Yes, there is a real gap. Standard AQI numbers are built from PM2.5 and PM10 mass only. Inside a Delhi auto-rickshaw, researchers measured black carbon at 3.6 times the ambient level and ultrafine particle count at 8.4 times ambient (Apte et al., 2011). Neither pollutant has a numeric guideline value anywhere, so no index, anywhere, reports either one.
Your phone says the air is moderate. You are stopped in traffic, level with the back of an auto-rickshaw, its exhaust pipe a foot from your front wheel.
That moderate reading is not wrong. It is just not describing the air at your face.
A 2011 study measured exactly that microenvironment and found two pollutants running far higher than the index that was supposed to be grading the day.
Key takeaways
- Inside a Delhi auto-rickshaw, black carbon ran at 3.6 times the ambient level and ultrafine particle count ran at 8.4 times ambient, measured over about 180 hours of real-time monitoring (Apte et al., 2011).
- Neither black carbon nor ultrafine particles has a numeric guideline value in WHO's own 2021 air quality guidelines, so no public index reports either one, anywhere (WHO, 2021).
- On an 8.3 km Delhi arterial, motorised two-wheelers and auto-rickshaws both ran about 30% above the ambient PM2.5 reading during morning rush hour (Goel et al., 2015).
- In Bengaluru, black carbon measured 10 micrograms per cubic metre on residential roads against 56 on highways, a 5.6 times spread no single city index number can show (Upadhya et al., 2024).
The research behind this
Peer-reviewed measurements this post is built on. Each card says what was measured, what was found, and what it can't tell you.
Delhi auto-rickshaw study
- What they measured
- About 180 hours of real-time PM2.5, black carbon and particle-number readings inside auto-rickshaws across roughly 60 trip averages in New Delhi, with simultaneous ambient monitoring outside the vehicle.
- What they found
- In-vehicle PM2.5 ran 1.5 times ambient, black carbon ran 3.6 times ambient, and particle number (the ultrafine count) ran 8.4 times ambient.
- Can't tell you: an auto-rickshaw is an open-sided three-wheeler, a fair proxy for a motorcycle but not identical, and the data is from 2008 to 2009 Delhi traffic.
Delhi 11-mode exposure study
- What they measured
- PM2.5 along a fixed 8.3 km Delhi arterial route during morning rush hour, January to May 2014, across 11 transport microenvironments including motorised two-wheelers, auto-rickshaws, buses, cars and walking.
- What they found
- On-road PM2.5 ran about 20 to 40% above ambient across every open-air mode. Motorised two-wheelers and auto-rickshaws both ran about 30% over ambient. Only an air-conditioned car and the metro beat ambient.
- Can't tell you: one arterial route and one rush-hour window, and the percentages come from the published abstract, not an independent re-check of the full dataset.
Bengaluru road-type mapping
- What they measured
- About a year of mobile monitoring of black carbon, ultrafine particle count and carbon dioxide across 150 unique kilometres of Bengaluru roads, with about 22 repeat passes per segment at 30 metre resolution.
- What they found
- Mean black carbon ran about 10 micrograms per cubic metre on residential roads, 22 on arterials, and 56 on highways, a 5.6 times spread between the lowest and highest road type.
- Can't tell you: the authors say multi-year data would be needed to confirm the pattern holds given normal day-to-day variability.
What the studies actually measured
None of these three studies set out to grade an index. They were measuring what a person in traffic actually breathes, and comparing it to the fixed station nearby.
Apte and colleagues rode along. Portable monitors ran for about 180 hours inside Delhi auto-rickshaws, open-sided three-wheelers that share a rider's exposed microenvironment. A separate station logged ambient air at the same time.
PM2.5 inside ran 1.5 times ambient. Black carbon ran 3.6 times ambient. Particle number, the ultrafine count, ran 8.4 times ambient (Apte et al., 2011).
Goel and colleagues tested whether that gap was a one-route fluke. Along a fixed 8.3 km Delhi arterial, across 11 transport modes, motorised two-wheelers and auto-rickshaws both ran about 30% above ambient PM2.5.
Only a sealed, air-conditioned car and the metro beat the ambient number (Goel et al., 2015).
Upadhya and colleagues then asked whether the gap moves with the road rather than the vehicle. A year of mobile mapping across Bengaluru found black carbon at 10 micrograms per cubic metre on quiet residential streets, climbing to 56 on highways.
That is a 5.6 times spread across roads a single commute can cross in minutes (Upadhya et al., 2024).
Put together, the three studies measure the same shape from three different angles: vehicle type, road type and route. The air at a rider's face runs consistently higher than the number describing the city around them.
Ride with filtered air
Filters haze PM2.5 on every ride
Easi Breezi clips onto any helmet and feeds you H11 HEPA-filtered air. Bike-powered, no batteries, no charging.
Delhi's moderate reading measures two pollutants, not the ones riding next to you
Delhi's official daily bulletins run on the Central Pollution Control Board's composite AQI, which is built from PM10 and PM2.5 sub-indices only (CPCB).
In the first week of October 2026, independent analysis of those same official bulletins put the average AQI at 171, the highest opening week of October since 2023 (CREA, 2026).
Individual readings still landed in the moderate band, weeks ahead of the city's usual November and December smog season.
A moderate reading is not a false one. It is an honest answer to a narrow question: how much PM10 and PM2.5 mass sat at a handful of fixed stations over the last 24 hours.
It was never built to answer a different question: what is inside the air pocket around a rider's helmet, stopped behind an auto-rickshaw, idling at a red light.
A moderate city-wide AQI does not rule out a bad fifteen minutes stuck in dense two-wheeler traffic. Treat idling behind an open-sided vehicle as its own exposure, not an extension of the day's average.
Your vents let in exactly what the index can't see
A passive-vent helmet moves air by forward motion. It has no filtration stage, so whatever concentration sits in the rider's air pocket enters unfiltered. The auto-rickshaw numbers describe that exact air pocket: an open-sided vehicle, surrounded by traffic, with no sealed cabin between the rider and the exhaust beside them.
The reason sits in how a standard AQI is built, not in where any one station sits.
Why a mass-based monitor misses most of the particle count
Schematic, illustrative mechanism, not a measured dataset. The underlying numbers come from Apte et al., 2011 and WHO's 2021 review.
PM2.5 mass and particle count are different measurements of the same air. Black carbon and ultrafine particles contribute almost nothing to PM2.5 mass, since each particle is tiny, but they make up the overwhelming majority of particle count and combustion toxicity (Upadhya et al., 2024).
WHO reviewed the evidence in 2021 and stated it plainly: there is insufficient quantitative evidence to set a numeric guideline for black carbon or ultrafine particles. Only general good-practice statements exist, no index-ready number (WHO, 2021).
A rider idling behind a truck or an auto-rickshaw is, physically, inside the exact microenvironment Apte's team measured. Not behind it. Not near it. Inside it. The number on their phone describes a station kilometres away, measuring neither of the two pollutants concentrated in that pocket of air.
Why the blind spot will not close by 2031
WHO's air quality guidelines took over 15 years to move from the 2005 edition to the 2021 update, and that update still left black carbon and ultrafine particles in the insufficient-evidence category (WHO, 2021).
This is a projection, not an official forecast. If that revision cadence holds, a rider in 2031 will very likely still be reading an index built from PM2.5 and PM10 alone.
The gap Apte measured in 2008 and 2009 Delhi traffic, 3.6 times ambient for black carbon and 8.4 times for ultrafine count, will still be real. It will still be invisible to the number on a phone screen.
Upadhya's Bengaluru data shows the gap moves with geography as much as with time. A rider crossing from a residential street onto a highway inside the same commute crosses a 5.6 times swing in black carbon.
No single city-wide index number, now or in 2031, is built to represent that swing at the road segment a rider actually rides (Upadhya et al., 2024).
Two-wheeler traffic keeps growing across the cities this gap is largest in. The index measuring that traffic is not scheduled to change at anything close to that pace.
Active filtration does not need the index to catch up
Easi Breezi's clip-on filter sits directly in the intake path and removes particulate by physical capture, not by reading a number first. It filters whatever is actually in the air a rider is breathing, including the black carbon and ultrafine fraction a PM2.5 mass index was never built to report.
It does not need a government feed, a satellite model or a fixed station to tell it the air is bad. It filters what is actually arriving at a rider's face, not what the last bulletin said about the whole city.
The same gap shows up outside Delhi. Hanoi riders breathe black carbon at levels no index reports either (read that study here).
The broader question of whether official air data can be trusted at all runs through an earlier post on a country that lost its only public monitor.
Who can you trust right now?
No single source replaces a dense government network, but these help you cross-check what you have, or fill in for a pollutant the index skips entirely.
- IQAir's AirVisual map pools government and independent stations worldwide, useful for comparing a nearby town's reading against your own. It still reports PM2.5 and PM10 only, so the black carbon and ultrafine gap stays invisible there too.
- Sensor.Community is a volunteer low-cost sensor network that runs far denser in some cities than any government grid, though individual sensors can drift without regular calibration checks.
- PurpleAir shows real-time readings from thousands of consumer sensors. Raw PurpleAir numbers usually read high and need a correction factor before comparing with an official AQI.
- OpenAQ aggregates open government and research-grade data worldwide. Coverage in any one country depends entirely on who has chosen to publish there.
- Easi Breezi's live haze page at easibreezi.com/pages/haze pulls a modelled, more localised estimate. It is a model estimate, not an official AQI reading, and says so on the page.
None of these five report black carbon or ultrafine particle count either. That gap does not close by switching sources. It closes by filtering at your own intake, regardless of what any station anywhere is reporting that day.
Frequently asked questions
Does the AQI include black carbon?
No. Standard air quality indices, including India's CPCB AQI, Singapore's PSI and Malaysia's API, are built from PM2.5 and PM10 mass plus a handful of gases.
None carries a sub-index for black carbon or ultrafine particle count, because WHO itself has no numeric guideline value for either (WHO, 2021).
Is it safe to ride a motorcycle when the AQI is moderate?
A moderate city-wide AQI means PM2.5 and PM10 mass were acceptable at the monitoring stations over the last 24 hours.
It says nothing about black carbon or ultrafine particles in dense traffic, where studies measured both running several times higher than the ambient reading (Apte et al., 2011).
Why does my phone's air quality app disagree with what I see on the road?
The app is almost certainly reporting the nearest fixed station's 24-hour PM2.5 or PM10 average. What you see and smell in traffic can include black carbon and ultrafine particles that the station does not measure at all, and that concentrate right where exhaust pipes sit, at handlebar height.
Does the auto-rickshaw study apply to a motorcycle?
The researchers themselves frame it as a fair proxy, not an identical match. Both are open-sided vehicles with no sealed cabin between the rider and surrounding traffic, which is the condition that produced the 3.6 and 8.4 times readings (Apte et al., 2011).
- Treat a moderate AQI as a floor, not a full picture. It only covers PM2.5 and PM10, not black carbon or ultrafine particles.
- Avoid idling directly behind an open-sided vehicle when you can. That position is close to the exact microenvironment the Delhi study measured.
- Cross-check two independent sources, such as IQAir and Sensor.Community, knowing both still skip black carbon and ultrafine counts.
- Filter at your intake, not the index. Easi Breezi's clip-on helmet air purifier filters the air at your face in real time, regardless of what any station reports that day.
Sources
- Apte, J.S., Kirchstetter, T.W., Reich, A.H., Deshpande, S.J., Kaushik, G., Chel, A., Marshall, J.D., Nazaroff, W.W. (2011). "Concentrations of fine, ultrafine, and black carbon particles in auto-rickshaws in New Delhi, India." Atmospheric Environment, 45(26), 4470 to 4480. doi.org/10.1016/j.atmosenv.2011.05.028
- Goel, R., Gani, S., Guttikunda, S.K., Wilson, D., Tiwari, G. (2015). "On-road PM2.5 pollution exposure in multiple transport microenvironments in Delhi." Atmospheric Environment, 123, 129 to 138. doi.org/10.1016/j.atmosenv.2015.10.037
- Upadhya, A.R., Kushwaha, M., Agrawal, P., Gingrich, J.D., Asundi, J., Sreekanth, V., Marshall, J.D., Apte, J.S. (2024). "Multi-season mobile monitoring campaign of on-road air pollution in Bengaluru, India." Science of the Total Environment, 914, 169987. doi.org/10.1016/j.scitotenv.2024.169987
- World Health Organization (2021). "WHO global air quality guidelines." who.int
- Central Pollution Control Board (2026). Daily AQI bulletins, Delhi. cpcb.nic.in
- Centre for Research on Energy and Clean Air (2026). Analysis of October 2026 Delhi AQI bulletins. creaair.org
Written by Ash, mechanical engineer and founder of Easi Breezi, building an active HEPA filtration system for motorcycle helmets (Patent Pending). Based between Hong Kong and Bali, riding daily in the traffic this blog writes about.