---
license: cc-by-4.0
language:
- en
tags:
- climate
- physics-informed ML
- surrogate modeling
pretty_name: >-
  Dataset for emulating the MPI-ESM1.2-LR ensemble mean annual temperature and
  precipitation
size_categories:
- n<1K
---

This is the repository for the Em-MPI dataset associated with 'The impact of internal variability on benchmarking deep learning climate emulators'. The total size is <10GB. 

To download the dataset run this Python script:
```
from huggingface_hub import snapshot_download # Requires pip install huggingface_hub

snapshot_download(repo_id='blutjens/em-mpi',
  repo_type='dataset', 
  local_dir='../em-mpi-data', # Local path to data 
)
```

Alternatively, one can download individual files on huggingface.com via:
```
# Click of 'Files and Versions'
# Click on the file you wish to download
# Click 'download'
```