arXiv Science⌕ Search

arXiv · 2611.12787

MarsNet: Learning the Martian Atmosphere from Observations Alone

Abstract

We present MarsNet, a machine-learned model of the Martian atmosphere trained only on observations. Existing reanalyses assimilate spacecraft retrievals into general circulation models (GCMs), so where observations are sparse their fields largely reflect the model. We therefore use extant reanalyses only for evaluation. MarsNet alternates a two-hour forecast with an update that assimilates the observations of each window, and both steps are learned jointly. Because the Martian record is too sparse for an unrestricted network, learning is confined within a fixed physical structure: the state is in physical units, the forecast uses physical tendencies of fixed form and learned strength, and the update is a bounded Kalman update. MarsNet is trained on Mars Years (MY) 24--31 and evaluated on independent observations. On three seasons of MGS radio-occultation temperatures assimilated by no dataset, MarsNet matches or improves on the three reanalyses in the lowest three scale heights, with a near-surface root-mean-square error (RMSE) of 4.2 K against 4.4--5.4 K, but is less accurate aloft. The model's surface pressures are the most accurate of the four products (RMSE 10 Pa against Curiosity in MY32, compared with 20--46 Pa). On withheld Mars Climate Sounder profiles of MY32, its temperatures beat an observed climatology and persistence, including in ten-sol forecasts without new observations, and its stated uncertainty is close to calibrated. Much of the temperature skill is present before training and appears to come from the structure and tables fitted to the observations, while training mainly improves the surface pressures, at some cost to dust. The model cycles stably through nine Mars years, and the influence of each observation can be traced. Its speed makes experiments on the observing record practical, and it offers an independent reference for GCMs and reanalyses.

Explore related subjects

Keep this discovery

Explore connections, maps & timelines

BibTeXRIS

Sebastian Mengwall, J. Michael Battalio. 2026-11-09. MarsNet: Learning the Martian Atmosphere from Observations Alone. https://arxiv.science/abs/2611.12787

Citations retain the original authors and DOI and include this site’s record URL. If this discovery service contributed to your research, you can also cite arXiv Science by Indiciae: BibTeX · RIS.

KEEP EXPLORING

Related papers

Testing the prevalence of hydrogen-silicate miscibility in young sub-Neptunes

Hydrogen-silicate miscibility can significantly alter the interior structure and thermal evolution of sub-Neptunes. We consider the interplay between this miscibility and stellar-driven atmospheric escape. We find that, for the first $\sim 100$ Myrs, sub-Neptunes store most of their hydrogen content within their miscible interiors, protecting it from escape. As hydrogen is removed from the top of the atmosphere, more hydrogen is exsolved from the miscible interior, resupplying the envelope mass and delaying the planet's contraction when compared with models that do not account for miscibility. We construct a population-level Bayesian hierarchical inference model which tests for the prevalence of fully miscible sub-Neptunes. With conservative measurement uncertainty and no completeness corrections, we predict that $\sim 100-200$ observed young sub-Neptunes with measured radii, masses and ages $\lesssim 1$ Gyr are required to differentiate between fully miscible and immiscible models. This survey size requirement decreases with improved measurement uncertainty. Furthermore, one can reliably infer the underlying distributions of planetary initial conditions if completeness corrections are included in the analysis.

astro-ph.EP↗

AEGIS: Differentiable Mars Climate Model with Neural Closures

General circulation models (GCMs) are the primary tool for simulating planetary atmospheres. They play a vital role in understanding Mars's atmosphere, as forecasting its unique weather is mission-critical for operations such as entry, descent, and landing. Mars poses unusual challenges for these models, as observations are sparse compared to Earth. In addition, a thin $CO_2$ atmosphere alongside a radiatively active dust cycle creates a volatile atmosphere with large diurnal temperature swings and no true terrestrial analog for validation. Existing Mars GCMs, including the LMD PCM, the NASA Ames Mars GCM, and PlanetWRF, are mature and physically detailed but are implemented in legacy Fortran with finite-difference or finite-volume solvers, and they do not expose gradients for calibration or machine learning. Here we present AEGIS, a modular differentiable Mars climate model that couples Mars's unique atmospheric physics to the Dinosaur dynamical core, with interfaces for neural closures. We showcase stable ten-Mars-year simulations that reproduce the seasonal $CO_2$ cycle while conserving the total $CO_2$ inventory, capture realistic large-scale surface-temperature structure, and produce surface pressure that follows Mars Orbiter Laser Altimeter (MOLA) topography. Gradients through coupled trajectories agree with finite differences and support physical calibration and neural training. We compare with conventional GCMs, highlighting the framework's computational efficiency and differentiability.

astro-ph.EP↗

Identifying deformed exoplanets with JWST and Ariel: A target selection metric for phase curve observations

The shape of an exoplanet encodes fundamental information about its internal structure. For close-in giant planets, strong stellar tidal forces induce measurable deformations that can be probed through high-precision photometry. These deformations are quantified by the fluid Love numbers, which characterise a planet's response to external perturbations and are directly linked to its internal density distribution. In this work, we assess the potential of JWST and the future Ariel mission to constrain the Love numbers of hot Jupiters through full-orbit phase curve observations. To identify the most promising targets for this science case, we introduce a new detection metric, the deformation metric (D), which combines the expected deformation amplitude with the anticipated photometric signal-to-noise ratio. We apply this metric to the Ariel target list and identify WASP-33 b, WASP-76 b, and HAT-P-32 b as the three most favourable targets that currently lack Love number measurements. Using WASP-76 b as a benchmark case, we perform end-to-end retrieval simulations and show that a single continuous JWST phase curve observation can constrain the Love number $h_{2f}$ to a precision of $σ_{h_2f} \approx 0.09$, while a single Ariel visit could achieve $σ_{h_{2f}} \approx 0.18$. Such precision is likely sufficient for Love number measurements to provide meaningful constraints on planetary interior models and to distinguish between competing interior-structure scenarios, opening a new observational avenue for probing the deep interiors of exoplanets.

astro-ph.EP↗