There is a newer version of the record available.

Published December 18, 2023 | Version v0.1.0
Software Open

FuXi-S2S

Authors/Creators

Description

This is the official repository for the FuXi-S2S paper.

A machine learning model that outperforms conventional global subseasonal forecast models

by Lei Chen, Xiaohui Zhong, Jie Wu, Deliang Chen, Shangping Xie, Qingchen Chao, Chensen Lin, Zixin Hu, Bo Lu, Hao Li, Yuan Qi

 

Installation

The Google Drive folder contains the FuXi-S2S model and sample input data, all of which are essential resources for this study. Currently, access to these resources is limited. For inquiries regarding the Google Drive link, kindly reach out to Professor Li Hao at the following email address: lihao_lh@fudan.edu.cn.

  1. Install xarray  
    conda install -c conda-forge xarray dask netCDF4 bottleneck
  1. Install pytorch   
    conda install pytorch pytorch-cuda=11.7 -c pytorch -c nvidia
 

Usage

 

python inference.py \
    --model model/fuxi_s2s.pt \
    --input data/input.nc \
    --total_step 42 \
    --total_member 51 \
    --save_dir output;
 

Input preparation

The input.nc file contains preprocessed data from the origin ERA5 files. The file has a shape of (2, 76, 721, 1440), where the first dimension represents two time steps. The second dimension represents all variable and level combinations, named in the following exact order:

 

['z1000', 'z925', 'z850', 'z700', 'z600', 'z500', 'z400', 'z300',
'z250', 'z200', 'z150', 'z100', 'z50', 't1000', 't925', 't850',
't700', 't600', 't500', 't400', 't300', 't250', 't200', 't150',
't100', 't50', 'u1000', 'u925', 'u850', 'u700', 'u600', 'u500',
'u400', 'u300', 'u250', 'u200', 'u150', 'u100', 'u50', 'v1000',
'v925', 'v850', 'v700', 'v600', 'v500', 'v400', 'v300', 'v250',
'v200', 'v150', 'v100', 'v50', 'q1000', 'q925', 'q850', 'q700',
'q600', 'q500', 'q400', 'q300', 'q250', 'q200', 'q150', 'q100',
'q50', 't2m', 'd2m', 'sst', 'ttr', '10u', '10v', '100u', '100v',
'msl', 'tcwv', 'tp']

Files

tpys/FuXi-S2S-v0.1.0.zip

Files (8.2 kB)

Name Size Download all
md5:d3ed5d6cae6bab34535eb581837d6027
8.2 kB Preview Download

Additional details

Related works

Is supplement to
Software: https://github.com/tpys/FuXi-S2S/tree/v0.1.0 (URL)