Add dataset card, link to paper
#2
by
nielsr
HF Staff
- opened
README.md
ADDED
|
@@ -0,0 +1,25 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
task_categories:
|
| 3 |
+
- image-segmentation
|
| 4 |
+
license: mit
|
| 5 |
+
tags:
|
| 6 |
+
- climate
|
| 7 |
+
- satellite imagery
|
| 8 |
+
---
|
| 9 |
+
|
| 10 |
+
# ClimateBench-M: A Multi-modal Climate Data Benchmark
|
| 11 |
+
|
| 12 |
+
[**📖 Paper**](https://arxiv.org/abs/2404.00225)
|
| 13 |
+
|
| 14 |
+
ClimateBench-M is the first multi-modal climate benchmark designed to support the development of artificial general intelligence (AGI) in climate applications. It aligns data across three critical modalities at a unified spatio-temporal resolution:
|
| 15 |
+
|
| 16 |
+
1. Time-series climate variables from ERA5.
|
| 17 |
+
2. Extreme weather event records from NOAA.
|
| 18 |
+
3. Satellite imagery from NASA HLS.
|
| 19 |
+
|
| 20 |
+
This dataset is used in the paper: https://huggingface.co/papers/2504.07394.
|
| 21 |
+
|
| 22 |
+
To facilitate broad accessibility and reproducibility, **ClimateBench-M** is hosted on [🤗 Hugging Face Datasets](https://huggingface.co/docs/datasets/index). Due to its size and multi-modal nature, the dataset is split into two parts:
|
| 23 |
+
|
| 24 |
+
- ClimateBench-M-TS: Climate time series with extreme events aligned and labeled.
|
| 25 |
+
- ClimateBench-M-IMG: Satellite imagery data
|