跳到正文
arXiv cs.AI· Sameera S Kashyap, Victor Cruz, Angel Yepez, Razvan Marinescu·· 4 小时前AI 评分26

AEGIS:带神经闭合接口的可微分火星气候模型

AEGIS: Differentiable Mars Climate Model with Neural Closures

AI 导读

AEGIS 是一个模块化可微分火星气候模型,将火星大气物理与 Dinosaur 动力核心耦合,并提供神经闭合接口。它完成稳定的十个火星年模拟,复现季节性 CO₂ 循环且保持 CO₂ 总量守恒,地表温度结构合理,地表气压随 MOLA 地形变化。耦合轨迹上的梯度与有限差分结果一致,可支持物理校准与神经网络训练。

正文 · 原文

View PDF HTML (experimental)

Abstract: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.
Subjects: Earth and Planetary Astrophysics (astro-ph.EP); Instrumentation and Methods for Astrophysics (astro-ph.IM); Artificial Intelligence (cs.AI)
Cite as: arXiv:2610.04081 [astro-ph.EP]
  (or arXiv:2610.04081v2 [astro-ph.EP] for this version)
  https://doi.org/10.48550/arXiv.2610.04081

arXiv-issued DOI via DataCite

Submission history

From: Sameera Sudarshan Kashyap [view email]
[v1] Fri, 2 Oct 2026 21:45:37 UTC (1,463 KB)
[v2] Tue, 17 Nov 2026 07:15:56 UTC (1,899 KB)

来源:arXiv cs.AI · arxiv.org