PM2.5 now, the median of the measurements
median of 4 sources · as of 29 October, 12:59
The sources disagree: the measurements show 76 µg/m³, the Copernicus model 17. The model works on a grid of about 40 km and usually misses local smog from heating and traffic, so it stays out of the summary.
Each source with its own link. Times are Tashkent time.
Uzhydromet station
An automatic Uzhydromet station. WAQI publishes its AQI on the 2012 scale; we convert the index back to µg/m³ and show the index on the 2024 scale.
73 µg/m³ PM2.5
AQI 163 · unhealthy
PM10: 118 µg/m³
converted from the index
Data: World Meteorological Organization, Uzhydromet, World Air Quality Index Project
Uzhydromet station
An automatic Uzhydromet station. WAQI publishes its AQI on the 2012 scale; we convert the index back to µg/m³ and show the index on the 2024 scale.
79 µg/m³ PM2.5
AQI 167 · unhealthy
PM10: 132 µg/m³
converted from the index
Data: World Meteorological Organization, Uzhydromet, World Air Quality Index Project
US Embassy
The US Embassy's PM2.5 monitor, via WAQI. At times there is no PM2.5 and the index comes from PM10.
no PM2.5 data
AQI 82 · moderate
PM10: 118 µg/m³
Data: Tashkent Air Quality Monitor - US Embassy, Uzhydromet, World Meteorological Organization, World Air Quality Index Project
community sensor
A resident's laser particle sensor (Sensor.Community). Cheap: above 70 % humidity it overstates.
50 µg/m³ PM2.5
AQI 137 · unhealthy for sensitive groups
PM10: 96 µg/m³
Data: Sensor.Community
aggregator
IQAir's average index for Tashkent. The free plan gives no concentrations, so we convert the index.
80 µg/m³ PM2.5
AQI 168 · unhealthy
converted from the index
Data: IQAir
model
The Copernicus CAMS model for the city centre, via Open-Meteo. A calculation, not a measurement: it misses local smog.
17 µg/m³ PM2.5
AQI 65 · moderate
PM10: 23 µg/m³
Data: Copernicus Atmosphere Monitoring Service (CAMS), Open-Meteo.com (CC BY 4.0)
PM2.5 by source, hourly means, µg/m³. The horizontal line is the WHO 24-hour guideline, 15 µg/m³. Dashed: the model, not a measurement.
| Time | Yunusabad | Chilanzar | Community sensors | IQAir | CAMS model |
|---|---|---|---|---|---|
| 29 October, 13:00 | — | — | — | — | 17 |
| 29 October, 12:00 | 73 | 79 | 53 | 80 | 19 |
| 29 October, 11:00 | — | — | — | — | 24 |
| 29 October, 10:00 | — | — | — | — | 31 |
| 29 October, 09:00 | — | — | — | — | 40 |
| 29 October, 08:00 | — | — | — | — | 43 |
| 29 October, 07:00 | — | — | — | — | 40 |
| 29 October, 06:00 | — | — | — | — | 35 |
| 29 October, 05:00 | — | — | — | — | 33 |
| 29 October, 04:00 | — | — | — | — | 32 |
| 29 October, 03:00 | — | — | — | — | 33 |
| 29 October, 02:00 | — | — | — | — | 39 |
| 29 October, 01:00 | — | — | — | — | 45 |
| 29 October, 00:00 | — | — | — | — | 49 |
| 28 October, 23:00 | — | — | — | — | 54 |
| 28 October, 22:00 | — | — | — | — | 55 |
| 28 October, 21:00 | — | — | — | — | 53 |
| 28 October, 20:00 | — | — | — | — | 52 |
| 28 October, 19:00 | — | — | — | — | 50 |
| 28 October, 18:00 | — | — | — | — | 45 |
| 28 October, 17:00 | — | — | — | — | 36 |
| 28 October, 16:00 | — | — | — | — | 38 |
| 28 October, 15:00 | — | — | — | — | 37 |
| 28 October, 14:00 | — | — | — | — | 38 |
| 28 October, 13:00 | — | — | — | — | 41 |
Hourly PM2.5 from the Copernicus CAMS model. In Tashkent the model usually runs low, so read it for which hours are cleaner rather than for the numbers themselves.
Best time for a walk: tomorrow 14:00–17:00, when the model expects about 15 µg/m³ (moderate).
| Day | Model, PM2.5 (µg/m³) | WAQI forecast, index |
|---|---|---|
| Thursday, 29 October | 14–33 | 115 (89–138) unhealthy for sensitive groups |
| Friday, 30 October | 14–45 | 97 (69–138) moderate |
| Saturday, 31 October | 17–49 | 84 (69–89) moderate |
| Sunday, 1 November | 17–48 | 72 (68–89) moderate |
IQAir, WAQI and most apps show the US AQI, the EPA scale as revised in 2024. Uzhydromet has its own PM2.5 scale, and it is much milder: a level the AQI already calls “unhealthy” (from 55.5 µg/m³) stays green on Uzhydromet's scale up to 60 µg/m³ and yellow up to 120. That is why the same morning looks “hazardous” in one app and “satisfactory” on the agency's map.
| Category | AQI | PM2.5, µg/m³ |
|---|---|---|
| Good | 0–50 | 0–9 |
| Moderate | 51–100 | 9.1–35.4 |
| Unhealthy for sensitive groups | 101–150 | 35.5–55.4 |
| Unhealthy ← now | 151–200 | 55.5–125.4 |
| Very unhealthy | 201–300 | 125.5–225.4 |
| Hazardous | 301–500 | 225.5+ |
| Colour PM2.5, µg/m³ | Uzhydromet's advice |
|---|---|
| Green 0–60 | Good air, no limits. |
| Yellow 60–120 ← now | Satisfactory. People with chronic respiratory disease or sensitive to pollution should do less outdoors if they get a sore throat, a cough or shortness of breath. Everyone else as usual. |
| Orange 120–180 | Moderate pollution. Children, pregnant and nursing women, older people and those with heart or lung disease should limit outdoor exertion where they can and carry the medicines their doctor prescribed. Others should ease off if the nose or throat gets irritated; a mask makes sense. |
| Red 180–300 | High pollution, a danger to healthy people too. Sensitive groups should sharply cut outdoor exertion and stay indoors where possible. Others may go out, but in a protective mask, and ease off if the nose, throat or eyes feel irritated. |
| Purple 300+ | Unsatisfactory. Everyone should avoid long spells outdoors; vulnerable people should stay home and skip sport. Remote work and study are advised, and outdoor events should be cancelled or postponed. |
Uzhydromet's advice, summarised from monitoring.meteo.uz.
The category now is “Unhealthy” on the US AQI. Advice after the EPA and the WHO:
Surgical and cloth masks barely stop fine PM2.5 particles. An FFP2, N95 or KN95 respirator protects if it seals to the face; with a beard or a loose fit the benefit drops sharply. Respirators rarely fit small children, so for them less time outside matters more. People with heart or lung disease should ask a doctor before wearing a respirator for long.
A purifier with a HEPA filter noticeably cuts PM2.5 in a room with the windows shut. Look at its CADR, the clean air it delivers per hour: a 20 m² room needs about 250 m³/h or more. Change the filter on the maker's schedule. Skip ionisers and devices that give off ozone: ozone itself irritates the lungs. Air the room during the hours the forecast shows as cleaner.
Based on AirNow (EPA) and WHO guidance.
The Tezkun figure is the median PM2.5 of the fresh measurements: Uzhydromet's stations and the US Embassy monitor (via WAQI), Sensor.Community's community sensors (together they count as one value) and IQAir's city average. A median resists outliers: one sensor that overstates cannot move it. Values older than 3 hours and community readings above 70 % humidity are left out.
WAQI and IQAir publish the US AQI, not concentrations. We convert the index back to µg/m³ with the table it was computed on: the EPA's 2012 table for WAQI (as its documentation states) and the 2024 revision for IQAir; the result is as precise as the index's rounding, about 1 µg/m³. The AQI of the summary and of the cards comes from the concentration through the 2024 table, from current values rather than a 24-hour mean, so it can differ from the index on aqicn.org.
The Copernicus CAMS model is a calculation, not a measurement: it gives the background for a square of about 40 km and a forecast several days ahead, but it does not see local smog from stoves, boiler houses and traffic. That is why the model is shown separately and never enters the median.
Updates: the collector polls the sources every 15 minutes; the stations and IQAir update hourly, the community sensors every few minutes. All times are Tashkent time.
WAQI's data may not be resold or redistributed as an archive; when quoting, credit the original source.
The city network air.tashkent.uz is not connected yet: it has no open API.
Tezkun has kept the history of measurements since 29 October 2026.