发表机构
University of California, Los Angeles; Hong Kong University of Science and Technology; University of Lausanne; Allen Institute for AI; University of Colorado Boulder; California Institute of Technology; NOAA/Geophysical Fluid Dynamics Laboratory; Google Research; New York University（加州大学洛杉矶分校; 香港科技大学; 洛桑大学; 艾伦人工智能研究所; 科罗拉多大学博尔德分校; 加州理工学院; NOAA地球物理流体动力学实验室; 谷歌研究院; 纽约大学）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
本文回顾深度学习天气预报进展，提出AI气候建模需显式纳入外强迫并经受分布外检验，通过自回归模拟器与混合模型实现低成本局部灾害评估。
AI 中文摘要
深度学习在天气预报领域取得了快速进展：基于大气再分析资料训练的自回归模型，如今在临近预报、中期预报以及次季节到季节预报的提前时间尺度上，已能与动力模型相媲美，并以更低的成本生成校准良好的集合预报。我们回顾了这些进展，并探讨将其扩展到气候时间尺度的可能性，在此尺度上，挑战从初始条件技巧转向在改变的外强迫下产生可靠的统计响应。基于人工智能的气候预测系统必须对通常超出观测记录的驱动因子（如温室气体、土地利用变化）产生可信的强迫响应。我们为人工智能在气候建模中提出两项最低要求：（i）外部强迫因子必须足够明确地进入模型，以支持它们独立变化的干预实验；（ii）必须在分布外情景下进行鲁棒性压力测试，包括极端事件和反事实轨迹。利用领先的人工智能自回归模拟器和混合物理-人工智能模型，我们识别了发展和耦合方面的挑战。将这些模型报告的吞吐量与移植到GPU上的动力模型进行比较，突显了人工智能如何通过仅以所需分辨率推进目标变量并使用更长的时间步长，而非积分完整的高频多变量状态，来缩短求解时间。多样的人工智能降尺度策略可以部分替代显式的细尺度分辨率，为跨预测时间尺度的低成本局部灾害评估铺平道路。
英文摘要
Deep learning has made rapid advances in weather forecasting: autoregressive models trained on atmospheric reanalyses now rival dynamical models across nowcasting, medium-range, and subseasonal-to-seasonal lead times, producing well-calibrated ensemble forecasts at reduced cost. We review these advances and consider their extension to climate horizons, where the challenge shifts from initial-condition skill to producing reliable statistical responses under altered forcings. AI-powered climate prediction systems must produce credible forced responses to drivers (e.g., greenhouse gases, land-use change) typically outside the observed record. We propose two minimum requirements for AI in climate modeling: (i) external forcing agents must enter explicitly enough to support interventions in which they vary independently; and (ii) robustness must be stress-tested in out-of-distribution regimes, including extremes and counterfactual trajectories. Using leading AI autoregressive emulators and hybrid physics-AI models, we identify development and coupling challenges. Comparing the reported throughput of these models with that of GPU-ported dynamical models highlights how AI can reduce time-to-solution by advancing only the target variables at the required resolution and using longer time steps, rather than integrating a full high-frequency, multivariate state. Diverse AI downscaling strategies can partially substitute for explicit fine-scale resolution, paving the way toward inexpensive local hazard assessment across prediction horizons.
Comments33 pages, 10 figures. Submitted to "Science Advances"
