Weather data to forecast the spread of COVID-19

Belgian Weather per day per province by using and data sources

The software collects the Weather of the Belgian communes from the OpenWeatherMap database. See

The weather data will be linked to other data to answer one question: “Does the usage of the Weather data influence (or not) the model forecasting the spread of the COVID-19?”

The context

Given the consequences of the COVID-19 pandemic for public health, many stakeholders in public and private sectors engage in global efforts to treat and understand those exposed to the virus and to contain the outbreak.

Universities and private companies contribute to the analysis of data useful to forecast the spread of the “Severe Acute Respiratory Syndrome CoronaVirus 2” (SARS-CoV-2).

Those persons include researchers and consultants who share their analysis to the public administration and members of governments who will shape the rules and regulations describing how the citizens should behave during the pandemic.

Why collecting the weather data?

In Belgium, the Group of Experts for an Exit Strategy (GEES) drafts proposals to envisage a gradual deconfinement. The GEES supports the Belgian Federal Government.

See also

The GEES uses data from researchers, amongst others. Those researchers build models of the spread of the virus by using diverse data: mobility, health, beds in hospitals, fatalities, current regulation …

One question remains: does the usage of the Weather data influence (or not) the spread of the virus?

One of those groups of researchers includes the Machine Learning Group (MLG) from the Université Libre de Bruxelles. The MLG supports the initiative by providing models made of Artificial Intelligence/Machine Learning algorithms. The models built by the MLG influence the Exit Strategy.

Technical challenges

OpenWeatherMap does not store the Postal code of the communes in Belgium. The Belgian Post provides the Belgian postal codes in French and Dutch

This project joins both databases to enable a link between diverse types of data: mobility, postal codes, etc.

The tools used to build the software

The sponsors of the project

EOSC Secretariat logo

The EOSC Secretariat supports the governance of the European Open Science Cloud (EOSC). See

The EOSC supports projects aiming to make data Findable, Accessible, Interoperable, and Reproducible (FAIR) for scientists; these combinations would lead to (unforeseen) reuse and faster development of science.

EOSC Secretariat granted funds to to create a “Crossroads Data Bank for COVID-19”.

Latest news from the European Commission: logo

OpenWeatherMap provides, free-of-charge, access to weather data to its historical weather data

OpenWeather Ltd is a British-based tech company that provides weather and satellite data worldwide. OpenWeather collects and processes raw data from a variety of sources, and gives its customers access to the archive. See

See Openweather LTD Helps The Fight To Overcome COVID-19

The supporters of the project

Machine Learning Group logo

The Machine Learning Group, founded in 2004 by Gianluca Bontempi, is a research unit of the Computer Science Department of the ULB (Université Libre de Bruxelles, Brussels, Belgium), Faculty of Sciences, currently co-headed by Prof. Gianluca Bontempi and Prof. Tom Lenaerts.

MLG targets machine learning and behavioral intelligence research focusing on time series analysis, big data mining, causal inference, network inference, decision-making models, and behavioral analysis with applications in data science, medicine, molecular biology, cybersecurity and social dynamics related to cooperation, emotions, and others.


Legal matters

The current software digests data that do not contain personal data.

The output of this project is linked to data that do not contain Personally Identifiable Information (PII). See

The data models

Belgian Postal Codes

Here are the columns describing the postal codes:

Index(['country code', 'postal code', 'place name', 'admin name1',
       'admin code1', 'admin name2', 'admin code2', 'admin name3',
       'admin code3', 'rounded_lat', 'rounded_lon', 'accuracy'],

OpenWeatherMap List of cities IDs

Here are the columns describing the cities for which OpenWeatherMap tracks the Weather:

Index(['id', '$numberLong', '', 'city.findname',
       '', 'city.coord.lon', '',
       'city.zoom.$numberLong', 'id.$numberLong', 'city.coord.lon.$numberLong',
       '$numberLong', 'rounded_lat', 'rounded_lon'],

OpenWeatherMap Historical data

OpenWeatherMap provides an API returning the hourly Weather. A call to the API returns a maximum of 7 consecutive days of data.

Here is an example of a call:{id}&type=hour&start={start}&end={end}&appid={YOUR_API_KEY}

The response from the history of OpenWeatherMap

The documentation of the response from the history of OpenWeatherMap is here:

    "dt": 1587312000,
    "main": {
        "temp": 287.44,
        "feels_like": 284.3,
        "pressure": 1015,
        "humidity": 76,
        "temp_min": 287.04,
        "temp_max": 288.15
    "wind": {
        "speed": 4.6,
        "deg": 40
    "clouds": {
        "all": 77
    "weather": [
            "id": 500,
            "main": "Rain",
            "description": "light rain",
            "icon": "10d"
    "rain": {
        "1h": 0.3

What are the data available in the final data set?


ALT-F1 SPRL logo

ALT-F1 SPRL has built the software. See

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.

<span>%d</span> bloggers like this: