Dr. Basil Kraft

Project Leader
Department Biogeochemical Integration (BGI)
Global diagnostic models
+49 3641 57-6233

Main Focus

Deep learning for time series modeling

Ecosystem and Earth System dynamics are often hard to model and predict and today, still many machine-learning approaches fail in encoding temporal dependencies that typically characterize dynamical natural systems. Yet, deep learning approaches in the time domain (e.g., Recurrent Neural Networks or Transformer models) are theoretically prepared to address such issues. I explore how these data-driven approaches can be used to better describe, predict, and understand Earth system processes on the land surface via the modeling of Earth observation data.

Hybrid modeling

Hybrid modeling combines physically-based modeling and machine learning. The approach allows to use prior physical knowledge to regularize data-driven models and ultimately, to achieve better predicability of complex Earth system processes. At the same time, hybrid models are partially interpretable, allowing to gain scientific insights.

Curriculum Vitae


Conferences & workshops

  • 2020 - EGU talk "Towards global hybrid hydrological modeling by fusing deep learning and a conceptual model"
  • 2019 - Workshop on ecological memory effects (talk, participant), Oak Ridge Natiaonal Labs (ORNL)
  • 2019 - Introduction to Deep Learning (talk), Institute of Coastal Research, HZG
  • 2019 - EGU talk "Identifying dynamic memory effects on vegetation state using recurrent neural networks"
  • 2019 - Workshop "Deep Learning 101" (host) at the Earth System PhD Conference (ESPC) at the Max Planck Institute for Biogeochemistry
  • 2019 - Co-organizer of the Earth System PhD Conference (ESPC) at the Max Planck Institute for Biogeochemistry
  • 2018 - Co-chair of the "Deep Learning for Environmental Science & Ecology" session at the International Conference on Ecological Informatics (ICEI)

Academic background

Go to Editor View