Open Access Publication of the Month – Kaveh Patakchi Yousefi (IBG-3) et al.

6 November 2024
To encourage us all to broaden our horizons, each month the Central Library selects one open access publication from the JuSER publications portal to be featured in the newsletter of Forschungszentrum Jülich.
In the open access publication of the month, Kaveh Patakchi Yousefi, Alexandre Belleflamme, Klaus Goergen and Stefan Kollet (all IBG-3) write about assessing the feasibility and impact of using U-Net, a type of convolutional neural network (CNN) artificial intelligence method, to enhance precipitation forecast within an integrated quasi-operational hydrological model (based on ParFlow/CLM) for central Europe.
CNNs are used for recognizing patterns in spatial data, making them well-suited for correcting weather forecast errors. They use U-Net to learn and correct the space-time errors of HRES weather forecasts provided by ECMWF (European Centre for Medium-Range Weather Forecasts) with near real-time data from H-SAF (Satellite Application Facility on Support to Operational Hydrology and Water Management) used as reference. The corrected precipitation dataset was used in an integrated hydrological model to assess its impact. The study highlights the potential of this method in improving precipitation forecasts, though high-quality reference data is very essential for this correction.
The open access publication entitled "Impact of deep learning-driven precipitation corrected data using near real-time satellite-based observations and model forecast in an integrated hydrological model" was published in the journal Frontiers in Water.
JuSER publications portal – Impact of deep learning-driven precipitation corrected data using near real-time satellite-based observations and model forecast in an integrated hydrological model
Internet – Institute of Bio- and Geosciences – Agrosphere (IBG-3)