PeakWeather

PeakWeather is a high-resolution, benchmark-ready dataset for spatiotemporal weather modeling.

Key Features

  • High-resolution observations: 10-minute interval data spanning 2017-2025 over 302 SwissMetNet stations distributed across Switzerland

  • Multiple variables: Temperature, pressure, humidity, wind, radiation, precipitation and more

  • Topographic descriptors: Elevation, slope, aspect, and surface roughness to describe the Swiss complex terrain

  • NWP baselines: Ensemble forecasts from ICON-CH1-EPS, the state-of-the-art numerical prediction model operational at MeteoSwiss

  • Ideal for: Many tasks including time series forecasting, missing data imputation, virtual sensing, and graph structure learning

Stations

Quickstart

Install the dataset library. The base package handles the station measurements and the NWP predictions. It can be install by running the following command

pip install git+https://github.com/MeteoSwiss/PeakWeather.git # Install base package

If access to the topographical descpitors is desired, then there are additional required libraries that can be installed via

pip install "peakweather[topography] @ git+https://github.com/MeteoSwiss/PeakWeather@main" # Install with extras

When the PeakWeatherDataset is instantiated for the first time, the weather data is downloaded.

from peakweather.dataset import PeakWeatherDataset
# Download the data in the current working directory
ds = PeakWeatherDataset(root=<PATH_TO_DATA>)

For detailed usage and parameter descriptions, please refer to documentation of the PeakWeatherDataset class, which provides extended documentation on its functionality and options.

Citation

If you use PeakWeather in your research, please cite:

@misc{zambon2025peakweather,
   title={PeakWeather: MeteoSwiss Weather Station Measurements for Spatiotemporal Deep Learning},
   author={Zambon, Daniele and Cattaneo, Michele and Marisca, Ivan and Bhend, Jonas and Nerini, Daniele and Alippi, Cesare},
   year={2025},
   eprint={2506.13652},
   archivePrefix={arXiv},
   primaryClass={cs.LG},
   url={https://arxiv.org/abs/2506.13652},
}

Documentation

Examples

API