Weather Data from Open Meteo API

This repository contains a tool to easily access weather data using the Open Meteo API. By specifying a location and a time period, you can import historic and upcoming hourly weather data from around the world with different measurements such as temperature, rain, snow, wind and more.

The weather data comes from the Open Meteo API:

• The Open Meteo Archive API, which provides historical weather data
• The Open Meteo Forecast API, which provides forecasted weather data

https://open-meteo.com/en/docs

Usage

The script contains four functions:

• get_latest_recorded_date() - Returns the latest recorded date in the Open Meteo Archive API database.
• get_archived_weather(lat, long, start, end, metrics) - Returns historical weather data for a given latitude and longitude, time range, and set of metrics.
• get_forecast_weather(lat, long, metrics) - Returns forecasted weather data for a given latitude and longitude and set of metrics.
• get_distance(lon1, lat1, lon2, lat2) - returns the distance in kilometers between two sets of longitude and latitude coordinates.

To use any of these functions, simply import the weather_data.py file and call the desired function with the appropriate parameters.

Example

Here's an example of how to use the get_archived_weather() function to retrieve historical weather data for a location:

import weather_data

lat = 43.296482 # Marseille latitude
long = 5.36978 # Marseille longitude
start = '2023-03-01'
end = '2023-03-25'
metrics = ['temperature', 'precipitation']

data = weather_data.get_archived_weather(lat, long, start, end, metrics)

This will return a dictionary containing hourly weather data for Marseille for the month of March 2023, including temperature and precipitation metrics.

Then you can view the data in the form of a dataframe as follows:

import pandas as pd

pd.DataFrame(data)

>>      latitude  longitude              time  temperature_2m  precipitation
   0    43.40001   5.300003  2023-03-01T00:00             4.5            0.0
   1    43.40001   5.300003  2023-03-01T01:00             4.1            0.0
   2    43.40001   5.300003  2023-03-01T02:00             3.7            0.0
   3    43.40001   5.300003  2023-03-01T03:00             3.1            0.0
   ..        ...        ...               ...             ...            ...
   596  43.40001   5.300003  2023-03-25T20:00            13.0            0.0
   597  43.40001   5.300003  2023-03-25T21:00            12.4            0.0
   598  43.40001   5.300003  2023-03-25T22:00            11.5            0.0
   599  43.40001   5.300003  2023-03-25T23:00            11.7            0.0

If you want to import weather data for different locations, you can do the following (with or without multiprocessing):

from multiprocessing import Pool
from functools import partial

params = [(48.85826,   2.294499),  # Eiffel Tower coordinates
          (40.689253, -74.044547), # Statue of Liberty coordinates
          (27.175012,  78.042097)] # Taj Mahal coordinates

start = '2023-03-01'
end = '2023-03-25'
metrics = ['snowfall', 'windspeed_10m']

with Pool(4) as pool:
    data = pool.starmap(partial(weather_data.get_archived_weather, start=start, end=end, metrics=metrics),
                        params)

pd.concat([pd.DataFrame(d) for d in data])

>>       latitude  longitude              time  snowfall  windspeed_10m
   0    48.900010   2.300003  2023-03-01T00:00       0.0           14.7
   1    48.900010   2.300003  2023-03-01T01:00       0.0           15.1
   2    48.900010   2.300003  2023-03-01T02:00       0.0           15.1
   3    48.900010   2.300003  2023-03-01T03:00       0.0           14.7
   ..         ...        ...               ...       ...            ...
   596  27.200005  78.000000  2023-03-25T20:00       0.0            6.9
   597  27.200005  78.000000  2023-03-25T21:00       0.0            8.2
   598  27.200005  78.000000  2023-03-25T22:00       0.0           11.9
   599  27.200005  78.000000  2023-03-25T23:00       0.0           12.8

Credits

This repository is maintained by Antoine PINTO ([email protected]). It is based on the Open Meteo API, which provides the weather data.