AI/ML pervades Geodesy, being commonly used for filling data gaps, disaggregating components within integral geodetic observational constraints, detecting signal patterns, defining input/output relationships from data, and building reduced order representations of complex dynamical systems.

Existing AI/ML approaches fail to leverage first principles of geodesy, risking interpretability and reliability. While geodetic data volume is increasing fast, these do not yet commonly inform AI/ML development. This suggests that AI/ML methods should be grounded in geodetic first principles and tailored to the field's unique needs. Progress in this direction will require closer collaboration between the geodesy and computer science communities. This session seeks contributions that can help accelerate such collaborative development.

Papers are invited that (i) apply AI/ML methods to Geodesy and its applications; (ii)  provide insights into the deficiencies of the common AI/ML methods; and (iii) report on early experiences in collaborative development of a new paradigm of Geodesy Informed Machine Learning.

• C - Cryosphere
• GC - Global Environmental Change
• H - Hydrology
• IN - Informatics