A 2025 study measured the actual air 151 to 153 motorcycle taxi drivers breathed in six Thai provinces. Their personal PM2.5 had no statistical link to the nearest official station. Pathum Thani drivers averaged 410.9 micrograms per cubic metre, against Thailand's 2023 national average of 23.3 (Samana et al., 2025; IQAir World Air Quality Report).
Your handlebars sit about a metre from the exhaust pipe of the car stopped in front of you. The nearest official air quality station sits on a rooftop, a kilometre or more away, pulling air that never once passed through a traffic jam.
A 2025 study put real air samplers on 151 to 153 motorcycle taxi drivers across six Thai provinces for a full shift each. The result: what a driver actually breathed had no statistical relationship to what the government's own monitor said that day (Samana et al., 2025).
Key takeaways
- Motorcycle taxi drivers in Pathum Thani personally sampled an average of 410.9 micrograms per cubic metre of PM2.5 during a single shift, the highest of six Thai provinces studied (Samana et al., 2025).
- Thailand's official national annual PM2.5 average was 23.3 micrograms per cubic metre in 2023, already more than 4 times the WHO annual guideline of 5 (IQAir World Air Quality Report 2023; WHO, 2021).
- The study's authors state personal PM10 and PM2.5 readings were not significantly correlated with the nearest official monitoring station or area air sampler, even though the station data correlated well with itself (Samana et al., 2025).
- In Hanoi, motorcyclists measured almost 3 times the black carbon of bus passengers riding the same streets at the same time, a pollutant no public index in the region reports at all (Quang et al., 2020).
The research behind this
Peer reviewed measurements this post is built on. Each card says what was measured, what was found, and what it cannot tell you.
Motorcycle taxi drivers, six Thai provinces
- What they measured
- 441 motorcycle taxi drivers interviewed across Bangkok, Nonthaburi, Pathum Thani, Samut Prakan, Samut Sakhon and Nakhon Pathom. Personal air samplers on the driver's body during one work shift (about 6.5 hours) measured PM10 for 151 drivers and PM2.5 for 153, January to March 2023.
- What they found
- Personal PM2.5 averaged 410.9 in Pathum Thani, 206.2 in Nonthaburi and 132.2 in Bangkok. Province differences were statistically significant. Personal readings were not significantly correlated with the nearest official air monitoring station or area sampler.
- What it can't tell you
- Each driver was sampled on one shift, not across a season. PM10 and PM2.5 came from different drivers on different days, which the authors note produced one physically odd result they do not fully resolve.
Motorcyclists vs bus passengers, black carbon
- What they measured
- Portable aethalometers carried on motorcycles and inside buses and cars in Hanoi traffic, sampling black carbon at the same time, in the same streets.
- What they found
- Mean black carbon was 29.4 micrograms per cubic metre on motorcycles against 10.1 in buses, almost 3 times higher. At peak traffic the gap was 34.7 against 12.1.
- What it can't tell you
- Pilot scale sample. Black carbon is not part of any public air quality index in the region, so this gap never reaches a reported number at all.
One year of mobile monitoring, Bengaluru roads
- What they measured
- About a year of mobile black carbon, ultrafine particle and CO2 monitoring across 150 kilometres of road, with roughly 22 repeat passes per segment, mapped at 30 metre resolution.
- What they found
- Mean black carbon was about 10 micrograms per cubic metre on residential roads, 22 on arterial roads and 56 on highways, a 5 times spread between road types a rider crosses in minutes.
- What it can't tell you
- Indian context, not Thai. Used here to show one fixed station cannot represent the spread between road types, a mechanism that is not specific to any one city.
Neighbourhood sensors vs the EPA monitor
- What they measured
- 12 low cost sensors placed across one Cincinnati neighbourhood, calibrated against the nearest EPA reference monitor, April to June 2022. 3 residents also wore personal monitors for 7 days.
- What they found
- 11 of the 12 neighbourhood sensors read a higher air quality index than the EPA site, with up to 6 days above the unhealthy threshold while the EPA monitor never crossed it once.
- What it can't tell you
- US case, not Southeast Asia. Included because it shows the same gap in a well funded monitoring system, so the mechanism is structural, not a Thailand specific failure.
What the studies actually measured
Start with the Thai study, built for this exact question. Researchers recruited 441 motorcycle taxi drivers across six provinces around Bangkok.
Not all of them carried a personal air sampler. 151 drivers carried one for PM10, 153 for PM2.5, strapped on for a full work shift, about 6.5 hours (Samana et al., 2025).
That is a real person's actual shift, not a lab estimate. The sampler sat with the driver through every queue, every idle stretch behind a diesel truck, every stretch of open road. At the end of each shift, the filter went to a lab.
Alongside that, the researchers pulled the readings from the nearest official air quality monitoring station and a separate area air sampler for the same province and period. Then they checked whether the two data sets moved together.
They did not. The paper states personal PM10 and PM2.5 concentrations were not significantly correlated with either the monitoring station or the area sampler data. The station and area sampler, by contrast, correlated well with each other (Samana et al., 2025).
Average personal PM2.5 for a motorcycle taxi driver's shift in Pathum Thani, the highest of six provinces studied.
Source: Samana et al., PLOS ONE 2025Three other studies show this isn't a Bangkok quirk. In Hanoi, motorcyclists' black carbon ran close to 3 times a bus passenger's on the same road (Quang et al., 2020).
In Bengaluru, black carbon on highways ran 5 times higher than on residential streets a few minutes away (Upadhya et al., 2024). In Cincinnati, a neighbourhood sensor network found unhealthy air on days its one EPA monitor read fine (Kubis et al., 2026).
Treat any single area reading as a rough guide, not a verdict on your own commute. Check your actual route and time of day, not only the city average.
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.
Thailand's official number has looked calm while a driver's shift did not
Thailand's reported national annual PM2.5 average was 18.1 micrograms per cubic metre in 2022, and 23.3 in 2023 (IQAir World Air Quality Report, cited via Thailand's government summary). Bangkok's own city figure for 2023 came in at 21.7 (IQAir World Air Quality Report 2023).
Both numbers already sit well above the World Health Organization's annual PM2.5 guideline of 5 micrograms per cubic metre (WHO, 2021). On paper, that is already a story: Thailand's air is several times over the global health guideline, every year on record.
But neither number describes a single shift. They are annual averages, smoothing a year of traffic jams and clear mornings into one tidy figure.
A Pathum Thani driver's 410.9 reading for one shift runs close to 18 times that 2023 national average. The two numbers were never measuring the same thing.
| Reading | Value (ug/m3 PM2.5) | What it represents |
|---|---|---|
| Thailand national average, 2023 | 23.3 | A year's worth of readings across the whole country, smoothed into one number |
| Bangkok city average, 2023 | 21.7 | A year's worth of readings in one city, smoothed the same way |
| Driver's personal shift, Bangkok | 132.2 | What one motorcycle taxi driver actually breathed over one shift |
| Driver's personal shift, Nonthaburi | 206.2 | What one driver actually breathed over one shift, a different province |
| Driver's personal shift, Pathum Thani | 410.9 | What one driver actually breathed over one shift, the highest of six provinces studied |
Why one official number can't see a motorcycle taxi driver's shift
Two sensors, two different jobs
A rooftop station represents a wide area over a long time. That is the right job for an annual report, not for one person's lungs on one shift.
A rider's face sits close to the vehicle ahead, where exhaust collects. A queue puts that rider in the exact pocket of air a rooftop sensor is sited to avoid.
If the city's air quality app says moderate, a motorcycle commute that day is probably fine.
The app is reporting a wide area average. A driver's own shift reading can run many times higher, and the study behind this post found no statistical link between the two (Samana et al., 2025).
What this means for your ride
If you commute by motorcycle or scooter in heavy traffic, the number on your phone is a starting point, not a verdict on your morning. It tells you roughly how the region is doing.
It does not tell you what is happening three feet from your face in a queue.
This is also why riders who filter the air they actually breathe tend to notice a difference an index reading alone would never explain. Filtering at the point of inhalation protects against the shift you are on, not the year someone else averaged.
Why 2031 probably still looks the same
Thailand's officially reported PM2.5 rose from 18.1 in 2022 to 23.3 in 2023 (IQAir World Air Quality Report). Nothing in the region's monitoring plans adds shift level personal sampling as a standard practice.
The gap Samana et al. documented is structural, not a short term glitch. It comes from comparing a wide area annual average to a single person's single shift.
Upadhya et al.'s Bengaluru data and Kubis et al.'s Cincinnati data show the same structural gap in two very different monitoring systems (Upadhya et al., 2024; Kubis et al., 2026). That is a strong sign the mechanism will not resolve itself in Thailand by 2031 either.
So the honest 5 year projection is not that Bangkok's air gets dramatically worse or better on paper. It is that the gap between the official number and a driver's actual shift stays roughly as wide as it is today.
Who can you trust for a number closer to your ride
No single source replaces a personal sampler strapped to your own commute. A few sources get closer than one city average.
- IQAir's AirVisual map aggregates many stations, some closer to street level, so you can check a specific spot instead of a city wide figure.
- Sensor.Community and PurpleAir are crowdsourced, low cost sensor networks with public, hourly data. Coverage depends on who has installed a sensor nearby.
- Chiang Mai University's DustBoy network is a real-time sensor network built for Thailand's haze season, faster and more local than an annual report.
- Thailand's own Air4Thai network is still worth checking, since it is the number health advisories are built on. Read it alongside an independent source, not instead of one.
None of these fixes the core problem: an area reading, however good, is still not a shift level personal reading.
Frequently asked questions
Is Bangkok's air quality data accurate?
Bangkok's official stations are accurate for a wide area average over time. They were never built to capture one commuter's shift, which is why a 2025 study found no statistical link between the two (Samana et al., 2025).
Why does my air quality app say moderate when traffic feels terrible?
Most apps report an area average from the nearest fixed station, often a rooftop away from the road. Your exposure in a queue, at handlebar height beside an exhaust pipe, can run far higher.
Do motorcycle taxi drivers breathe more pollution than other commuters?
The evidence says yes for Southeast Asian traffic. Motorcycle taxi drivers in Thailand personally sampled PM2.5 up to 410.9 micrograms per cubic metre in one province (Samana et al., 2025). Hanoi motorcyclists measured close to 3 times a bus passenger's black carbon on the same streets (Quang et al., 2020).
What is PM2.5 and why does it matter for riders?
PM2.5 refers to particles 2.5 micrometres wide or smaller, small enough to reach deep into the lungs. The World Health Organization's annual guideline is 5 micrograms per cubic metre, a level Thailand's national average has run well above for years (WHO, 2021).
Can filtering the air inside a helmet help if the official reading is already calm?
Filtering at the point you actually breathe protects against your shift's exposure, not the area average. The two can differ by close to 18 times, so filtering at your face addresses the number that actually reaches your lungs.
What to do this week
- Check two sources, not one. Compare your city's official reading against an independent map like IQAir's AirVisual or a local sensor network before you judge a commute as fine.
- Treat queued traffic as its own zone. The studies above show the biggest gaps appear in stop and go traffic, not on open road, so budget for that specifically.
- Watch for haze season spikes near you. Independent networks like Chiang Mai University's DustBoy update faster than annual reports and catch short term spikes an area average smooths away.
- Don't assume a low area number means a clean shift. This post's whole evidence base says the two can move independently, so plan your own protection regardless of what the app says that morning.
- Filter the air you actually breathe. An Easi Breezi helmet air filtration system filters at the point of inhalation, so it responds to your own ride, not a rooftop average a kilometre away.
Sources
- Samana K, Ito K, Suthienkul O, Ketsakorn A (2025). Analysis of PM-bound polycyclic aromatic hydrocarbons exposure among motorcycle taxi drivers in six central provinces in Thailand in winter. PLOS ONE 20(12): e0336587. doi.org/10.1371/journal.pone.0336587
- Quang TN, Hue NT, Dat MV, Tran LK, Phi TH, Morawska L, Thai PK (2020). Motorcyclists have much higher exposure to black carbon compared to other commuters in traffic of Hanoi, Vietnam. Atmospheric Environment. doi.org/10.1016/j.atmosenv.2020.118029
- Upadhya AR, Kushwaha M, Agrawal P, Gingrich JD, Asundi J, Sreekanth V, Marshall JD, Apte JS (2024). Multi season mobile monitoring campaign of on road air pollution in Bengaluru, India. Science of the Total Environment, PMID 38211861. doi.org/10.1016/j.scitotenv.2024.169987
- Kubis K, Edwards J, Ryan PH (2026). Ambient and personal air monitoring using low cost sensors to identify elevated air pollution in vulnerable communities. Science of the Total Environment, PMID 42061143. doi.org/10.1016/j.scitotenv.2026.181837
- World Health Organization (2021). WHO Global Air Quality Guidelines. who.int, Global Air Quality Guidelines 2021
- IQAir (2024). World Air Quality Report 2023, Thailand country and Bangkok city figures. Reported via The Nation Thailand
- Royal Thai Government (2023). Summary of IQAir World Air Quality Report 2022 national PM2.5 average. thailand.go.th
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.