Towards a Foundation Model for the Martian Atmosphere
Sujit Roy, Udayshankar Nair, Yuling Wu, Georgios Priftis, Liping Wang, Anastasia Georgiou, Anne Jones, Björn Lütjens, Johannes Schmude, Campbell Watson, Rachel A. Slank, Ankur Kumar, Anirbit Mukherjee, Procheta Sen, Ramin Lolachi, Haonan Chen, Manil Maskey, Juan Bernabé-Moreno, Rahul Ramachandran
astro-ph.EP, astro-ph.IM, cs.LG, physics.ao-ph
Submitted: 2026-05-16
License: http://creativecommons.org/licenses/by/4.0/
The gist: The martian atmosphere hosts dynamical phenomena ranging from planet-encircling dust storms to mesoscale orographic clouds and nocturnal low-level jets.
Terminology
Abstract
The martian atmosphere hosts dynamical phenomena ranging from planet-encircling dust storms to mesoscale orographic clouds and nocturnal low-level jets. General circulation model show capability to simulate these phenomena, but is computationally expensive at resolution needed to resolve mesoscale features. While assimilation of satellite remote sensing observation enable forecasting capabilities using such models, observation record is often sparse, short and fragmented across instrument generators. These constraints motivate the development of a data-driven foundation model for the Martian atmosphere. Foundation models live in a complex design landscape. There is an interplay between the available data, the physics of the underlying processes and corresponding developments in AI. Even though the idea of a foundation model is to address multiple use cases in a data- and compute-efficient manner, it is important to have a clear picture what applications can sensibly addressed by a single model. The purpose of this paper is to elucidate this design landscape. We discuss available data ranging from atmospheric retrievals to reanalysis datasets as well as existing physical models. Moreover, we identify a wide range of candidate downstream applications. Finally, we consider relevant recent developments in artificial intelligence (AI) that can be leveraged in this context. Here, we put a particular emphasis on AI models for atmospheric physics, data-driven approaches to data assimilation as well as methods to work in a limited data setting.
Sources
- FourCastNet: A Global Data-driven High-resolution Weather Model using Adaptive Fourier Neural Operators
- Forecasting Global Weather with Graph Neural Networks
- FengWu: Pushing the Skillful Global Medium-range Weather Forecast beyond 10 Days Lead
- AIFS -- ECMWF's data-driven forecasting system
- ClimaX: A foundation model for weather and climate
- AtmoRep: A stochastic model of atmosphere dynamics using large scale representation learning
- Prithvi WxC: Foundation Model for Weather and Climate
- AI Foundation Models for Weather and Climate: Applications, Design, and Implementation
- Surya: Foundation Model for Heliophysics
- An update to ECMWF's machine-learned weather forecast model AIFS
- Learning to Advect: A Neural Semi-Lagrangian Architecture for Weather Forecasting
- GenCast: Diffusion-based ensemble forecasting for medium-range weather
- Skillful joint probabilistic weather forecasting from marginals
- Demystifying Data-Driven Probabilistic Medium-Range Weather Forecasting
- Deep Learning for Day Forecasts from Sparse Observations
- HealDA: Highlighting the importance of initial errors in end-to-end AI weather forecasts
- GraphDOP: Towards skilful data-driven medium-range weather forecasts learnt and initialised directly from observations
- AI-based data assimilation: Learning the functional of analysis estimation
- DiffDA: a Diffusion Model for Weather-scale Data Assimilation
- Ambient Physics: Training Neural PDE Solvers with Partial Observations
Related papers
- PDS 70 c and SR 12 c: Observational Constraints on Giant-Planet and Satellite Formation
- Two-stage disruption of resonant chains
- Detectability of resolved hydrogen lines from the accretion shock at gas giants and their CPDs
- Binary-lens Microlensing Degeneracy: Impact on Planetary Sensitivity and Mass-ratio Function
- Atmospheric escape fractionates secondary but not primary atmospheres
- The Occurrence Rate of Nearby Planetary Companions to Hot Jupiters