General selection
What are air quality scales?
Air quality stations measure concentrations of pollutants in each country or region. Countries or regions define air quality indexes and categorise the raw data into a descriptive rating scale. These indexes make it easier to identify the level of pollution and if there’s any associated risk.
Different countries/regions use different scales to report air quality based on local pollution and health considerations. There are dozens of local indexes used across the globe. For example, some states in Australia use a number-based system while others use a category-based system. Canada, US and Japan define separate air quality indexes, as does the European Environment Agency.
As the air pollution worsens, public health risks increase. It especially affects children, the older adult population and other at-risk populations. During times of poor air quality, governmental agencies generally provide health recommendations related to indoor and outdoor activities.
How air quality indexes are calculated
The Air Quality local Indexes are based on measurements of air pollutants.
Some common pollutants that are tracked include:
- Particulate matter, like PM2.5 and PM10
- Ozone (O3)
- Nitrogen dioxide (NO2)
- Sulphur dioxide (SO2)
- Carbon monoxide (CO).
The concentration values of these pollutants, and sometimes other pollutants, are measured over time and used to calculate the Air Quality local Index. Different countries/regions measure different pollutants for the index definition. For example:
- The European AQI reports on the five separate pollutants mentioned above.
- The India AQI reports the pollutants above and on ammonia (NH3).
The Air Quality local Index is influenced by several other factors, such as:
- Wind speed and direction
- Terrain
- Smoke plumes
- Traffic
- Other sources that emit fine particle pollution
Important:
- Pollutant concentrations can be inconsistent over short distances and make air quality readings vary between locations.
- Each monitoring station may not measure every pollutant. This difference can sometimes lead to discrepancies between reported AQI, which are station-specific and reflect only pollutants measured at that station, and actual air quality.
You can learn more about the Air Quality Index in your area with information provided by our partners and third-party sources:
Australia
- Australian Capital Territory Government
- Environment Protection Authority (EPA) Victoria
- Environment Protection Authority (EPA) South Australia
- New South Wales Government
- Northern Territory Environment Protection Authority (EPA)
- Queensland Government
- Government of Western Australia
- Environment Protection Authority (EPA) Tasmania
Brazil
Chile
India
Israel
Mexico
Singapore
South Korea
US
Most common types of outdoor pollutants and their sources
The Air Quality local Indexes are based on measurements of air pollutants. The most commonly measured outdoor pollutants are:
- Particulate matter (PM): Small solid particles and liquid droplets found in the air. PM10 and PM2.5 are particles with a diameter smaller than 10 micrometres and 2.5 micrometres. It’s emitted from motor vehicles, wood heaters and industry. Fires and dust storms can also produce high concentrations of particulate matter.
- Nitrogen dioxide (NO2): A gas and a major component of central city air pollution. It mainly comes from vehicles, industry, power stations and heating.
- Ozone (O3): A gas found in the stratosphere. It protects us from harmful ultraviolet radiation and the troposphere. Ozone is a harmful pollutant produced by a chemical reaction between the sun light, organic gasses and nitrogen oxides released by:
- Cars
- Power plants
- Other sources
- Sulphur dioxide (SO2): A toxic gas with a pungent, irritating odour. It can come from electric industries that burn fossil fuels, petrol refineries, cement manufacturing and volcano emissions.
- Carbon monoxide (CO): A gas from motor vehicles or machinery that burns fossil fuels.
All these pollutants have health implications when they occur in high concentration. Learn more about these in the World Health Organization (WHO) website.
Air quality influences
Although complex, Air quality is influenced by several factors:
- Weather conditions like wind speed and direction, and relative humidity among others.
- Solar irradiance
- Wildfires and other types of fires
- Agriculture dust storms and dust emissions
- Industry and private households emissions
- Traffic emissions
- Other physical and chemical atmospheric processes
Station-based solution specific
How Air Quality Index (AQI) near you is selected
AQI levels are measured at particular air quality stations and do not necessarily reflect the AQI level at your location. For example, if wind is blowing, upstream stations might be more related to the AQI at your location than closer downstream stations. To avoid confusion, we show a map-based view to display the AQI level at given stations around you.
Due to space constraints, several Google products do present a single-station reading. In that case, the AQI value is selected according to the closest station to your location.
What do the smoke plumes mean?
Important: The map might show yesterday’s smoke while today’s smoke is still being analysed. In some cases, the AQI might be good while there’s a smoke plume. This can be caused in cases where the smoke plume doesn’t reach the ground surface and doesn’t affect the measured air quality.
Additional information about smoke in the US is provided based on satellites’ data from NOAA, available in Google Search and Maps.
The data includes medium and high levels of smoke density. Smoke plumes will be shown on the air quality map if data is available.
Air quality data sources
To ensure the quality of Google Air products, we reflect monitoring station’s data directly from these sources:
Australia
- Australian Capital Territory Government
- Environment Protection Authority (EPA) Victoria
- Environment Protection Authority (EPA) South Australia
- New South Wales Government
- Northern Territory Environment Protection Authority (EPA)
- Queensland Government
- Government of Western Australia
Brazil
Chile
India
Israel
Mexico
Singapore
South Korea
United States
Model-based solution specific
Selection of Air Quality Index (AQI) near you
To show the air quality at your location, Google applies its air quality model.
If you're viewing air quality for a city, like 'weather in London', the resulting air quality reading can be for a location far away from you, like in the city centre. This won't accurately reflect the situation around you, even if you're in the same city.
To get AQI for your location:
- Sign in to Google Maps.
- On the location header, select Choose area.
-
To use precise location, change location.
Data source and accuracy of Google’s Air Quality model
We use an air quality model based on a multi-layered approach known as fusion approach. This approach combines data from various input sources and weighs the layers in a sophisticated way. The input layers are:
- Governmental reference monitoring stations
- Commercial sensor networks
- Global and regional dispersion models
- Fire smoke and dust models
- Satellite information
- Traffic data
- Auxiliary information such as land cover
- Meteorology
The Google model provides air quality indices based on the most common pollutant concentrations mentioned above with the addition of NO, NOx and in some cases Non-Methane Hydrocarbons (NMHC). The model is calculated on a 500m × 500m grid.
Pollutant data from governmental or reference monitoring stations is the base layer and most trusted information in the model. To remove any irregular values and ensure high quality of data, the model performs quality assurance on measurements that are collected from monitors all over the world. Where there's a significant delay between the time of measurement and its publication, a nowcasting algorithm calculates pollutant concentrations for the present hour.
Model limitations
While every layer of information used by the Google model has errors associated with it, our approach significantly reduces the total error as the model conducts cross validation between different sources. However, every model has errors that could be:
- A short delay (one to two hours) in the air quality data in some cases.
- Local events, like a barbeque or a burning house, which aren't detected by the model.
- In some cases, the model could show the smoke a few miles from your location.
- Users can experience delayed AQ data in Mexico and Canada during specific times due to fire smoke.
Different AQI value from Google and monitoring station
There can be differences between the air quality index values on governmental monitoring stations and on Google because of these reasons:
- Not all governmental stations measure all kinds of pollutants.
- Governmental monitoring station data frequently has reporting delays so sudden changes in air quality can be missed.
- Governmental monitoring stations only measure what happens at the station’s location.
For example 2, Google’s model takes into account multiple data sources and predicts the air quality at the location of the station in real time for all six pollutants:
- Ground-level ozone (O3)
- PM2.5
- PM10 such as a dust event
- Carbon monoxide (CO)
- Sulphur dioxide (SO2)
- Nitrogen dioxide (NO2) at a 500m grid resolution
Example 1: During a dust event (PM10), a monitoring station will show good air quality since it only measures Ground-level ozone (O3). However, the Google model shows poor air quality since PM10 is included in it.
Example 2: A monitoring station measures all pollutants, Google and the station show the same dominant pollutant like Ozone, however, Google is showing an air quality index of 200 while the station shows 150. This can be because Ozone changes throughout the day, so the measurement that was taken two hours ago couldn’t be identical to Google’s real-time prediction.
The air quality that Google shows doesn't match with what I see on my (or nearest) commercial sensor. Why is that?
The number of pollutants reported
Most commercially available sensors report on PM2.5 and PM10 only, whereas Google provides air quality reporting for multiple pollutants, including Ground-level ozone (O3), PM2.5, PM10, Carbon monoxide (CO), Sulphur dioxide (SO2) and Nitrogen dioxide (NO2). In most commercial sensors, there's a decreasing efficiency for larger particles, meaning the PM10 measurement reliability could be very low due to the optical method limitation used in these sensors.
Environmental factors
External environmental factors, such as high relative humidity and temperature, could affect the reading of commercial optical sensors.
The location of the sensor can also have a major effect on the readings as they can be impacted by local pollution, which doesn't represent a wider area.
Our model incorporates readings from commercial sensors and identifies and removes invalid measurements.
The conversion process involved
Commercial sensors often adopt a 'count-based' method of PM2.5 measurement, while the standard format used for reporting air quality information is 'mass-based'.
The government monitoring stations and Google’s model report in terms of 'Mass per volume' by default. This conversion requires the density of the particle such as smoke or dust. This conversion could lead to large differences when compared to government monitors and Google’s model.
The averaging time
Different commercial sensor network providers display data with different averaging times. Google is calculating pollutant concentrations and AQIs on an hourly basis. Each AQI has their own averaging times based on the country, which is usually at least hourly, often multiple hours. For example, in case of a sudden increase in smoke pollution, you could see poor air quality being indicated on a commercial sensor provider website based on a 10-minute average, while Google would show the hourly average which in this case would be lower because of the preceding hours of lower pollution, and the official AQI value which reflects health implications is based on even longer averages.
Tip: The commercial sensor network isn't included in our model and could show different results.
Google’s Air Quality data differs to other providers
Different air quality indexes
Different countries and regions use air quality indexes based on different scales, for different purposes. If you’re looking to compare providers, it’s important to ensure that you're comparing reports that speak the same 'air quality language'.
For example, some providers use the US AQI hourly, while others use the US AQI as a set such as the averaging PM2.5 over 24 hours or the AirNow index.
Different reporting and measurement methods
There are differences in how different providers measure air quality, which affects the final results.
In some cases, air quality data is based on measurements. Some providers use commercial sensors, and others use models.
Google combines information from monitoring stations around the world with:
- Commercial sensor information
- Satellite data
- Weather patterns
- Traffic conditions reporting
- Wildfire
- Land cover information
When reporting air quality data, different providers can have different time aggregations so it’s possible to see time delays in reported pollutant or index values across different providers at the same location.
A pollutant is present but Google shows good air quality
Sometimes, the air quality information you see on Google can seem inconsistent with what you see or smell around you. There are usually a number of possible explanations:
- Noses are particularly sensitive to smell. It’s possible that you can smell a particular type of pollution, like gasses from smoke in very low concentrations, even if the air quality is considered safe enough from a health perspective. Air pollution created by Volatile Organic Compounds (VOCs) can produce noticeable smells, but aren't measured by government monitoring stations or included in our air quality reports.
- Smoke that affects visibility can often be seen at high altitudes even if it's not detectable at the ground level.
Learn more about Google model limitation.
Google shows bad air quality but everything is clear
While some types of pollution are visible such as dust storms or wildfire smoke, a lot of air pollution isn't visible to the naked eye. This is one of the fundamental ways air pollution differs from the weather, where often we can simply glance outside and see if it's sunny, rainy or windy.
For example, high levels of ozone or otherwise known as the 'beautiful day' effect, which requires the presence of sunlight to form, can be created at high altitudes like on a mountain top or on an otherwise nice and sunny day.
Another reason could be the model limitation as described in Air Quality Help.
Google reports differently to United States’ AirNow
Two of the most popular official sources for air quality information in the United States are the AirNow main website and the AirNow Fire and Smoke Map.
There are some key differences between AirNow and Google air quality reports:
- Number of pollutants monitored: Google is aligned with the US EPA with US EPA and covers more pollutants than AirNow shows in the website.
- Google uses more data sources.
- Google reports hyper-local information versus area-wide worst report by AirNow.
- Google and Air Now have differences in the Air Quality Index (AQI).
| AirNow | AirNow Fire & Smoke Map | ||
| Data sources and method |
Monitoring stations and interpolation model based on the worst reading in a region |
|
Various input data, models and different temporal and spatial prediction algorithms:
|
| Pollutants monitored |
|
Particulate matter (PM2.5) |
|
The US has two official ways of calculating the Air Quality Index (AQI):
- The US EPA’s AQI calculates the exposure over longer periods of time and incorporates six pollutants. See table below.
- NowCast, which is used to approximate the complete daily AQI during any given hour. It provides current conditions to enable people the power to take action to reduce outdoor activities and exposure when necessary and protect their health. The NowCast allows current conditions maps to align more closely with what people are actually seeing or experiencing. It's used to provide more timely information for Ozone, PM2.5 and PM10.
When reporting the AQI in the US, Google combines the two indexes.
Comparing Official USA AQIs and Google
Here's a more detailed breakdown of the different US AQI approaches, so you can compare them side by side:
| US AQI | AirNow NowCast | Google Hybrid US AQI | |
| Number of pollutants |
6 pollutants
|
3 pollutants
|
6 pollutants
|
|
Time periods considered |
The average time period covered varies by pollutant: e.g. ozone is calculated based on 8-hour average exposure levels as well as 1-hour exposure ranges. PM2.5 is reported based on 24-hour exposure averages. |
The AirNow NowCast averaging method gives more weight to recent hours to better reflect sudden changes in air pollution levels due to smoke from fires or dust storm events. Shorter average time periods compared to US AQI e.g. taking into account the last period of 12 hours. |
Google adopts the AirNow NowCast averaging method for Particulate Matter (PM10 and PM2.5), giving more weight to recent hours and sudden changes. But we also report on more pollutants compared to NowCast: Google takes the averages of other pollutants: O3, NO2, SO2, CO, runs calculations, then converts this into the US AQI format. |
Air quality data sources
Here are the sources where we gather information for the Google Air Quality model:
Global data sources
- Low-cost sensors data from PurpleAir.
- Modified information from Copernicus Atmosphere Monitoring Service.
- Modified information from Copernicus Global Land Cove.
- Public sector information licensed under the Open Government Licence v3.0. from Met Office.
- European EEA information under CC-BY-2.5 DK licence.
Belgium
- Modified IRCEL - CELINE information. Licence.
Canada
- Information from the Manitoba government, licensed under the OpenMB Information and Data Use Licence (Manitoba.ca/OpenMB).
- Information licensed under the Open Government Licence – Ontario, version 1.0.
Denmark
- DCE - National Center for Miljø og Energi. The data is raw data that isn't quality controlled.
Finland
France
- Pays de la Loire: Source of the data: Air Pays de la Loire.
- Geo D'Air.
Germany
- Contains modified German Environment Agency information. Changes that were made.
Guernsey
- © Crown 2023 copyright Defra via uk-air.defra.gov.uk
Italy
- Regione del Veneto. Licence.
Ireland
- Environmental Protection Agency (EPA): https://www.epa.ie/ https://airquality.ie/.
- Licence.
Japan
- Modified Soramame information.
Mexico
- The Air Quality information published by the Environment Secretariat of the Government of Mexico City is prepared from the information obtained from the Atmospheric Monitoring Network and its monitoring stations in the Metropolitan Zone of the Valley of Mexico, which are operated and managed by the Air Quality Monitoring Directorate of the General Directorate of Air Quality (SEDEMA). This information is available to the public, and it is subject to quality processes that could modify it. The dissemination or use of this information by third parties is under the responsibility of the person who publishes or uses it.
- SINAICA, https://sinaica.inecc.gob.mx/. Changes that were made.
Spain
- MeteoGalicia and the Ministry of the Environment, Territory and Housing of the Xunta de Galicia.
- Madrid Comunidad.
- Ministry of the Environment information under CC BY 4.0 licence.
Sweden
- Contains modified SMHI information.
United Kingdom
- © Crown 2023 copyright Defra via uk-air.defra.gov.uk.
- Contains LondonAir information. Licence.
- Northern Ireland Air.
United States
- Texas TCEQ.
- New York State, Department of Environmental Conservation: The data displayed here includes data obtained from http://www.nyaqinow.net that's preliminary and subject to change.