PDE foundation models are skillful AI weather emulators for the Martian atmosphere

2026-02-16

AI Summary

This study addresses the challenges of Martian atmospheric weather prediction, which are primarily constrained by data scarcity and limited computational resources. For the first time, the authors adapt Poseidon—a PDE-based foundation model originally pretrained on Earth’s atmospheric dynamics—to Martian meteorological modeling. They propose an architectural extension that scales the model from two to three dimensions and leverage transfer learning combined with fine-tuning to enable efficient training on only 34 GB of four-year Martian weather data. The approach achieves a 34.4% performance improvement over baseline methods on held-out annual test sets, with a median training cost of just 13 GPU hours. These results demonstrate the strong generalization capability and practical potential of PDE-based foundation models in real-world, complex physical systems such as the Martian atmosphere.

Technology Category

• Machine Learning: Deep Neural Architectures and Foundation Models
• Multiagent Systems: Other Foundations of Multi Agent Systems
• Cognitive Modeling & Cognitive Systems: Other Foundations of Cognitive Modeling & Systems

Application Category

• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• Web Mining and Content Analysis: Large pretrained models with web data
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications

Abstract

We show that AI foundation models that are pretrained on numerical solutions to a diverse corpus of partial differential equations can be adapted and fine-tuned to obtain skillful predictive weather emulators for the Martian atmosphere. We base our work on the Poseidon PDE foundation model for two-dimensional systems. We develop a method to extend Poseidon from two to three dimensions while keeping the pretraining information. Moreover, we investigate the performance of the model in the presence of sparse initial conditions. Our results make use of four Martian years (approx.~34 GB) of training data and a median compute budget of 13 GPU hours. We find that the combination of pretraining and model extension yields a performance increase of 34.4\% on a held-out year. This shows that PDEs-FMs can not only approximate solutions to (other) PDEs but also anchor models for real-world problems with complex interactions that lack a sufficient amount of training data or a suitable compute budget.

Problem

• Martian atmosphere
• weather prediction
• sparse data
• limited compute budget
• AI emulator

Innovation

• PDE foundation models
• Martian weather emulation
• 3D model extension
• sparse initial conditions
• transfer learning