• The National Oceanic and Atmospheric Administration (NOAA) established the
  Hurricane and Ocean Testbed (HOT) in 2021.
• HOT was developed to test new models and products to improve hurricane forecasts, but it was not designed to handle artificial intelligence (AI) models.
• This creates a disconnect between AI researchers developing promising models and the operational community that could benefit from them.

Plan the expansion of NOAA’s Hurricane and Ocean Testbed so artificial intelligence models can be evaluated using operational forecasting standards, workflows, and metrics.

Connect AI researchers with operational hurricane forecasters to align scientific innovation with real-world forecasting needs, constraints, and decision-making processes.

Advance hurricane prediction capability in support of public safety and resilience, consistent with NSF’s mission to promote national health, prosperity, and welfare.

Enable guest AI researchers to evaluate models within NOAA’s Hurricane and Ocean Testbed using operational workflows and standards.

• Invite guest AI researchers and serve as the liaison to the testbed.
• Develop a tiered test protocol for AI models.
• Identify AI infrastructure needs for expansion.
• Extract insights from model performance.

Convene stakeholders to define requirements and produce a practical plan for expanding HOT into an AI-ready testbed.

• Conduct meetings with testbed users and operational partners.
• Hold a workshop to design an AI-ready expansion framework.
• Develop an expansion plan and implementation roadmap.
• Assess which elements generalize to other testbeds.