graphical_summary

Precipitation Nowcasting

FUSEDCAST - fused data approach to the nowcasting of severe weather

Within the EUMETSAT Research Fellowship project FUSEDCAST (a fused data approach to the nowcasting of severe weather), Çağlar Küçük conducted research on AI-based solutions for precipitation nowcasting algorithms, aiming to narrow the gap between computer science and meteorology.

Using the European Weather Cloud, Çağlar developed two models for distinct purposes:

  1. A satellite-to-radar nowcasting model for predicting weather radar data based on satellite observations. This model is particularly useful for data-scarce regions, functioning as a radar emulator. (See publication and repository)
  2. A multi-source-to-precipitation nowcasting model for accurate precipitation prediction by integrating data from space- and ground-based observations with traditional numerical weather prediction outputs. This model is designed specifically for convective precipitation nowcasting in data-rich regions, such as much of the Alpine region in Europe. (See preprint and repository)

You can find more details from the presentation at the 7th EWC User Workshop 2024.

Organisation

Geosphere Austria
Hosted by
EUMETSAT logo