Why Delhi's Pollution Sources Don't Add Up

By Samarth Bansal / January 19, 2025

Note: While reporting our last week’s story on air pollution, we discovered something unexpected: even basic questions about pollution sources remain surprisingly contested. The Cutting Room is a new section on The Plank where we share important findings and unresolved questions that don’t fit our main stories but deserve attention. (Read more here.)

Every winter, Delhi’s air becomes a toxic haze, and headlines scream numbers: “40% of pollution is from stubble burning!” or “20% is from vehicles!”

These figures sound definitive.

But what if no one really knows the truth behind them? What if these numbers, rather than clarifying the problem, are just adding to the confusion?

If your goal is to really understand what’s going on, the best thing you can do is this: don’t get hung up on these percentages. Possibly, ignore them.

There are fundamental gaps in our understanding of the pollution problem—gaps so big that even the experts are still piecing together the puzzle.

I had no idea how contested this science was—and how deeply political this question is—until Anushka Mukherjee (my colleague) and I dug into the studies behind these numbers and talked to the scientists doing the actual work.

I’m sharing what we’ve learned here, hoping either someone can pick up the unresolved bits and explore further, or someone who knows can help us understand, so we can write about it.

I. The Science of Uncertainty

What makes air pollution such a fascinating puzzle is how it forces us to think differently about causation and measurement. We’re trying to understand a system where the very thing we’re measuring—particles in the air—has already lost some of the clues about where it came from.

Think about the complexity of the problem. Delhi is huge. What pollutes in Khan Market won’t be the same as what pollutes elsewhere.

Highways have different pollution signatures than quieter neighbourhoods. Sources shift with seasons, time of day, and weather—and as the city evolves over the years, so do the contributors. Some solutions might reduce certain sources, while new ones emerge.

This is why air pollution remains a complex puzzle to solve. It isn’t like tallying the number of cars on a highway, where you can eventually arrive at a precise count. With pollution, the ‘truth’ isn’t hiding behind better measurements or more data. Reality itself resists being pinned down to precise numbers.

But that’s okay. Science has ways of dealing with uncertainty. We have established methods which are used around the world on how to know what’s in our air. India is also using them—it’s just that the way we are doing it is creating more confusion than clarity.

Look at Delhi, our most studied city. You’d see polarised fights about the causes of the toxic winter haze: is it the farmers burning stubble or rich people buying more cars? Or is it just too much construction?

These debates are signs of uncertainty. Scientists resolve this in two main ways. One is like detective work with air samples—they capture the air and analyse its chemical fingerprints to trace it back to cars, factories, or dust. The other is more like accounting—they try to count everything that could cause pollution and calculate how much each might pollute.

The best studies factor in both. (I am skipping the technical details, but if you’re interested, look for receptor modelling and emission inventories.)

These are called “source apportionment” studies. There is a big push (and rightly so) to have these studies for hundreds of Indian cities so we can know where our pollution comes from.

This question is important from a prioritisation and resource allocation lens: if we know which source is hurting us the most, perhaps we can direct more energy there.

But how you study shapes what results you get.

Q: As someone who thinks in numbers, I do get the intuitive appeal of source apportionment, but I also wonder if we are over-indexing on its importance. We broadly know the key pollution sources and there is so much basic stuff to be done. Do we really need precision? What if these studies become an excuse for inaction? (Thinking about this.)

II. When Different Methods Tell Different Stories

I think what I am talking about here will make more sense if you look at the process itself. Between 2010 and 2018, Delhi had five major studies trying to figure this out. The Council on Energy, Environment and Water (CEEW) conducted an excellent review comparing these studies and explaining the differences.

Their analysis revealed the complexity of this counting. Take a look. Say you want to figure the contribution of vehicles in Delhi’s polluted air. How would you do that?

How do you estimate all vehicles of different types, their fuel usage, their age—across the entire city?

Some teams counted vehicles at 87 locations, others at 72, and one at just 10. Some counted in parking lots, others at fuel stations, others on roads. Some included tourist vehicles and mall traffic, others didn’t. Some worried about old vehicles polluting more, others just used averages.

The geography makes it even more jumbled. Some studies just looked at Delhi, but one team looked at the whole region—Delhi plus Gurgaon, Faridabad, and Ghaziabad.

Even questions that seem simple get tricky fast. If a car kicks up dust from the road, should we count that as vehicle pollution or road dust? When vehicles are stuck in traffic jams, do they pollute more? Do better roads mean less pollution?

This is why we can’t get a simple, clear answer about how much vehicles pollute Delhi’s air. When these studies tried to figure it out, one said around 20%, and another said 40%.

What this leaves us with a bunch of options to choose from. Any news story, activist, or government official can pick any of these studies to quote. Two people can argue fiercely about numbers, both certain they’re right, both with studies to back them up.

Think about how bizarre this is: I went looking for the one study everyone agrees we should use, the gold standard, with solid scientific reasons why it’s the best. It doesn’t exist.

Many researchers we spoke to said the same thing: everyone’s using different methods. We should have standardised protocols, they said, so there is a common base on which different research groups can build on. This exists in the US and EU. In the absence of that, we get stuck arguing about basics.

Until this gets fixed, we’re left with multiple versions of the truth about Delhi’s pollution. Each version makes sense on its own, but put them all together and you get a confusing picture.

And that’s exactly where the issue becomes political.

III. When Science Meets Politics

If you’re a politician tasked with cleaning your city’s air, studies showing that the problem originates elsewhere—or is driven by weather patterns—offer the perfect excuse to avoid responsibility. The lack of clarity creates a loophole for inaction.

These challenges aren’t just theoretical—they play out dramatically in Delhi. Take this sequence of events, for example.

In 2015, the Delhi government commissioned IIT Kanpur to identify major air pollution sources. This study shaped many of Delhi’s pollution policies. It came out in 2016.

But the government said they needed more: a real-time study that could tell us what exactly causes pollution at any given moment—not just from when data was collected. In 2018, Arvind Kejriwal announced plans for a “round-the-year” study. Officials talked about making this permanent if it worked.

In 2019, the government commissioned Washington University. The Environmental Minister emphasized in Feb 2020: “Without real-time data about pollution sources, we can’t create mechanisms to reduce them.”

Then a pattern began to emerge.

Washington University submitted its study in May 2020. The government had paid ₹90 lakh. But by September, a government committee raised mysterious red flags. We don’t know what: the study is public, but their objections aren’t.

In December, Deputy CM Manish Sisodia terminated the study. But the idea lived on. Officials were told to find a new agency for real-time source apportionment.

Enter IIT Kanpur again in 2021. Their study was costlier: ₹12.72 crore. They got ₹10.72 crore upfront for equipment. The study began.

Then in late 2023, plot twist: DPCC chairperson Ashwini Kumar—a BJP appointee—cancelled this study too, saying IIT Kanpur failed to justify their conclusions.

Again, the exact objections aren’t public. From what we can piece together through press reports, the core conflict was this: IIT Kanpur’s study pointed to pollution sources outside Delhi, but the DPCC chairperson saw this as AAP trying to deflect responsibility. It again became a BJP-AAP tussle and the study was halted.

In late 2024, DPCC was hunting for yet another organization to do this study.

This is six years of India’s capital city “trying” to figure out what’s polluting its air: crores spent, two studies cancelled.

Q: We couldn’t find details of the cancelled IIT Kanpur study. Did it get completed? Is a copy accessible? Did the AAP government used any of its findings to shape its action plan?

IV. The Price of Uncertainty

This sequence doesn’t just seem to be about bureaucratic inefficiency—it suggests something troubling about how scientific uncertainty can be weaponised.

But I hesitate. Because it’s a strong claim. What if the studies were genuinely flawed? Officials might have valid concerns. It’s just that the repeated cancellations and lack of transparency make it hard to trust the process.

With this disclaimer, I will take the liberty to say what this pattern is telling me: The problem isn’t just that we can’t get exact numbers — it’s that this very uncertainty becomes a tool for avoiding action and escaping accountability.

This brings us back to where we started: When you read news reports quoting specific percentages about pollution sources, remember this story. Behind every confident statistic lies a messy reality: conflicting methods, hidden objections, and (possibly) political convenience.

If you want to dive deeper, check this comprehensive review which puts together data from different studies to get ranges for the top sources: vehicle exhaust, road dust, construction dust, cooking and heating, open waste burning, and industries.

That’s it for today.

For questions, thoughts and feedback, write to me samarth@theplankmag.com.