PhD position: Develop Hybrid Machine Learning for Global Soil Mapping

In this role, you develop novel process-informed AI tools and advance hybrid soil mapping approaches integrating expert-knowledge. Curious? Check the vacancy!

Are you excited to develop cutting-edge AI methods that contribute to solving global environmental challenges? Join the Department of Physical Geography at Utrecht University as a PhD candidate and help shape the next generation of geospatial modelling.

Your job

Predicting ecosystem dynamics under global change requires accurate and consistent soil information at global scale. Current global soil maps are derived using empirical machine learning that often ignores known soil processes, resulting in inconsistencies across soil properties and underperformance in data-scarce regions.

This PhD project will develop next-generation machine learning methods for geospatial prediction by integrating process knowledge into modern AI tools. The research offers the opportunity to explore neural network arc...