Climate Physicists Face the Ghosts in Their Machines: Clouds

2026-02-20 · Source: artificial intelligence – Quanta Magazine · Field: Science & Research — Environmental Science & Earth Systems, Mathematics & Computational Sciences, Research Methodology & Innovation · Depth: Intermediate, short

Summary

The critical role of clouds in climate models is highlighted, with current simulations showing that over half of the variation in warming predictions (ranging from 2 to 6 degrees Celsius) stems from how clouds are treated. Despite advancements like the Frontier supercomputer running Navier-Stokes-based models at 3-kilometer resolution, directly modeling clouds, which can span meters, requires 100 billion times current computational power. Consequently, physicists currently estimate cloud influence by adding nonphysical "parameters" to equations, a process that relies on intuition and patchy data. To address this, Tapio Schneider established the Climate Modeling Alliance (CLIMA) in 2019, aiming to automate parameter selection using AI. This requires extensive cloud data, which is being generated through large-eddy simulations (LES) in collaboration with Google, as real-world data collection is limited and LES is computationally intensive.

Key takeaway

For AI Scientists developing climate models, the challenge of accurately representing cloud behavior is paramount. Current methods rely on imprecise parameter estimation, leading to significant prediction variability. You should explore integrating AI-driven parameterization with high-resolution large-eddy simulations to enhance model accuracy and reduce reliance on human intuition, thereby improving the reliability of future climate projections.

Key insights

Cloud modeling is a critical bottleneck in climate prediction, driving new AI-physics hybrid approaches.

Principles

• Cloud treatment dictates over 50% of climate model prediction variance.
• Direct cloud simulation is computationally infeasible with current tech.

Method

CLIMA integrates traditional physics with AI to automate parameter selection for cloud effects, using large-eddy simulations (LES) to generate necessary training data for machine learning models.

In practice

• Utilize AI for parameter optimization in complex physics models.
• Employ large-eddy simulations to generate synthetic data for training.

Topics

• AI in Climate Modeling
• Cloud Dynamics
• Climate Prediction
• Large-Eddy Simulations
• Navier-Stokes Equations

Best for: AI Scientist, AI Researcher, Research Scientist, AI Engineer

Related on AIssential

• Can AI models reliably forecast extreme weather events? — AI weather models offer speed but require rigorous, standardized testing for extreme event reliability.
• FLAME 2023: Diving into the Future of Fluid Dynamics and Machine Learning — AI techniques are transforming simulation by integrating physics-based modeling with data-driven machine learning.
• Sampling sea state using a diffusion model — Diffusion models offer a computationally efficient path to probabilistic global sea state estimation.
• How NVIDIA's Earth-2 uses AI to Accurately Predict Weather — NVIDIA Earth-2 provides an open, GPU-accelerated AI stack for faster, more accurate weather and climate forecasting.
• Can AI help coastal cities prepare for rising seas and extreme events? — A fast AI emulator accurately predicts extreme storm surges by learning from physics-based models, enhancing coastal flood risk assessment.
• Learning the coupled dynamics of global climate modes — UniCM unifies global climate mode forecasting by learning coupled ocean-atmosphere dynamics and inter-mode interactions, revealing emergent predictability.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by artificial intelligence – Quanta Magazine.