Live Air Quality
Up-to-date health advice and comparison with other major cities below. Updated hourly.
This station is currently experiencing data issues
This pollution monitoring station has not been reporting data for one or more pollutants required to calculate the current AQHI/AQI.
The data and charts below may be missing some elements.
Embee® filters for IQAir®
Health advice from the US AQI at 3:00 AM HKT
- Air quality is satisfactory, and air pollution poses little or no risk.
- Great day to be active outside!
Live PM2.5 Levels
PM2.5 are tiny particles less than 2.5 millionths of a metre in size. They are so small that when you breath in, they can pass directly through your lungs into your blood, and then around your body to organs including the brain.
Past 24-hours PM2.5 (μg/m³)
There is strong evidence linking PM2.5 pollution levels to heart disease, cancer, strokes, and COPD.
Southern PM2.5 vs. Major Cities(μg/m³)
WHO PM2.5 limits are based on a 24-hour average level, which is what is shown above. The current HK (past 1-hour) level is also shown with a black line.
Individual Pollutant Levels
As of 3:00 AM HKT(μg/m³)
View a different location
- London: ERG, Imperial College London under the terms of the Open Government License.
- Singapore: National Environment Agency via Data.gov.sg under the terms of the Singapore Open Data License
- Sydney: NSW Planning Industry & Environment under terms of the CC BY 4.0 license
- Hong Kong: Environmental Protection Department via data.gov.hk
- New York: U.S. EPA public domain data via the AirNow API
- 1 To reflect current conditions, the AQI value is calculated using latest 1-hour readings for PM2.5 and PM10 particulates. This differs from the typical US AQ calculation which uses a 24-hour average. Read more here.
- Heath advice and US AQI calculation is per the Technical Assistance Document for the Reporting of Daily Air Quality – the Air Quality Index (AQI) issued by the US EPA in September 2018.
- Hong Kong AQHI calculation is per methodology published by the EPD, with missing data interpolated using last known good value.