File size: 3,038 Bytes
8c20e29
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
---
license: apache-2.0
language:
- en
tags:
- OneScience
- Earth Science
- Extreme Weather
- Semantic Segmentation
- Tropical Cyclone
- Atmospheric River
frameworks: PyTorch
---

<p align="center"><strong><span style="font-size: 30px;">ClimateNet</span></strong></p>

# Model Introduction

ClimateNet is an expert-labeled extreme-weather dataset and pixel-level segmentation model for tropical cyclones and atmospheric rivers.

Paper: ClimateNet: an expert-labeled open dataset and deep learning architecture for enabling high-precision analyses of extreme weather  
https://doi.org/10.5194/gmd-14-107-2021

# Model Description

The model was proposed by teams from LBNL, UC Berkeley, ETH Zurich, NVIDIA, NCAR, and collaborators. It was trained with four-channel CAM5.1 fields and expert segmentation masks. DeepLabv3+ supports tropical-cyclone and atmospheric-river detection and conditional precipitation analysis.

# Use Cases

| Use Case | Description |
| :---: | :--- |
| Extreme segmentation | Identify background, TC, and AR pixels. |
| Climate scenarios | Transfer segmentation to warming experiments. |
| Conditional precipitation | Extract event-conditioned precipitation statistics. |
| ModelScope/OneCode execution | Validate data, training, inference, segmentation metrics, and visualization. |
| Multi-GPU training | Start multi-process training through `torchrun`. |

# Usage Instructions

```bash
hf download OneScience-Group/ClimateNet --local-dir ./ClimateNet
cd ClimateNet
```

### Environment Dependencies

**Hardware Requirements**

- A GPU or DCU is recommended.
- A CPU can be used for connectivity validation with the default small-sample configuration.
- DCU users should install DTK 25.04.2 or a compatible OneScience-recommended version first.

**DCU Environment**

```bash
# Activate DTK and Conda first
conda create -n onescience311 python=3.11 -y
conda activate onescience311
pip install onescience[earth-dcu] -i http://mirrors.onescience.ai:3141/pypi/simple/ --trusted-host mirrors.onescience.ai
```

**GPU Environment**

```bash
# Activate Conda first
conda create -n onescience311 python=3.11 -y libstdcxx-ng=12 libgcc-ng=12 gcc_linux-64=12 gxx_linux-64=12
conda activate onescience311
pip install onescience[earth-gpu] -i http://mirrors.onescience.ai:3141/pypi/simple/ --trusted-host mirrors.onescience.ai
```

```bash
python scripts/fake_data.py
python scripts/train.py
torchrun --standalone --nproc_per_node=2 scripts/train.py
python scripts/inference.py
python scripts/result.py
```
Training uses weighted cross-entropy. Inference returns finite class probabilities and evaluation reports per-class and mean IoU.
## Trained Weights
No weights are bundled under `weight/`. The authors provide trained models and data at https://portal.nersc.gov/project/ClimateNet/.
# Citation and License
This repository is an independent engineering reproduction of the public ClimateNet specifications.

The original paper is licensed under CC BY 4.0; official models, code, and data retain their respective terms.