Description:

This session will examine the integration of artificial intelligence (AI) and climate modeling, focusing on applications to predictions and projections, requirements and frameworks for responsible usage of AI, and the potential utilization for climate services. Participants will discuss practical examples, data requirements, transparency in data and methodologies, verification benchmarks, risks, biases, infrastructure needs, and the business model for sustainable and responsible progress that supports underpinning data, modeling and climate expertise. Participants will contribute to the dialogue regarding responsible integration of AI for climate applications.

1. AI Applications in Climate Modeling:
• Identifying data types used, including observational and climate model outputs.
• Comparing advantages and risks over conventional methodologies.
• Discussing the feasibility of AI-based climate models discussing data requirements and foundational principles.
2. Requirements for Responsible AI Usage in Climate Modeling:
• Addressing transparency, data quality, and methodology transparency.
• Discussing strategies and frameworks for independent verification, benchmarking and explainability.
• Discussing bias issues and equitable access.
• Identifying potential risks and biases.
• Discussing strategies for mitigation and responsible usage.
3. Credit and Return Mechanisms:
• Discussing crediting mechanisms for underpinning climate observations, science expertise, modeling and projections.
• Considering business models for sustained progress.
4. Explainability and Verification for Predictions and Projections:
• Examining need for explainable AI for reliable projections.
• Discussing methods for validating AI-generated climate projections.
5. Potential New Climate Services Enabled by AI:
• Exploring novel climate services.
• Analyzing advantages, risks, and deployment considerations.
• Discussing requirements for transparency and benchmarking.
6. Infrastructure and Workforce Development:
• Discussing infrastructure needs for AI-ready climate data and models.
• Exploring strategies for developing an AI-ready climate workforce.