New machine learning method for short-term forecasting of heavy rainfall

In urban areas, heavy rainfall can lead to flash floods and significant losses, which is why there is an acute need for accurate forecasting. The study developed a new, machine learning-based short-term forecasting model for precipitation.

Photo: Adobe Stock

The model is particularly suited for predicting strong and rapidly developing incidents of heavy rainfall.

Further information:

Researcher Jenna Ritvanen, Finnish Meteorological Institute, jenna.ritvanen@fmi.fi

Scientific article is available on IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing.

Reference: Ritvanen, J., Harnist, B., Aldana, M., Mäkinen, T., and Pulkkinen, S. (2023): Advection-FreeConvolutionalNeural Network for ConvectiveRainfallNowcasting, IEEE Journal of SelectedTopics in Applied Earth Observations and Remote Sensing, 16, 1654–1667, https://doi.org/10.1109/JSTARS.2023.3238016.

Science-newsResearchRainMachine learning