HyPhAICC v1.0: a hybrid physics–AI approach for probability fields advection shown through an application to cloud cover nowcasting

El Montassir, Rachid; Pannekoucke, Olivier; Lapeyre, Corentin

doi:https://doi.org/10.5194/gmd-17-6657-2024

Articles | Volume 17, issue 17

https://doi.org/10.5194/gmd-17-6657-2024

© Author(s) 2024. This work is distributed under
the Creative Commons Attribution 4.0 License.

https://doi.org/10.5194/gmd-17-6657-2024

© Author(s) 2024. This work is distributed under
the Creative Commons Attribution 4.0 License.

Articles | Volume 17, issue 17

Model description paper

|

10 Sep 2024

Model description paper |

| 10 Sep 2024

HyPhAICC v1.0: a hybrid physics–AI approach for probability fields advection shown through an application to cloud cover nowcasting

Rachid El Montassir, Olivier Pannekoucke, and Corentin Lapeyre

Data sets

A sample of the training data used in the paper "A Hybrid Physics-AI (HyPhAI) approach for probability fields advection: Application to cloud cover nowcasting" European Organisation for the Exploitation of Meteorological Satellites https://doi.org/10.5281/zenodo.10642094

Model code and software

relmonta/hyphai: Update paper information (v1.1.1) Rachid El Montassir https://doi.org/10.5281/zenodo.11518540

Pre-trained HyPhAICCast-1, HyPhAICCast-2 and U-Net's weights Rachid El Montassir et al. https://doi.org/10.5281/zenodo.10393415

Interactive computing environment

relmonta/hyphai: Update paper information (v1.1.1) Rachid El Montassir https://doi.org/10.5281/zenodo.11518540

Video supplement

HyPhAICCast-1 2-hour forecast on 01/01/2021 at 12:00 p.m. Rachid El Montassir et al. https://doi.org/10.5281/zenodo.10375284

Short summary

This study introduces a novel approach that combines physics and artificial intelligence (AI) for improved cloud cover forecasting. This approach outperforms traditional deep learning (DL) methods in producing realistic and physically consistent results while requiring less training data. This architecture provides a promising solution to overcome the limitations of classical AI methods and contributes to open up new possibilities for combining physical knowledge with deep learning models.