Meteociel Historical Scraper

Scrape hourly historical weather observations from meteociel.fr for any station and date range. Each row in the output is one hour of recorded data: temperature, humidity, pressure, wind, rainfall, and weather condition.

What does it scrape?

MeteoCiel publishes hourly observation tables for thousands of WMO and ICAO weather stations worldwide. This Actor fetches those pages day by day and turns them into structured JSON rows, one per hour.

Supported station types:

• French and European WMO official stations (13-column table layout)
• ICAO airport stations (11-column reduced layout)
• Personal and private network stations

Why use this Actor?

• Historical weather for any location — back-fill climate data, validate models, or enrich datasets with weather context
• Structured output — clean JSON ready for analysis, no HTML parsing on your side
• No API key needed — data comes directly from the public MeteoCiel pages
• Flexible station lookup — provide a station code directly, or just type a city name and the Actor resolves it automatically

Typical use cases: correlating sales or energy consumption with weather, training ML models, building climate dashboards, or archiving observations for a specific location.

How to use it

1. Go to the Input tab in Apify Console
2. Enter a city name (e.g. Paris) or a station code (e.g. 7156)
3. Set Start date and End date in YYYY-MM-DD format
4. Click Start — the Actor scrapes one page per day and writes hourly rows to the dataset

Input

Field	Type	Required	Description
stationCode	string	no	Meteociel numeric station code (e.g. 7156 for Paris-Montsouris). Auto-resolved from cityName if omitted.
cityName	string	no	City or station name used to look up the station (e.g. Lyon, Praha). Also used as a label in the output. At least one of stationCode or cityName is required.
dateFrom	string	yes	First day to scrape — YYYY-MM-DD
dateTo	string	yes	Last day to scrape — YYYY-MM-DD
delayMs	number	no	Milliseconds to wait between requests (default: 2000)
maxRetries	number	no	Max retries per failed request (default: 3)

Example input:

{
  "cityName": "Paris",
  "dateFrom": "2024-01-01",
  "dateTo": "2024-01-31"
}

Output

One JSON object per hourly observation, pushed to the default dataset. You can download it as JSON, CSV, Excel, or XML from the Apify Console.

Example row:

{
  "city": "Paris",
  "stationCode": "7156",
  "date": "2024-01-15",
  "hour": "12:00",
  "temperature_c": 8.2,
  "humidity_pct": 76,
  "pressure_hpa": 1012.4,
  "wind_speed_kmh": 15,
  "wind_direction": "SO",
  "rainfall_mm": 0,
  "weather_condition": "Couvert"
}

Data fields

Field	Format	Description
city	string	City or station name as provided in the input
stationCode	string	Meteociel numeric station identifier
date	YYYY-MM-DD	Observation date
hour	HH:MM	Observation hour (local time)
temperature_c	number	Temperature in °C
humidity_pct	number	Relative humidity in %
pressure_hpa	number	Atmospheric pressure in hPa
wind_speed_kmh	number	Average wind speed in km/h
wind_direction	string	Wind direction code (e.g. N, SSO, NNE)
rainfall_mm	number	Rainfall in mm (0 when none recorded)
weather_condition	string | null	Observed weather condition in French (full-layout stations only)

Cost and performance

Each day of data is one HTTP request. A one-year scrape for one station is 365 requests, typically completing in 15–20 minutes with the default 2-second delay between requests.

This Actor uses usage-based pricing:

Event	Price
Result (apify-default-dataset-item)	$0.50 / 1,000 items
Actor start (apify-actor-start)	$0.00001

A one-year scrape produces 8,760 hourly rows ($4.38) plus a one-time start fee.

Finding a station code

If you know the city name, just set cityName — the Actor queries the MeteoCiel station search API automatically and picks the best match (WMO official stations preferred over ICAO, then private networks).

To find a code manually, browse to a station page on meteociel.fr and note the code2= parameter in the URL.

Legal notice

This Actor scrapes publicly available pages on meteociel.fr for personal and research use. Always respect the site's Terms of Service and do not overload the server. The default delay of 2 seconds between requests is intentional. The authors are not affiliated with MeteoCiel.