The AtmoSwing (Analog Technique Model for Statistical Weather forecastING) software allows for real-time precipitation forecasting based on a downscaling method, the analogue technique. This method is used for hydro-meteorological forecasting in the framework of real-time flood management or electricity production. It allows for:

  • Automatic precipitation forecasting, 4-times a day
  • Displaying the forecast in a user-friendly GIS environment
  • Optimizing the analogue method for a provided timeseries.


Time Series for the Coming Days

Time series for the coming days for a given location can be quickly processed and displayed, along with the full probability distribution.

Data Visualization

The forecasts are displayed in a user-friendly GIS environment that allows for precipitation spatialization visualization. The software is interactive and handles several levels of synthesis or details.

Input Data

AtmoSwing is built in order to ease the integration of outputs from various global numerical weather prediction models. It focuses on precipitation forecasting, but can handle any other weather variable.

Data Exporting

Exports are either standard NetCDF files targeted for AtmoSwing Viewer, or XML files for integration in another tool, such as a web platform.


AtmoSwing is easy to use in operational forecasting and requires low IT infrastructure.

AtmoSwing is made of 3 tools: the Forecaster: automatically processes the forecast, the Viewer: displays the resulting files in a GIS environment, and the Optimizer: optimizes the method for a new region or a new weather variable. Terranum can provide support in various domains:

Specific development

There may be situations where AtmoSwing needs to be modified in order to meet the requirements of a certain project, or to model a specific weather variable. We are ready to make the necessary adjustments in AtmoSwing in order to meet your needs.

Operational forecasting

Terranum can provide assistance in setting up a physical or virtual server for operational forecasting with AtmoSwing.

Optimization of the method

We have various powerful optimization tools to adapt the method for a new region or a new weather variable. We have an exclusive global optimizer that relies on Genetic Algorithms, which revolutionizes the calibration of the method as it has been done until today.

Corrective or evolutive maintenance

AtmoSwing needs to continuously follow the changes in the providers protocols as well as evolutions of the method. We provide different levels of support and can take care for you of smooth migrations to new features.


Horton, P: AtmoSwing: Analog Technique Model for Statistical Weather forecastING and downscalING (v2.1.0). Geosci. Model Dev., 12(7), 2915–2940, 2019. doi:10.5194/gmd-12-2915-2019