Hourly pollutant concentrations and air quality indices on a 1 km European grid plus 3,000 station series, with four-day forecasts and history from 2015.
Data product · Listing updated 11 Nov 2026
Europe
Updated hourly; station data about one hour behind real time
1 km grid over the EU, UK, Norway and Switzerland; about 3,000 stations
Startups, Small and medium businesses, Mid-market, Enterprises, Public sector, Academia and research
Copernicus Atmosphere Monitoring Service, official monitoring stations reported to the European Environment Agency, Alpenglow downscaling
Hourly ingestion and bias correction against stations, machine-learning downscaling to 1 km
Street-level air quality for Europe, combining official monitoring stations with model data so you have a value for every location, every hour.
The base layer combines Copernicus Atmosphere Monitoring Service analyses and forecasts with near-real-time measurements from official monitoring networks reported to the European Environment Agency. A machine-learning model downscales the result to 1 km using land use, road density, elevation and meteorology, and bias-corrects against stations each hour. Grid values are estimates; agreement with held-out stations is strongest for ozone and weakest for NO2 near busy roads, and validation statistics by pollutant are published. Coverage includes the EU, the United Kingdom, Norway and Switzerland. The data describe places, not people.
12 fields as documented by Alpenglow Weather Systems.
| Field | Type | Description | Example |
|---|---|---|---|
| grid_id | string | 1 km grid cell ID | — |
| latitude | number | Cell centre latitude | — |
| longitude | number | Cell centre longitude | — |
| timestamp_utc | datetime | Hour, UTC | — |
| pm25_ugm3 | number | PM2.5 concentration, µg/m³ | — |
| pm10_ugm3 | number | PM10 concentration, µg/m³ | — |
| no2_ugm3 | number | NO2 concentration, µg/m³ | — |
| o3_ugm3 | number | Ozone concentration, µg/m³ | — |
| so2_ugm3 | number | SO2 concentration, µg/m³ | — |
| eaqi | integer | European Air Quality Index band, 1 (good) to 6 (extremely poor) | — |
| dominant_pollutant | string | Pollutant driving the index | — |
| source_type | string | station, model or forecast | — |
Free sample available on request
The provider sends a sample file so you can test fit and quality before you buy.
Prices in EUR, as published by Alpenglow Weather Systems
API, 250,000 calls a month
full grid history and bulk delivery
Converted amounts use DataStoa display rates and are approximate. Alpenglow Weather Systems quotes the final price, taxes and terms.
Request detailed pricingEurope.
Time coverage: 2015 to present.
Provider-reported
Latency: Updated hourly; station data about one hour behind real time
DataStoa has not audited these figures. Ask the provider how they are measured, or test them on a sample.
The licence is agreed directly between you and Alpenglow Weather Systems. This summary is provider-reported; the linked terms prevail.
Not listed: AI / ML model training, AI retrieval and grounding, redistribution allowed, marketing and outreach. Ask Alpenglow Weather Systems if you need any of these.
Weather, climate risk and air quality data for any location on Earth
Alpenglow Weather Systems runs a global weather API, physical climate risk models and a European air-quality service from Zurich. Its forecasts blend national weather service models with machine-learning post-processing, and every historical series is backed by station observations and reanalysis.
Also available on: Snowflake Marketplace, AWS Data Exchange, Databricks Marketplace, Microsoft Azure
What buyers say about Alpenglow Weather Systems. Reviews cover the provider, not only this listing.
from 3 reviews
Adding degree days and the weather-driven demand features to our model noticeably reduced forecast error for heat-sensitive products. Setup took a day since site coordinates were all they needed.
Lukas W., Demand planning manager, Beverage distributor
Each hazard score came with the underlying model resolution, the scenario and the time horizon, which let our actuaries decide where to rely on it. Subsidence scores are less granular than flood. Their scientists joined two calls to explain the method.
Isabel T., Catastrophe modeller, Specialty insurer
We started on the Starter tier and moved to Business when we launched. Response times are consistently fast, the history endpoint uses the same schema as forecasts, and the docs have examples in three languages.
Noah F., CTO, Agritech start-up