Skip to content

A global data-driven high-resolution weather model

License

Notifications You must be signed in to change notification settings

Peter-Bloomberg/FourCastNet

 
 

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

6 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Global Data-driven High-resolution Weather Model

简体中文 | English

本项目在幻方萤火超算集群上用 PyTorch 实现并优化了 FourCastNet 全球AI气象预报模型,首次使得AI气象模型能够与欧洲中期天气预报中心(ECMWF)的传统物理模型,高分辨率综合预测系统(IFS),进行直接比较。

台风路径预测与真实路径比较

汽水浓度预测与真实情况比较

Requirements

Training

原始数据来自欧洲中期天气预报中心(ECMWF)提供的一个公开可用的综合数据集 ERA5 , 幻方AI将其进行了整理,合入 hfai.datasets 数据集仓库中, 使用参考 hfai文档

  1. 预训练

    提交任务至萤火集群,使用64张A100训练

     hfai python train/pretrain.py -- -n 8 -p 30

    本地运行:

     python train/pretrain.py
  2. 微调训练

    提交任务至萤火集群,在之前的预训练模型上微调

     hfai python train/fine_tune.py -- -n 8 -p 30

    本地运行:

     python train/fine_tune.py
  3. 降水模型训练

    提交任务至萤火集群,以主干模型的输出作为输入,训练降水模型

     hfai python train/precipitation.py -- -n 8 -p 30

    本地运行:

     python train/precipitation.py

Citation

@article{pathak2022fourcastnet,
  title={Fourcastnet: A global data-driven high-resolution weather model using adaptive fourier neural operators},
  author={Pathak, Jaideep and Subramanian, Shashank and Harrington, Peter and Raja, Sanjeev and Chattopadhyay, Ashesh and Mardani, Morteza and Kurth, Thorsten and Hall, David and Li, Zongyi and Azizzadenesheli, Kamyar and others},
  journal={arXiv preprint arXiv:2202.11214},
  year={2022}
}

About

A global data-driven high-resolution weather model

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • Python 100.0%