Global Air Quality Forecast Scraper - Hourly PM2.5, Ozone & Pollen (No API Key)

Scrape hourly air quality forecasts and pollen counts for any coordinate on Earth - up to 7 days ahead, with no API key, no login and no proxy required. Feed the Actor a list of latitude/longitude pairs and it returns one clean, flat row per location per hour: PM2.5, PM10, ozone, nitrogen dioxide, sulphur dioxide, carbon monoxide, dust, UV index, European AQI, US AQI, an AQI category label, and alder, birch, grass and ragweed pollen counts.

Data comes from the Open-Meteo air quality API, which blends the CAMS European and CAMS global atmospheric composition models. It is free for non-commercial use and requires no credentials.

How is this different from the Air Quality Scraper?

This Actor is the forecast + pollen counterpart to Air Quality Scraper. They are complementary, not duplicates:

	Air Quality Scraper	Air Quality Forecast Scraper (this one)
Time dimension	Current conditions - one snapshot row per location, right now	Hourly forecast - one row per hour, 1-7 days ahead
Rows per location	1	24 x forecast days (up to 168)
Pollen	Not included	Alder, birch, grass, ragweed
Extra fields	Live measurement values	European AQI + US AQI + AQI category, UV index, dust
Typical use	"What is the air like right now?"	"What will the air and pollen be like tomorrow at 3pm?"

Use the current-conditions Actor for live dashboards and alerting on now. Use this one for planning, allergy calendars, scheduling and any model that needs a forward-looking hourly time series.

Features

• No API key, no account, no proxy - runs out of the box.
• Worldwide coverage - any latitude/longitude, 11 km resolution.
• Hourly granularity - one dataset item per location-hour, ready for time-series charts.
• Pollen forecasts - alder, birch, grass and ragweed grains/m3 (Europe).
• Both AQI standards - European AQI and US AQI in the same row.
• Human-readable AQI band - aqi_category maps European AQI onto Good / Fair / Moderate / Poor / Very poor / Extremely poor.
• Timezone-correct timestamps - each time value is full ISO-8601 with the location's real UTC offset.
• Clean, flat schema - no nested objects, exports straight to CSV, Excel, JSON or Google Sheets.

Input

Field	Type	Default	Description
mode	select	forecast	Scraping mode. Currently forecast - hourly air quality and pollen per location.
locations	array	["52.52,13.405,Berlin", "51.5074,-0.1278,London"]	One entry per location: "lat,lon" or "lat,lon,Label". The label is copied into location_name.
forecastDays	integer	3	Days of hourly forecast per location (24 rows per day). Max 7.
maxItems	integer	1000	Hard cap on total rows pushed across all locations. Max 10000.

Example input

{
  "mode": "forecast",
  "locations": [
    "52.52,13.405,Berlin",
    "51.5074,-0.1278,London",
    "40.7128,-74.0060,New York",
    "35.6762,139.6503,Tokyo"
  ],
  "forecastDays": 3,
  "maxItems": 1000
}

Output

One item per location-hour. Every field is nullable; hours where every pollutant is missing are dropped automatically.

{
  "location_name": "Berlin",
  "latitude": 52.5,
  "longitude": 13.400002,
  "timezone": "Europe/Berlin",
  "elevation": 37.0,
  "time": "2026-07-28T00:00:00+02:00",
  "pm10": 8.9,
  "pm2_5": 4.8,
  "carbon_monoxide": 134.0,
  "nitrogen_dioxide": 6.9,
  "sulphur_dioxide": 0.6,
  "ozone": 60.0,
  "dust": 0.0,
  "uv_index": 0.0,
  "european_aqi": 24,
  "us_aqi": 34,
  "alder_pollen": 0.0,
  "birch_pollen": 0.0,
  "grass_pollen": 6.0,
  "ragweed_pollen": 0.0,
  "aqi_category": "Fair",
  "source": "open-meteo",
  "scraped_at": "2026-07-28T13:45:35.738263+00:00"
}

Field reference

Field	Unit	Notes
location_name	-	Your label, or "lat,lon" if none supplied
latitude, longitude	degrees	Snapped to the model grid cell
timezone	IANA	Resolved automatically from coordinates
elevation	m	Grid-cell elevation
time	ISO-8601	Local wall-clock time with real UTC offset
pm10, pm2_5	ug/m3	Particulate matter
carbon_monoxide, nitrogen_dioxide, sulphur_dioxide, ozone	ug/m3	Gaseous pollutants
dust	ug/m3	Saharan/mineral dust
uv_index	index	Clear-sky UV index
european_aqi	EAQI	0-100+, EEA scale
us_aqi	US AQI	0-500, EPA scale
alder_pollen, birch_pollen, grass_pollen, ragweed_pollen	grains/m3	Europe only
aqi_category	-	Good / Fair / Moderate / Poor / Very poor / Extremely poor
source	-	Always open-meteo
scraped_at	ISO-8601	UTC run timestamp

AQI category bands (European AQI): 0-20 Good, 20-40 Fair, 40-60 Moderate, 60-80 Poor, 80-100 Very poor, above 100 Extremely poor.

Use cases

• Health apps - warn asthma, COPD and cardiovascular users hours before a PM2.5 or ozone spike instead of after it.
• Allergy forecasts - drive pollen calendars and medication reminders from real birch, grass, alder and ragweed counts.
• Environmental research - build reproducible multi-city hourly pollutant panels for exposure and epidemiology studies.
• Smart home - schedule air purifiers, HVAC recirculation and window-opening automations around forecast AQI.

Also useful for running/cycling apps, outdoor event planning, construction dust compliance, ESG reporting and insurance risk models.

Pricing

Roughly $0.5 per 1,000 rows, plus standard Apify platform usage. A 4-city, 3-day run is 288 rows - a few cents. One API call per location returns the entire forecast window, so cost scales with locations, not with hours.

Tips

• Fetching 7 days costs the same number of requests as 1 day; only row count changes.
• Set maxItems to locations x forecastDays x 24 to keep runs predictable.
• Pollen fields are modelled for Europe; expect nulls or zeros elsewhere.
• Schedule the Actor daily and append to the same dataset to build a forecast-accuracy archive.

Related Actors

• Air Quality Scraper - current air-quality conditions (the "now" counterpart to this Actor)
• MET Norway Weather Scraper - free global weather forecasts
• NOAA Weather Alerts Scraper - live US severe-weather alerts
• USGS Water Scraper - US river gauge and water-quality data

Data source and legal

Data by Open-Meteo.com, based on CAMS (Copernicus Atmosphere Monitoring Service) models, licensed under CC BY 4.0. Only public, non-personal environmental data is collected - no personal data, so no GDPR obligations arise from the output. Attribute Open-Meteo when you redistribute the data.

Development

python -m compileall -q src
python tests/test_smoke.py   # offline smoke tests, no network needed