AirCast-SR: A Foundation Model for Kilometer-Scale Atmospheric Super-Resolution via Latent Consistency Diffusion

Abstract: Operational weather prediction at kilometer scales remains computationally prohibitive for traditional numerical weather prediction (NWP) models, limiting forecast access for applications in energy, agriculture, and disaster management that require fine-grained spatiotemporal detail.

摘要： 对于传统的数值天气预报（NWP）模型而言，公里级的天气预报在计算上仍然极其昂贵，这限制了能源、农业和灾害管理等领域对预报数据的获取，而这些领域恰恰需要精细的时空细节。

Here we introduce AirCast-SR, a foundation model for atmospheric super-resolution that downscales global AI weather forecasts from 0.25 degree (~28 km) to 1 km horizontal resolution at hourly temporal resolution, producing 67-hour forecasts of eight coupled surface variables simultaneously.

在此，我们介绍了 AirCast-SR，这是一种用于大气超分辨率的基准模型。它能将全球 AI 天气预报从 0.25 度（约 28 公里）降尺度至 1 公里的水平分辨率，并保持小时级的时间分辨率，同时生成八个耦合地表变量的 67 小时预报。

EarthMind-SR employs a three-dimensional U-Net conditioned within a Latent Consistency Model (LCM) diffusion framework, trained on patch-based samples over the contiguous United States (CONUS) using GraphCast forecasts as input and NOAA’s Analysis of Record for Calibration (AORC) as the target.

EarthMind-SR 采用了一种在潜在一致性模型（LCM）扩散框架内进行条件约束的三维 U-Net 网络。该模型在美国本土（CONUS）的补丁样本上进行训练，以 GraphCast 的预报结果作为输入，并以美国国家海洋和大气管理局（NOAA）的记录分析（AORC）作为目标数据。

The model achieves near-zero bias across all variables and lead times, and its radial power spectral density analysis demonstrates preservation of fine-scale atmospheric structure at wavelengths of 10 km to 100 km where coarser models lose spectral power.

该模型在所有变量和预报时效上均实现了近乎零的偏差；其径向功率谱密度分析表明，在粗糙模型失去谱功率的 10 公里至 100 公里波长范围内，该模型仍能有效保留精细的大气结构。

We validate EarthMind-SR across three CONUS case studies spanning winter, summer, and spring seasons, and demonstrate zero-shot global transferability over India and Germany using independent surface station observations without any retraining or fine-tuning.

我们在涵盖冬季、夏季和春季的三个美国本土案例研究中验证了 EarthMind-SR，并利用印度和德国的独立地面站观测数据，在无需任何重新训练或微调的情况下，展示了其零样本的全球迁移能力。

As an open-weights foundation model, EarthMind-SR establishes a new paradigm for kilometer-scale AI weather prediction and provides a platform for regional fine-tuning, distillation, and downstream applications in climate services and hazard forecasting.

作为一种开放权重的基准模型，EarthMind-SR 为公里级 AI 天气预报确立了新的范式，并为气候服务和灾害预报领域的区域微调、蒸馏及下游应用提供了一个平台。