Frequently asked questions (FAQ)

  • We cover the main pollutants that matter to your health: PM2.5, PM10, NO2 and O3.

    • Particulate matter (PM2.5 and PM10) are small, solid particles that come from human activity, such as traffic and combustion, and from natural phenomena such as volcanic eruptions. They penetrate your airways, lungs and, in the case of the smallest particles, even blood vessels. This can be responsible for cardiovascular and respiratory diseases, as well as lung cancers.
    • Nitrogen dioxide (NO2) is a suffocating and irritating red-brown colored gas that comes mainly from combustion, 50% of which is due to road traffic. It causes bronchitis, asthma and other respiratory diseases.
    • Ozone (O3) is a gas formed by a chemical reaction between other pollutants. Its concentration is high when there’s a combination of strong sunlight and high traffic. It is responsible for respiratory and heart diseases, asthma and eye irritation.

  • When you make a request to the API for a given location (latitude and longitude), it returns hour-by-hour estimates for each pollutant, for the current time and over the next 24 hours. Pollutant levels are expressed in absolute concentrations (µg/m3), as well as in various air quality indexes: the Chinese AQI, the US AQI, and our own Plume Index.

  • Our Plume Index is designed to mirror international health recommendations related to air pollution exposure. It takes into account the pollutants that have a proven impact on our health and thresholds are determined in line with WHO guidelines. Each of the 7 categories of the Plume Index, from Fresh Air to Airpocalypse, indicates something specific about the length of time one can be exposed to such pollutant rates without an adverse impact on health. At “High Pollution” levels, for example, an exposure longer than 24 hours is harmful for human health.

  • The Plume API covers the entire world. You can get air quality data for any given location – not a single point on Earth is left out.

  • In most parts of the world, our resolution ranges from 10km2 to 40km2. Where street-level city maps are available, the resolution can be as high as 20m2.

  • You may access historical data via on-demand request — they are not directly provided through our API. If you need historical reports, please get in touch with us.

  • We source air pollution data from more than 12,000 environmental monitoring stations, run by national and local governments and agencies, in 60 countries around the world. Data from these stations is typically reported by local authorities with a delay of up to 4 hours. Therefore, unlike some of our competitors, we don’t simply pass these readings on to you and pretend it’s “real-time data” (which would be inaccurate). Instead, we use our own forecasting algorithms to predict truly real-time pollution levels. In areas located too far from ground-level monitoring stations, we use satellite data, weather information and models backed up by state-of-the-art academic research to provide background air quality estimates. This means we collect, analyze and quality-control millions of data points every single hour.

  • Forecasting air quality is very complex and requires developing a specific set of tools and skills – that’s exactly why we have a whole team dedicated to it at Plume Labs. Our forecasts are built by combining the latest advances in machine learning, atmospheric sciences and big data. We take into account a wide array of local features that have an influence on air pollution levels such as their natural variations (seasonal and intra-day cycles), weather patterns, weather forecasts, geographical specifics, human activities – and of course, we also factor in the latest data from ground stations and our historical datasets. Our AI algorithms run 24/7 and our forecasts at least once an hour so that you always get the latest and most accurate data.

  • Our air quality index groups pollution levels under 7 categories: fresh air, moderate, high, very high, excessive, extreme and airpocalypse. Across the board, 84% of our air pollution forecasts correctly predict pollution levels over a 24-hour period — and our accuracy is particularly high for certain pollutants (97% for nitrogen dioxide, 96% for ozone). Considering that we rely on weather forecasts that are themselves potentially subject to inaccuracy, you start to get a sense of just how accomplished our data sciences team really is :) We are constantly striving to improve our predictions and push the limits of what our environmental artificial intelligence can achieve.

  • We rely on air quality data feeds from ground stations (operated by public agents) to forecast pollution. These stations are extremely sophisticated but require maintenance, and are subject to unexpected shutdowns. Also, they sometimes report incorrect or inconsistent data. To make sure that our datasets are 100% robust, our algorithms constantly screen ground station data and identify outliers (based on a statistical approach and cross-checks with nearby stations). We compensate for station downtime in real-time with our own predictions, to avoid any API service interruption. And because our forecast accuracy is so important to you, we closely monitor our performance and leverage everything we’ve learned over the past 3 years to improve it.

  • We have redundant back-ends and fallback strategies to ensure our API stays online whatever happens. We can provide you with an SLA under our Enterprise Plan. Please get in touch if needed.

  • Our infrastructure can support millions of requests per month. It’s currently being used by large-scale client applications, as well as our air quality forecast app, the Plume Air Report.

  • We provide a unique environmental API that empowers you to develop commercial and noncommercial applications around air quality information. By agreeing to our Terms of Use, you commit yourself to not relicensing, reselling or transferring our data to any third party.

  • We offer different packages according to your needs: with varying call volume, add-on features, levels of support, etc. You can check our available offers here, and contact us for more details.

  • Once we’ve reached an agreement and you’ve agreed to our Terms of Use, we will send you an invoice so you’ll be able to pay via bank transfer. We will soon make electronic payment available through our website — we’ll keep you updated on that.

  • We cover the main pollutants that matter to your health: PM2.5, PM10, NO2 and O3.

    • Particulate matter (PM2.5 and PM10) are small, solid particles that come from human activity, such as traffic and combustion, and from natural phenomena such as volcanic eruptions. They penetrate your airways, lungs and, in the case of the smallest particles, even blood vessels. This can be responsible for cardiovascular and respiratory diseases, as well as lung cancers.
    • Nitrogen dioxide (NO2) is a suffocating and irritating red-brown colored gas that comes mainly from combustion, 50% of which is due to road traffic. It causes bronchitis, asthma and other respiratory diseases.
    • Ozone (O3) is a gas formed by a chemical reaction between other pollutants. Its concentration is high when there’s a combination of strong sunlight and high traffic. It is responsible for respiratory and heart diseases, asthma and eye irritation.

  • Our Plume Index is designed to mirror international health recommendations related to air pollution exposure. It takes into account the pollutants that have a proven impact on our health and thresholds are determined in line with WHO guidelines. Each of the 7 categories of the Plume Index, from Fresh Air to Airpocalypse, indicates something specific about the length of time one can be exposed to such pollutant rates without an adverse impact on health. At “High Pollution” levels, for example, an exposure longer than 24 hours is harmful for human health.

  • We source air pollution data from more than 12,000 environmental monitoring stations, run by national and local governments and agencies, in 60 countries around the world. Data from these stations is typically reported by local authorities with a delay of up to 4 hours. Therefore, unlike some of our competitors, we don’t simply pass these readings on to you and pretend it’s “real-time data” (which would be inaccurate). Instead, we use our own forecasting algorithms to predict truly real-time pollution levels. In areas located too far from ground-level monitoring stations, we use satellite data, weather information and models backed up by state-of-the-art academic research to provide background air quality estimates. This means we collect, analyze and quality-control millions of data points every single hour.

  • Forecasting air quality is very complex and requires developing a specific set of tools and skills – that’s exactly why we have a whole team dedicated to it at Plume Labs. Our forecasts are built by combining the latest advances in machine learning, atmospheric sciences and big data. We take into account a wide array of local features that have an influence on air pollution levels such as their natural variations (seasonal and intra-day cycles), weather patterns, weather forecasts, geographical specifics, human activities – and of course, we also factor in the latest data from ground stations and our historical datasets. Our AI algorithms run 24/7 and our forecasts at least once an hour so that you always get the latest and most accurate data.

Plume Labs
Plume Labs
11 Bis rue Bachaumont
75002 Paris
France