| --- |
| license: mit |
| datasets: |
| - CEMAC/synthetic_lee_waves |
| metrics: |
| - mse |
| pipeline_tag: image-to-image |
| library_name: fastai |
| tags: |
| - climate |
| --- |
| |
| # Model Card for LeeWaveNet |
|
|
| <!-- Provide a quick summary of the model. --> |
|
|
| This repository contains four neural-network models, trained using [fastai](https://docs.fast.ai/), for detecting and determining characteristics of trapped lee waves using maps of 700 hPa vertical velocity as input. |
|
|
| * The base model [segmodel.pkl](https://huggingface.co/CEMAC/LeeWaveNet/blob/main/segmodel.pkl) generates a segmentation mask indicating where trapped lee waves are present. This model uses a U-Net architecture with Resnet-34 (pre-trained on ImageNet) as the encoder model. |
| * Three alternative model heads have been trained on synthetic data: [amplitude_0.0625.pkl](https://huggingface.co/CEMAC/LeeWaveNet/blob/main/amplitude_0.0625.pkl), [wavelength_0.125.pkl](https://huggingface.co/CEMAC/LeeWaveNet/blob/main/wavelength_0.125.pkl) and [orientation_0.25.pkl](https://huggingface.co/CEMAC/LeeWaveNet/blob/main/orientation_0.25.pkl). These predict the amplitude, wavelength and orientation of detected waves respectively. |
|
|
| For full details, please see the article by [Coney et al. (2023)](https://doi.org/10.1002/qj.4592). |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
