Enabling Access to Air Quality Data: Global Air Quality Monitoring and Forecasting Products

This is the fifth article in the Blog Article Series on Air Pollution.

June 4, 2025

Colorful globe surrounded by charts and graphs, set in a vibrant outdoor landscape. — This image was generated with the assistance of DALL·E 3

Authors

Khairunnisa Yahya

Air Pollution Expert, UNDP Global Centre Singapore

Rinkle Dang

Climate and Sustainability Intern, UNDP Global Centre Singapore

Previous articles have examined how Low-Cost Sensor (LCS) data, satellite products, and air dispersion models enhance air quality data collection, with a case study from Singapore. While these advanced applications require expertise in data processing and analysis, alternatively, there are also readily available air quality datasets and forecasting products online that have been processed and compiled. These datasets can help bridge air quality data gaps in regions with few or no sensor networks, particularly in resource-limited countries. In addition, they can support a diverse range of air quality data users across public, private, and nonprofit sectors, including researchers, scientists, educators, students, policymakers, journalists, and the public. They enable these stakeholders to study, develop, implement, and communicate effective air quality measures.

Most importantly, the availability of high-quality, publicly accessible data fosters discussions on air quality monitoring and strengthens advocacy for pollution control policies. This is especially crucial, as approximately 1.4 billion people live in countries where no publicly available air quality data exists.

The air pollution datasets can generally be categorised into three main types: historical, real-time, and forecasting.

Historical Air Pollution Data

Historical air pollution data is typically compiled over time to assess long-term pollution trends, such as annual averages. This allows cities to evaluate their air quality against established guidelines or standards to determine whether pollution levels are considered safe. Such data is also useful for tracking air pollution events such as wildfires, and monitoring air quality changes over time to assess the effectiveness of pollution control measures. Additionally, historical air pollutant data plays a crucial role in epidemiological research, helping scientists quantify exposure to harmful pollutants and analyse associated health risks. By integrating this data with local health statistics, researchers can estimate the number of deaths (mortality rate) and disabilities or illnesses (morbidity rates) that are attributable to air pollution.

Annual average air pollution levels from previous years are compiled and shared by organisations like the World Health Organisation (WHO). This data is used to support various global databases, including the State of Global Air, and Sustainable Development Goal Indicator (SDG) 11.6.2, Air Quality in Cities.

There are several free historical air pollution datasets and some of them are listed below:

Source: Websites of respective datasets compiled by authors. See footnotes at the end of the blog.

World map showing concentrations of fine particulate matter (PM2.5) with varying shades of green. — Figure 1: SDG Indicator 11.6.2: Annual mean levels of PM2.5 and PM10 in cities (population-weighted)

WHO Ambient Air quality database

Real-Time Air Quality Data

Real-time air quality data is collected and reported hourly, either through national reference monitoring stations, typically managed by governments, or more recently through low-cost air quality sensors. This data helps individuals understand local air pollution levels and make informed decisions about daily activities. For example, when pollution levels are high, people can adjust their outdoor plans to minimise exposure.

City-wide air quality awareness also empowers communities to advocate for localised pollution control measures. Additionally, timely and accurate data allows authorities to issue early warnings during periods of poor air quality, helping protect public health. Real-time data sources, which often provide downloadable datasets and visualisations, further support air quality forecasters in refining their forecasts.

Some of the sources of real-time air quality data are listed below:

Table collating the sources of real-time air quality data — Source: Websites of respective datasets compiled by authors. See footnotes at the end of the blog.

Color-coded map of air quality across the region — IQAir Live Animated Air Quality Map (AQI, PM2.5...)

Air Quality Forecasting Data

Air Quality Forecast Data, derived from air dispersion models, provides air quality predictions up to a week in advance. These models provide insights into how air pollutants like PM2.5 are transported and dispersed in the atmosphere by combining emissions data with weather forecasts and real-time air quality measurements. Predicting air quality is a complex task, as it relies on diverse data sources that vary across locations and time, making data integration challenging.

Despite these challenges, accurate forecasts are essential for tracking pollution trends and providing early warnings for incidents, such as wildfire smoke dispersion. A global air quality model forecasting system can deliver reliable air quality information to everyone, enabling proactive measures to improve air quality.

Various global and regional forecasting models exist, and some examples are provided below. A more detailed global and regional list of forecasting models can be accessed here.

Global Air Quality Forecasting Data

GEOS-CF by NASA : NASA’s GEOS Composition Forecast (GEOS-CF) model provides global air-quality forecasts for pollutants PM2.5, O3, NO2, SO2 in near-real time for up to five days in advance, with a spatial resolution of approximately 25km. The model output can also be used to create health air-quality index for a country/city. Additionally, the data can be accessed in different formats, including maps, or visualisations which can be viewed on FLUID and are also available for download.
Copernicus Atmosphere Monitoring Service (CAMS): The Copernicus Atmosphere Monitoring Service provides five-day downloadable forecast charts for European air quality and globally for pollutants such as PM2.5, PM10, O3, NO2, SO2, greenhouse gases, aerosols, fires, and other reactive gases, based on satellite based observations and non-satellite data and modelling.

Regional Air Quality Forecasting Data

South-East Asia: Meteorological Service Singapore utilises a Multi-Model Ensemble to forecast data for the next 48-hour period at 3-hourly intervals. These forecasts combine individual models for estimating smoke aerosol optical depth (AOD) and concentrations of surface PM 2.5 and PM 10.
Global, including Europe and South-East Asia: The System for Integrated Modelling of Atmospheric Composition (SILAM by Finnish Meteorological Institute) is an open-source system that provides 4-day forecasts for major air pollutants, including SO2, NO, NO2, O3, PM2.5, and PM10.

Conclusion:

Given the risk that air pollution poses to the human welfare and the environment, access to reliable and publicly available air quality data is crucial for researching, designing, advocating, and communicating interventions for clean air. There are many readily available sources of historical, real-time, and forecasting air quality data, which can be utilised as needed. While some databases may have coverage gaps, particularly in developing regions like South Asia and Africa, ongoing efforts continue to expand monitoring infrastructure and improve accessibility. Additionally, various techniques used in these databases enhance the understanding of air pollution, despite occasional challenges such as accessibility or language barriers. In cases where low-cost sensors are used, they can provide valuable air quality insights especially for communities where air quality monitoring remains limited, by enhancing their accuracy and reliability and validating them against reference sensors. As advancements progress, these data sources serve as essential tools in the global pursuit of cleaner air and a healthier future for all.

Sources cited in the tables:

¹ https://www.who.int/data/gho/data/themes/air-pollution/who-air-quality-database

² https://pmc.ncbi.nlm.nih.gov/articles/PMC10680116/

³ https://www.stateofglobalair.org/resources/archived/state-global-air-report-2020

⁴ http://www.healthdata.org/gbd

⁵ https://www.stateofglobalair.org/data/estimate-exposure

⁶ https://www.iqair.com/us/world-most-polluted-cities

⁷ https://www.iqair.com/us/world-air-quality-report

⁸ https://www.iqair.com/us/world-air-quality-report

⁹ https://www.iqair.com/air-quality-map

¹⁰ https://www.airnow.gov/aqi/aqi-basics/

¹¹ https://www.iqair.com/newsroom/what-determines-the-major-cities-ranking