Machine Learning for Hydrological and Earth Systems
Dr. Shijie Jiang
Mission
In the face of increasing environmental and climatic pressures, understanding the variability and causality of biogeochemical cycles across spatial and temporal scales is essential. The Machine Learning for Hydrological and Earth Systems (ML4HES) group, part of the ELLIS Unit Jena, explores how environmental and climate sciences can benefit from advances in machine learning integrated with domain science. Our research combines data-driven insights with domain knowledge to investigate feedbacks, interactions, and responses within Earth systems, with a particular focus on advancing our understanding of coupled hydrological, ecological, and climate systems.
Focus areas
Ecohydrological and climate systems
We investigate the interactions between ecohydrological and climatic processes that drive water and energy fluxes across regions and scales. Our research focuses on understanding the role of vegetation and its feedback mechanisms in shaping hydroclimatic dynamics, including extreme events.
Soil-plant hydraulics
We study water transport and regulation within the soil-plant-atmosphere continuum, focusing on how vegetation responds to and influences environmental changes. Our work emphasizes root zone processes and uses diverse datasets to explore plant-water interactions, particularly under drought conditions.
Watershed hydro-biogeochemistry
Our research examines how hydrological and biogeochemical processes operate and interact at the watershed scale. We study the cycling of water, energy, carbon, and nutrients, with an emphasis on watershed responses to both natural and human-induced variability.
Research approach
Methodologically, our group combines advanced data-driven techniques with classical Earth system analysis to improve the modeling and understanding of complex dynamics within interconnected Earth subsystems. We emphasize the integration of machine learning with domain knowledge to develop robust methods for identifying dependencies, interactions, couplings, feedback mechanisms, and causal relationships across biogeochemical processes to address key scientific questions.
News
Team
Associated group members
