Earth’s ‘Digital Twin’ Designed to Simulate Terrestrial Water Cycle Could Help Predict Possible Natural Disasters Before They Strike

The water cycle may seem simple in theory, but it is hard to predict in practice. This process can get complicated when a continuously changing climate becomes involved. Experts will need high-resolution data across an immense expanse to model water on Earth.

(Photo: Wikimedia Commons/ NASA/JPL)

The modeling must be sophisticated enough to include everything from soil moisture in valleys to snow caps on mountains. Just recently, researchers have made a step forward by building the most detailed models created to date.

Digital Earth Simulation

Led by Dr. Luca Brocca of the National Research Council of Italy, a team of researchers created digital twin case studies for the water cycle in the Mediterranean Basin. They aim to design a system that enables non-experts to run interactive computer simulations. The result of their work is described in the study "A Digital Twin of the Terrestrial Water Cycle: a glimpse into the future through high-resolution Earth observations."

By constantly updating with new data, the digital twin of the Earth will permit scientists to simulate best- and worst-case scenarios. It will also allow them to assess risks and monitor the development of hazardous conditions before they occur. This information is very important for sustainable development and protecting vulnerable populations.

Brocca and his team used extraordinary volumes of satellite data to build the digital twin models. They combined them with new Earth observation data that quantify precipitation, soil measurements, river discharge, evaporation, and snow depth. The newly available data includes measurements obtained much more frequently across space and time, as often as once an hour and once a kilometer.

The higher-resolution data creates a more detailed image, like a screen with more pixels. Scientists use this data to develop their model, which is integrated into a cloud-based platform for simulations and visualizations. Ultimately, this leads to the creation of an interactive tool that anyone can use to map risks such as landslides and floods.

Planning the Future With Simulation

The research team began by modeling the Po River valley. Then, they expanded the digital twin to other parts of the Mediterranean region. In the future, the experts plan to extend it to cover all of Europe, with collaborations that will enable the same principles to be applied throughout the globe.

The platform's primary use is to provide enhanced flood and landslide prediction and optimized water resource management. More granular data and sophisticated modeling will be needed to make this project work better on a more local level. For example, maximizing the potential of a digital twin for agriculture would require data resolution measured in tens of meters instead of hundreds.

Additional challenges should also be considered. These include more ground observations for satellite data validation, delays in transferring satellite data to the model, and the increasing complexity of algorithms needed for data handling.

No model is perfect, and so is the Earth's digital twin. Furthermore, satellite data can contain uncertainties that must be properly characterized to give users an accurate picture of the model's reliability. According to Brocca, machine learning, and artificial intelligence can significantly address these challenges. This can be done by streamlining data quality assessment and by enhanced data analysis, collection, and processing speed.