---
license: apache-2.0
language:
- en
pretty_name: ClimateSuite
tags:
- climate
- climate-emulation
- earth-system-models
- zarr
- spatiotemporal
---

<h5 align="center">If you use ClimateSuite, please ❤️ like this dataset and ⭐ star <a href="https://github.com/stanfordmlgroup/spf">the SPF GitHub repository</a> for updates.</h5>

# ClimateSuite

ClimateSuite is a large collection of Earth system model simulations for climate emulation. It was curated for [Spatiotemporal Pyramid Flow Matching for Climate Emulation](https://arxiv.org/abs/2512.02268), which introduces Spatiotemporal Pyramid Flows (SPF) for efficient probabilistic climate emulation across temporal and spatial scales.

This repository stores ClimateSuite as archive-sharded Zarr data in the following format:

```text
grid.nc
inputs/<timescale>/*.zarr
outputs/<climate_model>/<timescale>/*.zarr
```

The current release includes decadal, yearly, and monthly data for ten climate models: `BCC-CSM2-MR`, `CESM2`, `CESM2-WACCM`, `CMCC-CM2-SR5`, `CMCC-ESM2`, `GFDL-ESM4`, `IPSL-CM6A-LR`, `MRI-ESM2-0`, `NorESM2-LM`, and `UKESM1-0-LL`.

### Downloading the data

The recommended path for SPF users is the repository helper, which downloads matching archive shards and extracts them automatically:

```python
from emulator.src.data.huggingface import download_climate_dataset

cache_dir = "/path/to/fast/cache"

# Download and extract the full dataset.
data_dir = download_climate_dataset(
    "jirvin16/ClimateSuite",
    cache_dir=cache_dir,
)

# Download and extract a specific model/timescale subset.
data_dir = download_climate_dataset(
    "jirvin16/ClimateSuite",
    cache_dir=cache_dir,
    climate_models=["NorESM2-LM"],
    timescales=["yearly"],
)
```

You can also download raw archive shards directly with `huggingface_hub`:

```python
from huggingface_hub import snapshot_download

cache_dir = "/path/to/fast/cache"

# Full archive download.
snapshot_download(
    repo_id="jirvin16/ClimateSuite",
    repo_type="dataset",
    cache_dir=cache_dir,
)

# Specific subset: shared yearly inputs and yearly NorESM2-LM outputs.
snapshot_download(
    repo_id="jirvin16/ClimateSuite",
    repo_type="dataset",
    cache_dir=cache_dir,
    allow_patterns=[
        "README.md",
        "manifest.json",
        "grid.nc",
        "inputs/yearly/*.zarr.tar.zst",
        "outputs/NorESM2-LM/yearly/*.zarr.tar.zst",
    ],
)
```

### Relevant Links

- **Paper:** https://arxiv.org/abs/2512.02268
- **Code:** https://github.com/stanfordmlgroup/spf

### Citation

```bibtex
@inproceedings{irvin2026spatiotemporal,
  title={Spatiotemporal Pyramid Flow Matching for Climate Emulation},
  author={Irvin, Jeremy Andrew and Han, Jiaqi and Wang, Zikui and Alharbi, Abdulaziz and Zhao, Yufei and Bayarsaikhan, Nomin-Erdene and Visioni, Daniele and Ng, Andrew Y. and Watson-Parris, Duncan},
  booktitle={Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)},
  year={2026}
}
```

## Contact

jirvin16@cs.stanford.edu