives on hold for Covid-19. But air pollution is much more deadly, yet everything continues as business as usual.

Burke takes pains to emphasise that he does not advocate endless lockdown, but only to emphasise the startling contradiction: “…the often-hidden health consequences of the status quo… the substantial costs that our current way of doing things exacts on our health and livelihoods.”

OSTRICHES AND OWLS

If there is ever to be a serious effort to combat smog and air pollution, monitoring would be the critical first step. It is impossible to come up with an informed and intelligent response without good data. 

The absence of data at this point in time is a lapse that simply fails to compute: Pakistan is the fifth most populated country in the world, a bona fide nuclear power, the menace of smog has been building up in our biggest cities for almost a decade now, we have an official Ministry of Climate Change and Environmental Coordination — this list could go on and on.

Perhaps, this is the real tragedy here. We are faced with a life and death situation that necessitates a serious and sustained long-term strategy, but our institutions and policymakers are not capable of much beyond short-term band-aid solutions. A letter-to-the-editor in Dawn from last year dubs this the “famed ostrich approach” — let’s stick our heads in the sand and hope the problem goes away on its own.

Our citizens have no clear pathways for activism. A few months ago, when Abid Omar, founder of the Pakistan Air Quality Initiative, sought out air pollution activists in Lahore for a collaborative art project to highlight the issue of smog, he was unable to find even a single one.

This season has made it abundantly clear that smog will continue to get worse. This crisis is, in a sense, also perhaps symbolic of our overall predicament as a country, with respect to our politics, our culture, our very identity as a nation. All have been severely tested in recent times. It is tempting to view smog as the physical embodiment of our darkest demons, the ones we ignore at our peril.

Unfortunately, smog season is here to stay.

Taha Ali is an associate professor at the NUST School of Electrical Engineering and Computer Science in Islamabad. X: @agrammaton

Abeha Hussain, Noor Fatima and Rida Tayyab are second-year undergrad students at the NUST School of Electrical Engineering and Computer Science.

Umair Shahid is a concerned citizen residing in Islamabad. He is working on setting up a low-cost community monitoring network in Islamabad (https://sensor.xtendum.com/list/).
He can be reached at umair.shahid@gmail.com

Published in Dawn, EOS, January 19th, 2025