Objectives and Expected Breakthroughs
in addressing the challenges in AI generalization for hydro-climatic extremes
-
Objective 2: Establish uncertainty quantification frameworks for a responsible use of AI in out-of-domain settings
-
Objective 3: Enhance Earth system understanding and predictability of extreme events
Breakthrough 3: Extract Insights
Graphic: KI generatedBreakthrough EXTRACT INSIGHTS:
Leverage causal modeling, hybrid modeling, and equation discovery to extract scientifically meaningful insights on the drivers and effects of hydro-climatic extremes, improving physical interpretability and predictive power.
-
Objective 4: Translate AI advances into actionable environmental assessment and management
Breakthrough 4: Demonstrators
Graphic: KI generatedBreakthrough DEMONSTRATORS:
Bridge fundamental AI research with real-world applications by developing generalizable AI demonstrators in the chosen hydro-climatic application domains related to hazards (floods, landslides) and ecosystem responses (phenological shifts, ecosystem fluxes).