UrbanAIR: Designing Resilient Cities with the Earth’s Digital Twin

With heat wave records being broken across Europe and the world, more uncertain and unpredictable weather patterns and more frequent and violent storms,  the global community is rightly concerned about the effects of climate change.

To mitigate these impacts, we need more than just data—we need the ability to test solutions in a virtual environment before implementing them in our physical streets.

Bridging the Gap: From Global Data to Local Streets

Destination Earth (DestinE) is the European Union's flagship programme to build a Digital Twin (DT) of the entire Earth. Digital Twins create a digital model of the real-world component that they “twin", allowing modellers to 

• Visualise the current state environmental systems

• Simulate future scenarios based on varying climate inputs.

For example, a DT of an aircraft engine helps monitor the engine for any issues that may arise before they become problems, and to let engineers service the engine to help keep it performing, which in turn helps keep flight passengers safe. 

Engineers cannot, sadly, bring the Earth in for servicing, but we can model future climate change scenarios at various scales.

The UrbanAIR project works with a set of “pilot cities” to study the impact of climate change on urban environments.  Working with Barcelona, Paris, Bristol, Antwerp and Rotterdam, we want to enable urban planners to make their cities safer and improve public health.  Can we make cities greener?  Can we have better air quality? Can we protect vulnerable people from extreme heat? The UrbanAIR project can enable urban planners to do this. 

The Technical Engine: Data Lakes and City Models

^{Jupyter Notebooks on DestinE are interactive development environments for code and data} 

Part of DAFNI’s role is to provide technical expertise on how to use the DestinE’s services. We act as a bridge between specialist scientists and local authority urban planners. One of the important resources is the DestinE Data Lake, which brings it all together: climate data, city data, compute capacity for simulations and modelling, and user management.

The city models can include detailed data —often down to 1–2 metres — such as land use, elevation, soil types, demographics, buildings, road types, rivers, green areas, traffic, etc. All of these come in different formats, shapes and sizes. To bridge the gap between climate data and  city models, UrbanAIR  utilises advanced infrastructure: 

• Object Stores (S3): For robust data storage.

• Islet (Cloud Edge): For efficient pre-processing and downscaling of data.

Object Store on DestinE 

What’s Next for UrbanAIR?

Notebook structure​

Our demo from last year successfully ran VITO's models for Antwerp within a notebook in the Data Lake. Since then, we have expanded our use of the Data Lake to the object stores, the Islet cloud, and services for building and running data processing and pipelines. The project team at STFC is now working with Barcelona and Paris and look forward to sharing these cities demonstrations later this year.

The goal is to give the city modellers a user-friendly front end to enable them to visualise their city and model how to improve the city and mitigate the local effects of climate change.