Skip to content

Earth Data Science: Integrating datasets and machine learning techniques to understand geospatial processes

Earth science and engineering is at the core of several pressing societal questions. Climate change impacts the dynamic of our planet, from the atmosphere to the oceans and the continents. The need to move away from fossil fuel also means we will need to rely more on so called ‘Geo-energies’, such as hydrothermal heat, geological reservoir of hydrogen, and carbon capture and storage in the subsurface. And humanity’s constant pull towards exploration of the planets of our solar systems also mean that an understanding of planetary processes is needed.

Earth Data Science is characterized by challenging datasets: the data encompasses a large range of temporal and spatial scales, it typically is large (‘Big Data’) but also typically spatially sparse. Machine learning and deep learning approaches show great promises in this field, but optimal application of these approaches in Earth Data Science is challenging: this is what our proposal aims to achieve. Our research initiative leverages on existing research and takes it to the next level by promoting integration across fields and across departments to relevant societal problems, and builds upon our strong foundation of teaching master level courses in data science and machine learning.

Led by Dr Cédric M. John.


Our research