Topic: Artificial Intelligence in Earth Sciences, Dos and Don’ts

Time: 3:30-5:30pm January 13, 2025

Venue: Meeting Room A A901, IGSNRR

Speaker: Professor Vahid Nourani, University of Tabriz, Iran

Abstract:

The rapid development of AI technology has unlocked new opportunities in Earth sciences, offering innovative ways to model complex processes and phenomena. However, like all mathematical tools, AI models have limitations, particularly due to their inherent "black-box" nature. These limitations necessitate cautious and well-informed applications to prevent potential shortcomings. To address these challenges, it is essential for users to possess a deep understanding of the physical principles underlying the processes. Reliable interpretation of results is crucial for ensuring that AI-driven insights are scientifically sound and actionable. This lecture aims to i) present critical perspectives on the application of AI in Earth sciences, highlighting potential pitfalls and limitations, ii) Explain, in straightforward language, how AI models are trained and deployed for modeling complex processes—especially in cases where physically based models are infeasible or insufficient, iii) Provide practical advice and recommendations for achieving reliable, high-performance outcomes from AI-driven models. By combining AI tools with domain expertise, scientists can leverage their potential while maintaining the integrity and reliability of Earth science research.